Case Study

HOW TO DO YOUR
CASE
Study
Sara Miller McCune founded SAGE Publishing in 1965 to support
the dissemination of usable knowledge and educate a global
community. SAGE publishes more than 1000 journals and over
800 new books each year, spanning a wide range of subject areas.
Our growing selection of library products includes archives, data,
case studies and video. SAGE remains majority owned by our
founder and after her lifetime will become owned by a charitable
trust that secures the company’s continued independence.
Los Angeles | London | New Delhi | Singapore | Washington DC | Melbourne

Third Edition
HOW TO DO YOUR
CASE
Study
Gary
Thomas
SAGE Publications Ltd © Gary Thomas 2021
1 Oliver’s Yard
55 City Road This edition first published 2021
London EC1Y 1SP
Apart from any fair dealing for the purposes of research,
SAGE Publications Inc. private study, or criticism or review, as permitted under the
2455 Teller Road Copyright, Designs and Patents Act, 1988, this publication
Thousand Oaks, California 91320 may not be reproduced, stored or transmitted in any form,
or by any means, without the prior permission in writing of
SAGE Publications India Pvt Ltd the publisher, or in the case of reprographic reproduction,
B 1/I 1 Mohan Cooperative Industrial Area in accordance with the terms of licences issued by the
Mathura Road Copyright Licensing Agency. Enquiries concerning
New Delhi 110 044 reproduction outside those terms should be sent to the
publisher.
SAGE Publications Asia-Pacific Pte Ltd
3 Church Street
#10-04 Samsung Hub
Singapore 049483
Editor: Jai Seaman

Library of Congress Control Number: 2020941944
Senior assistant editor: Charlotte Bush
Assistant editor, digital: Sunita Patel
British Library Cataloguing in Publication data
Production editor: Katherine Haw
Copyeditor: Neville Hankins
Proofreader: Rebecca Storr
Marketing manager: Ben Griffin-Sherwood
Cover design: Shaun Mercier A catalogue record for this book is available from the
Typeset by: C&M Digitals (P) Ltd, Chennai, India British Library
Printed in the UK
ISBN 978-1-5297-0496-9
ISBN 978-1-5297-0495-2 (pbk)
At SAGE we take sustainability seriously. Most of our products are printed in the UK using responsibly sourced
papers and boards. When we print overseas we ensure sustainable papers are used as measured by the PREPS
grading system. We undertake an annual audit to monitor our sustainability.
CONTENTS
About the author x
Preface xi
Online resources xiii
A roadmap for conducting a case study, using this book xiv
STAGE 1 GETTING YOUR BEARINGS xviii
1 What is a case study? 3

Is the case study scientific? 7
Some definitions 9
What is a case? 12
The subject and object of a case study 14
Forms of case study…and six steps to follow in devising your own 21
If you take only one thing from this chapter, take this … 24
Reflective activity 25
Further reading 25
2 The case study and research design 27

First things first: your purpose 27
Next, your question 28
Literature review 30
Using the literature and a storyboard to help you design your case study 31
This case study in the typology 38
How to go from idea to question to case study 39
Questions and different approaches to research 44
Design frames and methods 47
Further reading 48
3 The whole is more than the sum of the parts: seeing a

complete picture 49
Break things down or see them as wholes? 49
Gestalt psychology 52
vi HOW TO DO YOUR CASE STUDY
Dramas, theatres and stages 54

Ecological psychology 56
Systems thinking 59
Further reading 64
4 Rigour and quality in your case study: what is important? 66

Is the ‘sample’ important in case study? 66
Do I have to worry about reliability and validity in a
case study? 68
Triangulation71
Positionality73
Generalisation73
Theory76
Finding or regularising 79
Quality80
Further reading 83
5 Ethics 85
Your participants 85
Vulnerable groups 86
Participants’ activity 87
Deception or concealment 88
Confidentiality and anonymity 88
Data security and stewardship 90
Consent90
Risk92
Contacting participants 98
Ethics and social media 98
Care for your participants – and yourself 99
Where do I put discussion about ethics in my case study write-up? 99
Getting clearance – ethical review 99
Further reading 103
CONTENTS vii
STAGE 2 CHOOSING A CASE AND STRUCTURING YOUR STUDY 106
6 Different kinds of case studies: selecting a subject for

your case study 109
How do you select your case study subject? 110
Same starting points, different paths – there is no single right way 122
The typology in practice 123
Further reading 125
7 Your purpose: thinking about the object of your study 127

Intrinsic128
Instrumental128
Evaluative129
Explanatory131
Exploratory135
Further reading 142
8 Your approach: theory testing or theory building;

interpretation or illustration 144
Theory testing or theory building 145
Building a theory 145
Testing a theory 149
Drawing a picture – illustrative–demonstrative 152
Interpretative159
Experimental165
Further reading 172
9 Your process: the shape, style and manner of your case study 174
The single case 175
Time as a dimension of the case study 178
The multiple or collective or comparative case (or cross-case analysis) 186
viii HOW TO DO YOUR CASE STUDY
Nested case studies 191

Parallel and sequential studies 195
Further reading 199
STAGE 3 COLLECTING EVIDENCE, ANALYSING AND WRITING IT UP 200
10 Out in the field: some ways to collect data and evidence 203
Interviews205
Accounts206
Diaries207
Group interviews and focus groups 210
Interrogating documents 211
Questionnaires212
Observation215
Image-based methods 217
Measurements and tests 218
Official statistics and other numerical data 219
Using social media 220
Further reading 222
11 Analysis: a toolkit for analysing and thinking in case study 223

Interpretative inquiry: eliciting themes 224
Sociograms236
Systems thinking 238
Drawing storyboards – the nuts and bolts 240
Developing your theory 242
Using narrative 248
Think drama 254
Being intuitive and imaginative 256
Further reading 260
CONTENTS ix
12 Writing your study 262

Structure262
Writing up your case study 265
Two examples of good analysis, argument and writing 270
Some rules for writers 275
Remember you have an audience 277
… and this – a final thought 279
Further reading 280
Other reading 283
References 284
Index 294
ABOUT THE AUTHOR
Gary Thomas began his working life as a teacher, then he
became a teacher, then an educational psychologist, then
a professor of education at the University of Birmingham,
where his teaching, research and writing now focus on
inclusive education and the methods used in social science
research.
He has led a wide range of research projects and has
received awards from the AHRC, the ESRC, the Nuffield
Foundation, the Leverhulme Trust, the Department for
Education, charities such as Barnardo’s and the Cadmean
Trust, local authorities and a range of other organisations. He has written or edited
20 books and lots of boring academic articles.
He has two grown-up daughters, one small daughter and two grandchildren. He
likes, in alphabetical order, cats, chess, cycling, dogs, gardening and writing. He
dislikes Chelsea tractors, pomposity and people who try to make things sound more
complicated than they are (in that order).
Despite supporting Aston Villa football club, he maintains an optimistic outlook
on life.
PREFACE
I’ve been delighted with the reception of the first and second editions of How to Do
Your Case Study. This third edition contains much more material on ethics, risk and
data collection, new examples, and – among other changes – has new sections on
critical thinking, reflective journals, social media, coding and discourse analysis. As
in the previous two editions, my main aim has been to offer clear advice to students
on how best to conduct a case study inquiry, with lots of examples of how it can be
used most fruitfully.
In my examples I have drawn on instances of real or possible case studies that
could be done by real students doing real inquiries. I have done this rather than
use high-profile, well-known case studies carried out by teams of sociologists, each
taking several years to complete. These iconic examples can provide pointers and
flashes of insight that can inspire and give ideas, but, if I read too many of these,
I find I can become daunted rather than inspired, thinking, ‘How on earth could
I ever do anything like this?’ So, the examples I have given are, in the main, real,
adapted (and many thanks to the students from whom I have stolen them) or
sometimes imagined.
I’ve assumed people will be reading this book because they want to do a case
study, not because they want to find out about the case study method per se. There
must, I suppose, be a very limited number of research methodologists who are
interested in the case study in and of itself, but my assumption is that these are not
the principal audience for this book. Looking at the books available on the topic,
though, it is almost as if it were the other way round. It is almost as though authors
are writing books on case studies as if most people want to pursue an academic
interest in the case study method rather than want to find out about it so that they
can do one better.
I’ve also wanted in this book to provide not just a primer on the case study but
also something rather more eclectic than is to be found in most of the books avail-
able on the topic. In my experience, nearly all of the academic writing on the case
study has been done by sociologists seeing it as a form of inquiry for sociology
alone. While the case study has a long and fascinating tradition in sociology, it
is not owned by sociology. Indeed, as I’ve noted, the method can be used almost
anywhere, and I try to make this point in the book by taking examples from differ-
ent areas of inquiry. I also try to show how different insights about holistic inquiry
from different academic domains can have wider application and how ideas from
across a spectrum of disciplines have commonality.
xii HOW TO DO YOUR CASE STUDY
I should note that parts of Chapter 4 in this book have been published in the jour-
nals Qualitative Inquiry and the Oxford Review of Education (Thomas, 2011a, 2011b).
There are too many colleagues, friends and students to thank by name, but I
do really want to thank them all for their ideas and time. Thanks especially to
Jai Seaman at SAGE, who commissioned this third edition and coordinated read-
ers’ comments to provide such an enormously helpful plan of action for the new
volume. And thanks to those anonymous reviewers for their valuable comments.
Many thanks to all of Jai’s colleagues at SAGE for their help in putting the new
book together, especially Katherine Haw, Neville Hankins and their colleagues in
the production department at SAGE.
ONLINE RESOURCES
Discover your textbook’s online resources!
The third edition of How to Do Your Case Study offers a range of online resources
to help you confidently conduct a case study. Find them at: https://study.sage
pub.com/thomascasestudy3e
Videos from author Gary Thomas give you friendly guidance on important
considerations like how to structure your case study.
Checklists help you track your progress throughout the research process and make
sure you don’t miss a step.
Template consent forms that you can tailor to your own case study give you a
strong starting point for doing ethically sound research.
Weblinks showcase handy resources on a range of topics, from doing case study
research with social media to thinking critically and conducting ethical research.
Selected journal articles discussing key topics such as critical thinking help you
expand your knowledge and reinforce your learning.
A ROADMAP FOR
CONDUCTING A CASE
STUDY, USING THIS
BOOK
There are three stages to conducting a case study.
1 F
irst you need to know why you are doing a case
study. Is it really the best way of addressing your
research questions? What can it offer?
2 S
econd, you need to know how case studies are
typically structured and how best to choose your
case study subject.
3 T
hird, you need to know about processes for
collecting data and analysing them.
So, this book is organised by these stages. Here’s

the roadmap:
Road
map
STAGE 1. GETTING YOUR BEARINGS
Chapters 1 to 5 cover:
What a case is study all about

What a case study is good for
STAGE 2. CHOOSING A CASE AND

STRUCTURING YOUR STUDY
How case studies are categorised

How you can identify a case
How you should structure your case study
STAGE 3. COLLECTING EVIDENCE,

ANALYSING AND WRITING IT UP
Ways to collect data for a case study

Ways to analyse those data
Writing up the study
STAGE 1
GETTING YOUR BEARINGS

Chapters 1 to 5
What a case study is all about


Chapters 6 to 9


Chapters 10 to 12

In this first part of the book, I look at:
• w
hat the case study is, and the kinds of inquiries
for which it may be useful
• s ome ground rules for the conduct and design of

a case study
• w
hat we are trying to do with the case study as a
form of inquiry – that is, look at a situation in its
completeness and all its complexity
• w
hy, throughout the history of inquiry, people
have felt that it is a good idea to examine
something in its completeness
• s ome of the ways in which the success of a case

study may be measured
• the ethical dimensions of case study.

1
WHAT IS A CASE
STUDY?
At its best, the case study provides the most vivid, the most inspirational analysis
that an inquiry can offer. Einstein did it; Newton did it. Sociologists do it; psycholo-
gists do it. Doctors do it; teachers do it; lawyers do it; nurses do it. It is done across
the disciplinary and methodological spectrum, and even though this book is, first
and foremost, for those in the applied social sciences and humanities, I hope it will
also be useful to students and researchers in other fields.
The case study provides a form of inquiry that elevates a view of life in its com-
plexity. The contrasting view, of course, is that an inquiry is best conducted by
breaking life up into digestible, study-sized chunks. We all like easily digestible
food, but it is not often the most nutritious. It is the realisation that complexity in
social affairs is frequently indivisible, which has led to the case study having the
status of one of the most popular and most fertile design frames for researchers’
work.
The case study has branched out and blossomed from its origins, which were,
according to White (1992), in the professional training of lawyers at the Harvard
Law School in the nineteenth century. It was the appropriately named Christopher
Columbus Langdell, the first dean of that law school, who set the ball rolling in
1870. Actually, I’ve taken White’s word for it that it was Christopher Columbus
Langdell who named the case study method, but, of course, the ‘method’ is so
manifest, so obvious, that it is difficult to make an argument for its needing to be
‘pioneered’ or invented at all. Rather in the way that America was bound to have
been discovered with or without the voyage of the first, and rather more famous,
Christopher Columbus, the case study method would have been stumbled upon
even without the insights of his nineteenth-century namesake.
Others could quite legitimately claim to have been the first to develop a sys-
tematic method for looking at single cases, even if they did not think to call their
4 HOW TO DO YOUR CASE STUDY
method the ‘case method’ or ‘case study’. Jean Marc Gaspard Itard and his study
of Victor, the ‘wild boy of Aveyron’, at the beginning of the nineteenth century
could equally take the accolade, as could Frédéric Le Play in his celebrated stud-
ies of French working people in the middle of that same century. (Le Play used
a case method wherein he would invite himself to reside with ordinary working
families for extended periods until he got to know them intimately. Whether
or not his uninvited sojourns in the cottages of poor coalminers of the Jura
were exuberantly welcomed by his hosts is unrecorded – one can imagine some
Pythonesque tableaux – but his rich analyses were lauded in both political and
literary circles.)
If you use case study as the design frame for your research you will be concen-
trating on one thing: looking at it in detail. When you do a case study, you are
interested in that thing in itself, as a whole. I’m using the word ‘thing’ advisedly
since the thing may be a person, a group, an institution, a country, an event, a
period in time or whatever. You may be looking at the process of a medical diag-
nosis with one patient or you may be looking at relationships among a gang of
teenagers or you may be looking at one student’s learning in class. You may be
looking at a family or at IKEA’s development as a business or at the USA’s politi-
cal stance at the time of the Cuban Missile Crisis. Any of these ‘things’, these
phenomena, could form the subject of a case study. What is of interest is the
uniqueness of the thing and the thing in its completeness.
The main feature of your choice of case will
A case study is about the particular be the interest that you have in the subject of
rather than the general. the study. Why are you interested in it? It could
be interesting because …
• you are intimately connected with it – perhaps it is a child in your class, a patient
in your hospital or a project in your company
• it is a conspicuously good example of something in which you are interested – for
example, IKEA as an example of a successful business
• it is different from what is typical – it is an ‘outlier’ that shows something remark-
able by virtue of its difference from what seems to be the norm, such as, perhaps,
Kerala as a state in India where, despite widespread poverty, people have a far
longer life expectancy than they do elsewhere in India.
Singleness is the watchword with a case study.

Should you worry about this singleness? You may, after all, have heard people
being dismissive about ‘anecdotal evidence’ – in other words, evidence that comes
from just one situation or one event. Such anecdotal evidence is uncorroborated by
WHAT IS A CASE STUDY? 5
evidence from different experiences. Social science is about getting beyond what is
merely anecdotal, some would say.
Certainly, when we want to generalise – to say that this phenomenon has
occurred in this sample of people and we can generalise this finding to the wider
population – clearly we cannot use just one person’s experience or a single set of
uncorroborated observations as the basis for this. That is why such trouble has to
be taken in many kinds of research to establish the sample as being representative
of the wider population. If we want to generalise we need to make the basis of our
generalisation clear.
What we are talking about with a case study, though, is a different kind of inquiry
from those where generalisation is expected to follow. Here we are talking about
understanding how and why something might have happened or why it might be
the case. The assumption in a case study is that, with a great deal of intricate study
of one case, looking at your subject from many and varied angles, you can get closer
to the ‘why’ and the ‘how’. As Becker (2014: 3) puts it: ‘Everything present in or
connected to a situation I want to understand should be taken account of and made
use of. If it’s there, it’s doing something, however unimportant that thing seems, no
matter how unobtrusive it is.’
You will notice that although the assumptions
about valid processes of inquiry and analysis dif- Two important things about a case
study:
fer from those in many kinds of social research
(for we are not working from samples that enable 1. you drill down further
us to generalise), assumptions about the use of 2. you create a three-dimensional
picture – or what Foucault
evidence do not change. With a case study there called a ‘polyhedron of
is still the assumption that we must collect good intelligibility’.
evidence and lots of it. In the modern phrase, we
must ‘drill down’ as deep as we possibly can to get evidence, penetrating into every
nook and cranny, and squeezing out every little bit that can be found.
To mix our metaphors even more, we must look at our subject from many and
varied angles, to develop what the great historian–philosopher Michel Foucault
(1981: 4) called ‘a polyhedron of intelligibility’. By this he meant that inquiries
in the humanities and social sciences are too often one-dimensional, as if we are
looking at our subject just from one direction. In looking from several directions, a
more rounded, richer, more balanced picture of our subject is developed – we get a
three-dimensional view.
It is not just in the social sciences and the humanities that this multifaceted view
is valuable. In what are taken to be ‘harder’ sciences, such as biology, astronomy
and geology, and in many applied sciences, such as palaeoanthropology (the study
of humanoid fossils), the case study is one of the principal methods used.
Sometimes, this is called an idiographic approach. An idiograph is a little figure,

or a picture. The idea here is that we can do research by looking at this picture in
detail. We can better understand a phenomenon or a process by looking at this
particular picture, this idiograph, this case, in
Idiograph – a symbol, a drawing or detail, rather than looking at tens or hundreds
figure that stands for an idea (from or thousands of cases in not very much detail
the Greek ‘idea’ meaning … er …
‘idea’, and graphein meaning ‘to
and trying to generalise from that large num-
write’). ber. (I look in more detail at the idiographic
approach in Chapter 3.)
The key, I think is in what Flyvbjerg (2001: 132) calls ‘getting close to reality’.
By this he means keeping in contact with the subject of study and thinking with
your own experience and your own intelligence. It is this ‘staying real’ that the case
study is particularly good at encouraging, for it eschews methodological formulae
and endorses and stimulates a critical, creative approach to problem-solving.
Flyvbjerg puts the emphasis in this ‘getting close to reality’ on ‘little questions’
and ‘thick description’ (of which more in Chapter 11). We sometimes ignore these
little questions in the ‘Big Science’ view of what research should be like. Flyvbjerg
(2001: 133) quotes Nietzsche as saying that ‘all the problems of politics, of social
organization, and of education have been falsified through and through … because
one learned to despise “little” things, which means the basic concerns of life itself.’
As Flyvbjerg points out, small questions often lead to big answers.
For me, the truth of Flyvbjerg’s point is exemplified in one of the best case
studies I have ever read: A Glasgow Gang Observed (Patrick, 1973). It was written
by a young sociologist named James Patrick (a pseudonym, for reasons that will
become obvious), who had infiltrated a gang in the Maryhill district of Glasgow.
By hoodwinking members of the gang into believing that he was like them, then
joining the gang and becoming a participant in its activities, Patrick was able to
paint a detailed picture of the way that it operated. Although it is a case study –
there is only a single gang, not 50, from which to generalise – it gives, through
sparkling analysis, an understanding of gangs. Even though there is no pretence
that this is a representative picture or that all gangs are like this, we can neverthe-
less get a rich understanding of the dynamics, tensions and motivation of gangs
in general.
I’ll let Flyvbjerg (2001: 135–6) have the last word on this since he puts it so nicely:
practical rationality and judgment evolve and operate primarily by vir-

tue of deep-going case experiences. Practical rationality, therefore, is best
understood through cases – experienced or narrated – just as judgment is
best cultivated and communicated via the exposition of cases … which
is why … Richard Rorty [says]: ‘the way to re-enchant the world … is to

stick to the concrete’.
A lot can be achieved by recourse to the concrete. We escape from a tendency too
often found in academic writing to obfuscate with abstractions rather than clarify
with specificity – to bring a fog over the topic in hand with abstract words and the
seeking of generalisation where none is possible and none is helpful.
The great writer Harold Evans (2000: 32) said that the abstract should be ‘chased
out’ in favour of the specific. He was talking of journalism rather than academic
inquiry and writing, but I think his point is just as relevant in the academy, if not
more so. It is a point that is especially germane to the use of the case study, where
the reason-for-being concerns the validity of the concrete and the specific. Evans
(2000: 33) quotes from C.E. Montague: ‘The great escape should be from “mere
intellectualism, with its universals and essences, to concrete particulars, the smell
of human breath, the sound of voices, the stir of living”.’
You should, in a case study, be able to smell human breath and hear the sound
of voices. Nothing is lost in their refraction through our own understanding as
interpreting inquirers. In fact, much is gained as we add a separate viewpoint – one
that moulds and melds the experiences of others through our own understandings.
Is the case study scientific?

If people are ever dismissive about the case study with you – or, indeed, if ever you
have doubts yourself about the validity of this method – then you should remind
yourself of the pedigree of the case study in the natural sciences. It has a lineage of
highly significant advances coming out of case studies – idiographic analysis.
The interesting thing about scientists’ own reflections on their work and the
methodological traditions in which it grew is that it is generally accepted there
is no particular, correct or proper way of generating or marshalling evidence and
undertaking inquiries. As Einstein put it, the creative scientist must be an ‘unscru-
pulous opportunist’. The essence of science,
he said, is the seeking ‘in whatever manner is There are many ways of being
suitable, a simplified and lucid image of the scientific. The case study is one of
them.
world … There is no logical path, but only
intuition’ (cited in Holton, 1995: 168).
Some students in the social sciences seem unaware of this methodological
eclecticism. In fact, they may assume that ‘serious’ scientific inquiry is of a diff
erent character from that of the case study and necessarily involves quantification,
isolation and manipulation of variables and carefully controlled experiments.

This, however, is a product only of a particular kind of mid-twentieth-century
thinking on research and method.
In a classic article reviewing this mindset, two social scientists, Parlett and
Hamilton (1987), suggested that social scientists – particularly in the fields of edu-
cation and psychology – had assumed the methods of social science should, at their
best, properly involve work with large datasets from which generalisations could
be made.
Parlett and Hamilton saw this attitude as wrong-headed. They saw the mistake
as having emerged from an expropriation by the social sciences of the outlook and
methods of certain branches of the biological sciences. In particular, they said, the
methods of agricultural scientists and medics had been seen by large sections of the
social science establishment as the template within which to fit. They described this
outlook, this way of doing social science, as conforming to what they called the agri-
cultural–botany paradigm. Plant scientists do most of their work comparing the growth
of fields of wheat or potatoes or soya beans that have been subjected to different
treatments, but you cannot, they suggested, treat social analysis in the same way that
you treat the analysis of plant growth.
Their point was that the evaluative methods of agricultural scientists are fine
and dandy for studying agriculture, but the trouble is that people are not ears of
wheat, nor are they potatoes or even soya beans. We cannot, when we set up social
science experiments, make assumptions about ‘before’ and ‘after’ conditions in the
same way that agricultural scientists do, because there are no befores, middles and
afters as there are in fields of crops being given extra doses of fertiliser or different
insecticides.
People do odd things, in a way that potatoes do not. People are, in the jar-
gon, agents in their own destiny (in the way that potatoes are not) and in the
habit of making subtle or even drastic changes to the conditions for a trial. The
measures used to assess change cannot be taken with a tape-measure or a pair of
callipers; they have to be undertaken with test scores, attitude ratings and so on.
The trouble is that the latter are assumed to be of the same order of robustness
as the assessments – simple measures of length and weight – that are made of
plant growth. They are not. These social and psychological measures are not the
same as centimetres and grams (which get gold star status in the measuring stakes
because a centimetre is always a centimetre); they are far less trustworthy – in fact,
they can be downright deceitful. In short, Parlett and Hamilton argued, studies
undertaken in education and social sciences in this tradition fall short of their
own claims that they are controlled, exact and unambiguous. (See also my own
musings on this subject: Thomas, 2009, 2012, 2016, 2020.)
The long and the short of this is that there is no one way to be scientific and the
case study is just as scientific as the next way to carry out an inquiry. As Einstein
said (see above), science is not about a method, but about intuition or thinking – it
is about supplying answers to questions with good evidence and good reasoning,
which can be done in a variety of ways, with the principal feature of importance
being the thought and analysis that go into providing those answers.
We sometimes get too hung up, I think, on the shibboleths of science – that
is, the things taken to be core to the enterprise of science, such as causation and
generalisation. Einstein’s point is that it is not these that are at the core of good
scientific thinking. Rather, it is about making connections and having insights
and testing these out, in whatever way. The great sociologist Howard Becker
(1998: 41) put it this way when he wrote about how the connections we seek
are multistranded and multidirectional, such that causality is a less than helpful
concept in social science:
there are many modes of connection, for which we use words like ‘influ-
ence’ or ‘causality’ or ‘dependence’. All these words point to variation.
Something will vary and something else, dependent on what happens
to the first thing, will undergo some change as well. The things that
so vary will often influence each other in complicated ways, so that
‘causality’ is not really an appropriate way to talk about what we want
to emphasize.
Some definitions
The case study is not a method in itself. Rather, it is a focus and the focus is on one
thing, looked at in depth and from many angles. Bob Stake (2005: 443, empahsis in
original) puts it this way:
Case study is not a methodological choice but a choice of what is

to be studied … By whatever methods, we choose to study the case.
We could study it analytically or holistically, entirely by repeated
measures or hermeneutically, organically or culturally, and by mixed
methods – but we concentrate, at least for the time being, on the
case.
So, you have the focus for the case and you choose methods to help you inquire
into the subject. I’ll discuss some of those methods in Chapters 10 and 11.
Stake goes on to emphasise the importance of the singular in the case. So, a doc-
tor may be a case to be studied, but it is difficult to see how methods of ‘doctoring’
could be a case. Nor could the reasons for child neglect be considered a case, says
Stake (2005: 444), since ‘We think of those topics as generalities rather than specifi-
cities. The case is a specific One’ – note the upper-case ‘O’.
Helen Simons (2009: 21) sums this up well in her definition of the case study:
Case study is an in-depth exploration from multiple perspectives of the

complexity and uniqueness of a particular project, policy, institution, pro-
gramme or system in a ‘real life’ context. It is research-based, inclusive of
different methods and is evidence-led. The primary purpose is to generate
in-depth understanding of a specific topic (as in a thesis), programme,
policy, institution or system to generate knowledge and/or inform policy
development, professional practice and civil or community action.
Stake (1995: xi) makes a similar point: ‘Case study is the study of the particularity
and complexity of a single case, coming to understand its activity within important
circumstances.’
Both of these experts stress particularity and complexity and the real-life cir-
cumstances within which the research occurs. Looking at a number of definitions
of the case study, Simons says that what unites them is a commitment to study-
ing the complexity that is involved in real situations and defining the case study
other than by methods.
Still on the question of definition, Hammersley and Gomm (2000: 2) make the
point that all research is, in a sense, case study because ‘there is always some unit,
or set of units, in relation to which data are collected and/or analysed’. They suggest
that the important distinction between the case study and other kinds of research
is the number of cases investigated and the amount of detailed information which
can therefore be collected about each one. Whereas with social surveys, for exam-
ple, you find a little bit – perhaps just one set of questionnaire responses – from
each of the many people participating, with a case study you are finding much,
much more, but about a very limited number.
A summary of Hammersley and Gomm’s table outlining the differences between
the case study and two other forms of research (experiment and survey) is given in
Table 1.1.
The choice of one case (or a small number) is made with a trade-off in mind,
they say. You choose your very restricted sample so that you can gain greater
detail, but this is at the expense of being able to make generalisations about your
findings.
Table 1.1 A comparison of the case study with other forms of inquiry (liberally adapted
from Hammersley and Gomm, 2000)
Case study Experiment Survey

Investigates … one case or a small a relatively large a relatively large number
number of cases number of cases of cases
Data collected and a large number of a small number of a small number of features
analysed about … features of each features of each of each case
case case
Study of … naturally occurring cases where the naturally occurring
cases where the aim is to control the cases selected to
aim is not to control important variables maximise the sample’s
variables representativeness of a
wider population
Quantification of is not a priority is a priority is a priority
data …
To Hammersley and Gomm’s table I would add …
Case study Experiment Survey

Using … many methods and sources one method one method
of data
Aiming to … look at relationships and look at causation look for generalisation
processes
Like Hammersley and Gomm, Ragin (1992: 5) also offers a definition by present-
ing a contrast with what we might call ‘variable-led research’:
The … case-oriented approach places cases, not variables, centre stage.

But what is a case? Comparative social science has a ready-made, con-
ventionalized answer to this question: Boundaries around places and
time periods define cases (for example, Italy after World War II).
Ragin’s definition does two things. It contrasts the emphasis on cases with an
emphasis on variables in other kinds of research, but it also introduces the impor-
tant idea of the boundary.
An important feature of the case study is drawn by Wieviorka (1992), who says
that straightforward examination of the case itself – the child, or the institution or
the period in time, or whatever – is not, at least as far as research inquiry is con-
cerned, enough on its own to constitute a case study. A case study has to have an
additional ‘angle’ to it: the case has to be a case of something – it has to illuminate
and offer a degree of explanation to some wider theme. Wieviorka puts it thus:
For a ‘case’ to exist, we must be able to identify a characteristic unit …

This unit must be observed, but it has no meaning in itself. It is sig-
nificant only if an observer … can refer it to an analytical category or
theory. It does not suffice to observe a social phenomenon, historical
event, or set of behaviors in order to declare them to be ‘cases’. If you
want to talk about a ‘case’, you also need the means of interpreting it or
placing it in a context. (1992: 160)
Wieviorka’s insight is an important one, stressing the ‘analytical category’ – the

theme that the case itself is enabling you to view. I examine it further under ‘The
importance of the analytical frame’, later in this chapter.
***
In summary, definitions of case study stress

Case studies are analyses of
singularity and in-depth inquiry with a view persons, events, decisions, periods,
to a fully contextualised understanding of projects, policies, institutions or
other phenomena which are studied
some situation or theme. The definition that holistically by one or more methods
I shall adopt for case study in this book, tak- to illuminate and explicate some
analytical theme.
ing account of all those above, is as follows:
What is a case?
When I was writing this book, a moment of delightful serendipity occurred when, by
mistake, I pressed the thesaurus button on my word processor while the cursor was
stationed over ‘case’. I give a summary of the meanings it came up with in Table 1.2.
Table 1.2 Meanings given for ‘case’
Container Situation Argument

box instance reason
casing event defence
crate occurrence justification
suitcase state of affairs rationale
folder circumstances basis
I had thought that I knew what ‘case’ meant, but this serendipitous opening
led me to some personal research into the meanings of ‘case’. It turns out (if
my less-than-scrupulously-thorough research is correct) that the rather different
meanings of ‘case’ arise because the word is taken from two similar-sounding
Latin roots that have arrived, after corruption, at the same word, ‘case’, in English.
The two roots are:
• capsa, meaning ‘box’, ‘container’ and, thus, ‘case’

• casus, meaning ‘event’, ‘fall’, ‘accident’ and, thus, ‘case’.
Kelly (1993: 35) notes that the Oxford Latin Dictionary lists 11 different meanings for
the word ‘case’, suggesting that ‘the term was just as confusing in Latin as in English’.
Anyway, this personal etymological quest led me to realise that the use of ‘case’ in
‘case study’ ticked two important boxes – by good luck, probably, rather than good
planning – about the essence of this kind of inquiry. Quite fortuitously, therefore, it
is an entirely appropriate term for what is done in a case inquiry. Let’s look at these
two meanings, plus the other one identified by my computer’s thesaurus.
The case as container

The first column in Table 1.2, headed ‘Container’, is about the physical, concrete
meaning of ‘case’. While I hadn’t at first considered that this was relevant to the
case study, interestingly it emphasises the containment of a case study. A case, in the
example of a crate or suitcase or even a pencil case, is ‘bounded’: you close it and
clip the latches and that’s it. The case – think ‘suitcase’ – is everything that is in it:
T-shirts, jeans, socks, underwear, washbag, flip-flops, sunglasses, and so on.
At the risk of going from the ridiculous to the sublime, this is, in fact, similar to
something the great philosopher Ludwig Wittgenstein said in the opening line of
his famous Tractatus: ‘The world is everything that is the case’ (see Kenny, 1993: 3).
Wittgenstein’s language is sometimes thought to be a little mysterious, enigmatic
or even poetic – as if he were teasing his merely mortal readers with riddles, saying,
‘I’m not giving you answers, peasants: go away and think about it.’
Wittgenstein’s enigmatic statement is relevant to the case study. Wittgenstein
scholars (see Biletzki, 2009) have understood his ‘The world is everything that is the
case’ to mean that the world in which we are interested, as inquirers, is composed
of facts and states of affairs and objects, all of which are in a constant interrelation-
ship with one another. Like the zillions of atoms in a box, they bounce into one
another, they fit or don’t fit with each other and intermesh in myriad ways, and
the result is inherently highly complex and could have taken any one of a number
of other forms in our box. This is, in a sense, what the case study is about: the focus
is on the complexity. That is, we have a state of affairs bounded by the case (think
‘suitcase’ again) and we study the complexity of what is in there.
The case as situation, event

Here, ‘case’ means a particular instance, an event, a happening, and the set of cir-
cumstances that surround this. So, this is different from the bounded (suit)case
since it is less about the parameters that define the extent of the subject and more
about a set of conditions or a state of affairs.
This, too, very nicely defines what happens when a case study is undertaken.
A case study is about a set of circumstances in its completeness and the case is
described – marked out – by those circumstances. It is the circumstances of the
instance that are being studied. Where did it happen? When? What had happened
before? Who was around? What was in the news? How did all of this affect what
was going on and how events turned out?
An important feature of the case here is that, as Mitchell (2006: 31) puts it, there is, in
ordinary use of English, ‘a strong connotation that the word “case” implies a chance or
haphazard occurrence’ and this haphazardness entails specificity or particularity. This
is an important point to make, for, as I discuss in Chapter 4 regarding ‘samples’, there
can be no assumption that the case is in any way representative of a wider whole – it
is a one-off, defined by the peculiar circumstances that you, the researcher, describe.
The case as argument

The clue to the meaning of ‘case’ here is in statements such as ‘The case for … is as
follows’ or ‘Here is the case for the prosecution’. ‘Case’ is thus now being taken to
mean a rationale, an argument.
This is not, if I am honest, likely to be relevant to the origins of the term ‘case study’;
nor is it likely to have been influential in the term’s take-up as describing a form of
inquiry. It is interestingly relevant, however, since a case study involves the rationale
of one thing relating to another or possibly causing another. It involves your train of
reasoning about the interrelationships between the elements of your study. It involves
your justifying your reasoning and the conclusions with which you emerge, using evi-
dence drawn from your empirical work. It is essentially about the arguments that you
make to connect the elements of your observations. The arguments that you pose are
like the fibres and ropes that hold all of the disparate elements together.
The subject and object of a case study

I noted earlier Stake’s distinction between a study of a doctor, who could form the
subject of a case study, and doctoring, which could not. On his or her own, though,
even the doctor would not be a case – at least, not a very interesting or informative
one. To be a case – an interesting one – we would have to be able to say that the
doctor was a case of something (you should have an idea of what he or she is a case
of at the outset). It will then be this analytical focus that crystallises, thickens or
develops as the study proceeds. Indeed, it is the way this emerges, grows and devel-
ops that is at the heart of the study.
So, the Korean War would not be a case unless it was a case of something. Is it
a case of a war? If so, can you say that it is a case of an especially remarkable or
unusual kind of war? Perhaps it may be, by contrast, essentially a case of a border
dispute, but one that has grown out of proportion to its original significance. Is
it a case of US resistance to what was assumed to be communist expansion? Just
chronicling the story of the Korean War would not be a case study. An exploration
of any one of these ‘ofs’ – any of these analytical backdrops – would, however, make
a fascinating case study.
How the subject and object connect

Wieviorka (1992: 159) explains the distinctions I have drawn in my Korean War
example by noting that when we talk about a case we, in fact, are talking about two
elements:
1. what he calls a ‘practical, historical unity’ (the Korean War in my example) – this is,
in essence, the subject
2. what he calls the ‘theoretical, scientific basis’ of the case (such as US resistance
to communist expansion, in my example) – this is the analytical frame or what we
could call the object.
So, just extending the war theme for a moment, we could not have a case study
entitled ‘World War II: a case study’. Here, World War II is not a case of anything.
However, ‘World War II: a case study of a “just
war”’ would contain both of the elements Case study research comprises two
parts:
outlined by Wieviorka. It would make an inter-
esting piece of work since it could examine in 1. a ‘subject’ – the case itself
detail the notion of a just war, played out and 2. an analytical frame, or ‘object’.
exemplified through the case of World War II.
Alternatively, take the example of a hospital ward. Would a description of this
ward constitute a case study? In my opinion, it would not – it is not a case of any-
thing. It becomes a case of something when you can explain the analytical frame
through which you might be viewing it. It might:
• be a good example of that analytical frame

• demonstrate something interesting in terms of your analysis because of its
peculiarity
• be an example of an analytical focus that arises by virtue of your personal experience.
As I indicate in Table 1.3, the analytical focus must extend beyond mere description.
Table 1.3 What is and is not a case?
Subject (the ‘practical What would not (on its own)

unity’) constitute a case study? What would be a case study?
Jesson Ward at Parktown A simple description of the ward An analysis of why it is thought to be
Children’s Hospital an outstanding children’s ward
Editorials in the Daily A content analysis of each A case study analysis of a
Globe for the six days in editorial newspaper proprietor’s influence
week beginning 5 March over the content of editorials
Aleksandr the meerkat: The rise of this advertising A case study analysis of the
a successful advertising phenomenon exemplary use of personality and a
campaign storyline in advertising
Amelia: a quiet child A simple description of Amelia A case study analysis of Amelia over
and her circumstances a set period in class, with a view
to understanding, illuminating and
gaining insight into the phenomenon
of ‘quiet children’
So, a case study is like one of those capsules with two halves: each half, each
ingredient, is necessary in order for the other half to work, as shown in Figure 1.1.
It has to contain the ‘practical, historical unity’, as Wieviorka (1992: 159) puts
it (or ‘the subject’ in plain English), and it has to contain the analytical frame
(or ‘object’). It is not complete without both parts in place: one will not work
without the other.
The subject:
the person, The analytical
place ... frame or object
Figure 1.1 The two parts of a case study, each part needing the other
Let’s take another example and put it in the two-part capsule. Let’s take as our
subject the 1923 hung parliament in the UK. A narrative of this on its own is not a
case study. How can we put it in the context of an analytical frame?
We first of all have to remember that we are looking at this one hung
parliament – the 1923 one. We are looking at it as one of a set of hung parlia-
ments. Why are we looking at this particular one? Perhaps it is an especially
interesting one or an unusual one concerning the negotiations and the ‘horse-
trading’ that occur among party leaders at the beginning of a hung parliament,
just after the election result has been finalised and announced. This, then,
would be our analytical frame. Clearly, a range of analytical frames could poten-
tially accompany the subject, but we would need to choose one (or more) of
these (see Figure 1.2).
Analytical frame:
Subject: negotiation
the 1923 hung process at the
parliament beginning of the
parliament
Figure 1.2 Putting the two parts together
In the way that Wieviorka sees the case study, as subject and object, the subject is
almost less important than the object. What I mean by this is that the subject, the
case itself, is the lens through which you are looking at the object. Let’s unpack one
of the examples from Table 1.3 to illustrate what I mean …
A case study of a quiet child: Bringing subject

and object together
Imagine that you are interested in the phenomenon of young children
who are withdrawn – sometimes called ‘quiet’ or ‘silent’ children. They
rarely seek adult attention and will avoid playing with or approaching
others. They prefer solitary play or may just tag along on the edge of a
(Continued)
group. It is hard to get them to talk or to relate with either adults or other
children.
If ‘quiet children’ were your topic of interest, they would be your ‘object’.
Given this object – this topic of interest that you wanted to research – it
would be useful to use as a case study subject a child who showed the kinds
of behaviours I’ve just described (and of course it may be this particular child
who has led you to this topic of interest).
How would a case study proceed with this child – let’s call her Amelia. With
all ethical considerations taken into account and procedures completed
(see Chapter 5), you would get a rich picture of Amelia, her life, her relation-
ships with adults and children, how she spends her day, what she does and
does not like doing. You would observe her in class and at home, you would
interview her parents and her teachers, you might ask the classteacher to get
the children in Amelia’s class to indicate their favoured partners for a socio-
gram (see page 236). You might volunteer to be a playground helper and/
or a classroom assistant for a day a week for a term, to write an ethnography
(see page 159) of the ‘class world’.
And then you would analyse all the information, the data, that you
had collected about Amelia – your case study subject – but remem-
bering always that you would be analysing those data with the ‘object’
in mind. In other words, you would be analysing your data in trying to
The object is the thing to

Enquirer be explicated and analysed
(you) (namely, the concept of the
quiet child)
The subject
(Amelia, the
‘quiet child’,
in this
example) is
the lens
through which
we view the
object
Figure 1.3 Lens analogy

understand, illuminate and gain insight into the phenomenon of ‘quiet

children’. How does Amelia communicate in different environments?
Does she become very much more communicative, even noisy, at home?
Are her home and her relationships at home so comfortable that she
has never developed the resilience needed to cope with the rough-and-
tumble of school? Or, by contrast, might she be much the same at home,
pointing perhaps to a withdrawal from what she has learned are always
going to be stressful or negative interactions with other people?
Using the lens analogy, we can see the case study as in Figure 1.3.
A rich picture – with boundaries

A case study offers you a rich picture with many kinds of insights coming from different
angles, from different kinds of information. So, you may go into your case study and do
interviews, make observations, keep a diary, look at statistics. Nothing is ruled out. The
case study is a frame that offers a boundary to your research. Equally, you could think
of the case study as representing the end of a searchlight beam. Everything at the end
of the beam is seen in bright light and thrown into sharp relief, shadows and all. You
study what is in the beam of light and your subject of interest in it is, ‘What happened
here? How did it happen? What was connected to what? Why did it happen?’
The end of a beam of light has a boundary, an edge, just as a case study has a
boundary. Your case study is defined not so much by the methods that you are
using to do the study, but by the edges you put around your case – the direction in
which you want your research to go and how far. Perhaps you will say, ‘I want to
look at the development of highly successful companies’ and choose to do this by
looking through the lens of the home furnishing company IKEA. Here, then, the
object is ‘the growth of highly successful companies’ and the subject is IKEA.
Let’s look at this particular example, defined by these parameters.
A case study of IKEA: What are we looking

for in a case study?
In this case study your principal theme of interest will be growth in highly
successful companies: this is the object of your study.
(Continued)
With this focus on growth, this case study will be focusing on change
over time, beginning perhaps with a reconnoitre of the available statistics, to
establish the grounds for your case. You will want to establish that IKEA really
is a highly successful retailer and to show how it has grown over time, exam-
ining its turnover and profit figures since its creation by Ingvar Kamprad in
Sweden in 1943. What does the graph of turnover figures look like over the
years? Alternatively, you could show, on a series of maps, each representing
a 10-year period, the countries in which IKEA stores have existed. You may
be able to find out the figures you need quite easily or you may need to go
to some quite obscure sources.
This gathering and plotting of basic data on turnover and profit over the
years since 1943 provides the boundary for your case study. It says, ‘This has
happened: IKEA really is the biggest and it has become the biggest in a
period of less than 50 years.’
This, though, is not enough for a case study. Simply to show the growth
over time is not sufficient. You have the boundaries defined by the growth, but
now you want to know how and why the growth has happened and, while you
will not be able to provide a definitive answer, the case study should provide
you with enough evidence to make a convincing argument. You will need to
click together bits of information like pieces in a jigsaw, bits of evidence that
you collect in support of any one of a number of tentative ideas – sometimes
called hypotheses or theories – about the growth of the company.
If you are thinking about the growth of the company, what might be plau-
sible ideas to explain its growth? Let’s have a look at a few.
The company may have grown because its founder had one big idea or a
series of good ideas, such as realising the:
• potential of sawdust and chipboard for making cheap furniture

• willingness of consumers to build their own furniture from a kit if the
price and design are right
• ability and willingness of consumers to carry away their own flat-packed
furniture with the advent of mass car ownership.
If any of these were important, where did Kamprad’s ideas come from? Did
he work in a sawmill and notice the abundance of sawdust? Did he work in
retailing and realise the wastage involved in ‘transporting air’ in assembled
furniture? Was it, instead, down to the retailing technique employed in IKEA
stores, wherein consumers are encouraged to walk along a certain route
through the store, being encouraged to buy as they make their extended
and unexpected walk?
How could you go about finding out which of these potential ideas was
important? Clearly the ideal would be an interview with Ingvar Kamprad,
to ask him about your tentative theories, but I doubt if he gave many inter-
views. You will almost certainly have to rely on printed biographies of him.
The Wikipedia entry for him, for example, tells us that in 1976 he wrote a
manifesto detailing the IKEA concept of frugality and enthusiasm. We learn
that he also worked with Swedish journalist Bertil Torekull on the book
Leading by Design: The IKEA Story (Kamprad, 2000). In the autobiographi-
cal account, he further describes his philosophies and the trials and triumphs
of the founding of IKEA. Also, that he began to make a lot of money when
he was a child by buying matches cheaply in bulk and then reselling them at
a profit. He expanded to fish, ballpoint pens and pencils.
Already we can see that he seems to be one of those people who almost
cannot help but make money. It is not as if he had an especially good idea
about furniture or an especially wonderful insight into how it could be mass-
produced. Rather, the seed seems to have come from his dynamism in
finding ways to sell. Alongside this, he apparently drove a 15-year-old car,
flew only economy class and encouraged IKEA employees always to write
on both sides of a piece of paper. So, an obsession with economy, paring
expenses down to the very last little bit, also comes through.
This is the kind of emerging information on which the narrative in a case
study is built. The analysis begins to ‘thicken’ in this way around one theme.
Remember, though, that the topic of theoretical interest that took you to
this case study subject – the analytical frame – is ‘growth in highly successful
companies’. The case study you are conducting is intended to throw light on
this topic and your discussion will be orientated to its illumination.
Forms of case study … and six steps to

follow in devising your own
There is not just one kind of case study design. There are many, depending on
inquirers’ starting points, their purpose in doing a case study, and the approach
that they are likely to take. The case study takes shape contingent upon these issues:
starting points, and choices that inquirers take about purposes and approach. What,
then, are the paths that inquirers can follow?
To try to help answer this question, I have summarised elsewhere (Thomas, 2011a)
some of the categorisations that methodologists have made about case study. Out
of these commentators’ analyses I have distilled a typology (see Table 1.4), which
comprises some elements I have touched upon already and some which I have not
yet discussed in detail, but will explore further in Chapters 6 to 9. The important
feature of this table is that subject, object, purpose, approach, methods and process
are separated, for the purposes of understanding how case studies may be structured.
Table 1.4 Kinds of case studies – a typology
Methods
Object
Subject Purpose Approach Process

• Local • Intrinsic • Testing a Retrospective
knowledge theory
in data collection and

• Instrumental Single or Snapshot
methods to be used
case •
Choices about the
Choices about the

• Evaluative Building a multiple
• Key case theory Diachronic
analytical focus
• Explanatory
• Outlier case • Illustrative Nested
• Exploratory
analysis
Parallel
Sequential
Throughout this book I shall be drawing on this system for thinking about the
classification of case studies. I hope this will be useful in helping you to think about
the different forms of case study, what their purposes are, and how they might be
constructed. As I say, I go through this in more detail in Chapters 6 to 9 but I’ll give
you a brief introduction to it now so that you can think about how the various case
study examples I offer as the book progresses differ in aim and structure.
The classification system – or ‘typology’ – I have drawn up splits thinking about
case studies into six elements, which translate into six steps in the construction of
a case study:
1. The subject of the study – what is the case you are actually looking at? Is it a class-
room, a child, an institution, a war, a country, a political event, or what? Where
does this subject come from? It may be one that you know a lot about from your
own experience – a local knowledge case. Or it may be a key case in the literature.
Or it may be one that is interesting because it is an outlier – different from the
norm. Whichever, this is the lens through which you are examining the topic of
theoretical interest that you are focusing on, namely the object of the study.
2. The object of the study – what is the theoretical topic at the heart of your question
that the subject of your study is allowing you to explore in detail?
3. The purpose of the study – why are you doing it? Is it to make an evaluation? Is it
out of curiosity, or is it with a particular understanding in mind?
4. The approach of the study – is it to test a theory that you or someone else has
developed about a topic? Or is it to develop a theory from new about that topic?
5. The methods you will use to do the case study. What data collection tools might
you use, and how might you analyse the data that you gather?
6. The process you will use to conduct the case study. Will it be a single or multiple
study, and will you look at the case during one particular moment in time, or over
a longer period?
The choices you make about your subject, object, purpose, approach, methodology
and process will interact in various ways to blend into one unique study out of an
array of possible permutations.
Let’s now use the typology to offer an organisational scaffold for the case study
of Amelia that I gave earlier in this chapter. What is happening here? After noticing
something worth investigating, namely the strange quietness of certain students at
school, the teacher read around the area and wanted to develop insights into the
psychology of these quiet children. Using the typology, the six steps, as a mental
scaffold, she …
1. Identified as a case study subject a student whom she knew well – Amelia. Amelia
exemplified the topic in which she was interested, namely quiet children. This would
be a case study subject that emerged out of her local knowledge and experience.
2. Developed and explicated her interest in quiet children and embarked on a full
literature review of the topic – offering detail on the analytical frame, or object, of
her study.
3. Made clear her purpose in undertaking the study – this would be an instrumental
study with the aim of gaining insights and understanding of Amelia’s quietness
with a view to illuminating the topic more generally.
4. Decided on the approach she would take in undertaking the study. Would she be
testing theories from the literature about quiet children? There was not, she discov-
ered, very much in the way of theory to test. She concluded that her approach was
theory building: trying to formulate ideas about Amelia’s context and behaviour
that offered insights and understandings – theory – about her quietness, and pos-
sibly that of other children.
5. Decided on means and methods of data collection and analysis – her methodology.
This would comprise observation, interviews, a sociogram and analysis involving
theme mapping and thick description.
6. Decided that this would be a single case study spanning a particular period in
time – a snapshot of Amelia’s behaviour, context, interests, relationships, over half
a term.
Mapping these to Table 1.4, we can show the path for Amelia’s case study (Figure 1.4).
Methods
Object
• Local • Intrinsic • Testing a  Retrospective

knowledge theory  Snapshot
in data collection and

• Instrumental Single or
methods to be used
case 
Choices about the
Choices about the

• Evaluative • Building a multiple  Diachronic
analytical focus
• Key case theory 

• Explanatory  Nested
• Outlier case • Illustrative 
• Exploratory  Parallel
analysis
 Sequential
Figure 1.4 The path of Amelia’s case study
If you take only one thing from this chapter,

take this …
• A case study is about seeing something in its completeness, looking at it from
many angles. This is good science.
• Definitions of case study stress singularity and in-depth inquiry. The definition that
I shall adopt for case study is as follows:
Case studies are analyses of persons, events, decisions, periods, projects,

policies, institutions or other phenomena which are studied holistically by
one or more methods to illuminate and explicate some analytical theme.
• Although we cannot generalise from a case study, generalisation is not always

what is wanted from the inquiry process. We do not always want or need to gener-
alise, and some of the most inspired and insightful research, of any kind, has come
about as the result of case studies.
• What the case study is especially good for is getting a rich picture and gaining ana-
lytical insights from it. The ‘analytical’ bit is important: each study has a subject of
interest (a person, place, event or phenomenon) – the case itself – and an analytical
frame, the object, within which it is studied and which you are hoping to elucidate
in some way.
• Case study may take different forms, depending on how and why you select your
case and what you intend to achieve. You may choose to conduct a single case
study, or one involving several subjects, and you may choose to look at the case
(or cases) over a short or a more extended period.
• Following the six steps from the typology will help you to scaffold your thinking
about how to construct a study depending on your own circumstances, questions
and expectations.
REFLECTIVE ACTIVITY
At the end of this chapter I used the example of Amelia to exemplify mapping
that case study to the typology shown in Table 1.4. In other words, I broke down
the Amelia case study into subject, object, purpose, etc.
Try analysing another case study – perhaps one of the other examples from
Table 1.3, or the boxed example of IKEA – according to the typology. Or use a case
study in your own area of interest – one that you know about, or one that you
might conduct yourself. Draw arrows, as I have, between the different elements.
Explain why the case study is taking this path through the typology. Tip: There’s
no single right answer – you can justify the choices that you make.
Further Reading
Cohen, M.F. (2007) An Introduction to Logic and Scientific Method. New York:
Harcourt, Brace.
Originally published (with Ernest Nagel) in 1934, this is a magnificent, if slightly
dated, outline of what constitutes science. It is good on the duty of doubt in sci-
ence. See especially Chapters 2 and 12.
Gerring, J. (2016) Case Study Research: Principles and Practices. Cambridge:

Cambridge University Press.
An excellent primer.
Gomm, R., Hammersley, M. and Foster, P. (eds) (2000) Case Study Method. London:
Sage.
This book offers, as the editors put it, the most influential and important articles
on the case study. See especially the introduction by Hammersley and Gomm for
an excellent overview.
Hammersley, M. (1992) What’s Wrong with Ethnography? London: Routledge.

See Chapter 11, ‘So, what are case studies?’ By comparing the case study with sur-
veys and experiments, Hammersley aims to show that these methods are not as
different from each other as we sometimes assume and the differences that do exist
between these ways of structuring research hinge on case selection.
Patrick, J. (1973) A Glasgow Gang Observed. London: Methuen.

This is a book I never tire of recommending. Like Becker et al.’s Boys in White
(1961/1980), it is as good as a novel, describing the author’s infiltration of a gang
of young men in Glasgow.
Ragin, C.C. and Becker, H.S. (eds) (1992) What Is a Case? Cambridge: Cambridge
University Press.
Many excellent articles are included here. See especially Ragin’s introduction.
Simons, H. (2009) Case Study Research in Practice. London: Sage.

An excellent, practical book on conducting a sociologically framed case study.
Taken from the point of view of a student making her way through the process.
Stake, R.E. (1995) The Art of Case Study Research. Thousand Oaks, CA: Sage.
Stake, R.E. (2005) ‘Qualitative case studies’, in N.K. Denzin and Y.S. Lincoln (eds),
The Sage Handbook of Qualitative Research (3rd edn). Thousand Oaks, CA: Sage.
Stake is one of the doyens of the case study method and his work is well worth reading.
Thomas, G. (ed.) (2014) Case Study Methods in Education. Volume 1: Methodological

Issues around the Use of Case Studies in Social Science. London: Sage.
In this volume I have collected some iconic references for case study, including
some of the key ones drawn upon in this chapter.
Thomas, G. and Myers, K. (2015) The Anatomy of the Case Study. London: Sage.
Here, Kevin Myers and I outline in more detail the ways in which case studies work,
and we give several examples of case study in action.
Tight, M. (2017) Understanding Case Study Research: Small-scale Research with Meaning.
London: Sage.
A well-organised, practical primer to doing case study research.
2
THE CASE STUDY AND
RESEARCH DESIGN
First things first: your purpose
Research begins with a purpose and a question, not a research design. You have a
reason for needing to find something out (your purpose) and this leads to a ques-
tion. It is this question that will be at the heart of your research. The design follows
on from that, rather than the other way round.
Investing time and thought in your research question will yield a design solution.
But because of all the initial uncertainty about where the research question might
lead, sometimes people will be tempted to start by thinking of a design route before
they have given their research questions more thought. This means that they will
approach their research problem from entirely the wrong direction – looking for a
method first or going in with a favoured technique – while giving relatively little
attention to the research issue.
You must do the opposite: you must put the question before the method and let
the question determine the method you use to answer it. You cannot say, ‘Here’s a
great method. I’m sure I can use this in my research.’ So, you should not, if you are
sensible, begin by saying, ‘I want to do a case study.’
This mistake of putting the cart before the horse in research design reminds me
of the famous story of a man who lost his keys. He was desperately searching the
gutter under a street lamp. A friend came along and, seeing him hunting around,
asked, ‘Have you lost something?’
‘I’ve lost my car keys. I’ve been looking for them for half an hour’, the man
answered.
The friend joined in to help him, but after some time she couldn’t see any sign of
the keys. ‘Are you positive that it was here that you lost them?’ she asked.
‘No,’ he replied. ‘I lost them over there in that dark alleyway.’
‘Well, why on earth don’t you look over there, then?’ the friend asked
incredulously.
‘Because this is where the light is, of course!’ the man answered.
The moral of this story is that we should not concentrate our research and all of
our analytical effort on where we have good research instruments, good light, as
they will tell us nothing on their own. The main issue must be ‘What do I want to
find out?’ (‘What am I trying to find? Oh yes, my keys!’) The issue cannot be ‘What
is the best instrument for finding things out?’ until we have decided on that first
question.
Next, your question

As I just noted, it is easy to get hypnotised by the idea of methods in research –
almost as if these methods were at the core of the research. They are not. Methods
are a way of answering a question, but the key – at the very start – is to think about
the question you want to answer and consider how it can be answered.
If someone asks you a question in real life, such as, ‘Do you know the way to
San Jose?’, your immediate response will surely be to think briefly and then try to
answer to the best of your ability. If you are in Fremont, California, a little way up
the road from San Jose (the legendary one), you will say, ‘Take Route 880 south for
10 miles, and you’re there, my friend!’
If, however, you are asked the same question in Fremantle, Australia, 3000
miles from San Jose, your response will depend on your mood and your assess-
ment of the intent of the questioner. Is the questioner teasing you? Is she drunk?
Is she genuinely confused? If she is giggling with her boyfriend and you think
she is pulling your leg, you might laugh – grudgingly. If in a good mood, you
might clear your throat, smile and, in your best Dionne Warwick, trill out, ‘I’ve
been away so long I may go wrong and lose my way’. If you assess that she is
drunk, you might just ignore her. If you think she is confused, you might call
for help.
Whatever the question, you will assess it on the basis of your immediate context
and respond to it in a way that accords with that context and to the best of your
ability with the tools available to you at the time. You are not a computer, pro-
grammed with stimulus–response contingencies, so your immediate reaction to the
San Jose question will not be …
Ah! This is a question about directions. What do I know about methods to do
with solving directional questions? Maps! Now, do I need a map with a Mercator
CASE STUDY AND RESEARCH DESIGN 29
projection, a transverse Mercator projection, a Collignon projection, a Mollweide

projection or, let me see, I really ought to check out the benefits of the Lambert
conformal conic projection.
No, you do not do any of this. You think straight away about how to answer based
on your intelligence and common sense – and this must be your first response to
your research question. One thing it is important to remember is that, because you
are studying something in its completeness in a case study (that is, you are not just
studying one or two relevant variables), it is all too easy to try to do too much in
both the setting and the answering of the question. If you do this, though, you just
end up with a mushy heap.
The key to avoiding mushy heap syndrome is to find a storyline through the
general topic in which you are interested – one that is relevant to your particular
interests. In Chapter 3 I shall go on to note the importance of seeing the whole
situation as an ecology, and one of the pioneers of the ecological psychology move-
ment, Roger Barker (1968: 12), makes the point well. He highlights the legitimacy
of alternative forms of inquiry and analysis by invoking an example of alternative,
but equally valid, explanations of the same event. He asks us to imagine the move-
ment of a train of wheat across the Kansas plains. How is this movement to be
explained? An economist will explain it in one way, while an engineer will explain
it in another. ‘Both the laws of economics and the laws of engineering are true; both
operate in predictable ways on the train.’
In Chapter 6 I shall go on to examine the kinds of case studies that can be under-
taken once you have addressed your question in detail. I suggest that you use the
categorisation in Chapter 6 as a kind of check once you have made decisions about
how you will answer the question. First, though, it is important to think about the
channels open to you for answering your question.
In sum, you should not start a project assuming that you are going to do a case
study. Rather, you should go into it with a question that the research study will
answer. The case study may or may not be the design form that you choose once
you are clear about your question.
Kinds of questions
So, it is the research issue – your question – that is the starting point for your design
journey. It is the pivot for your research: everything rests on it. Your first port of
call in doing your research is not to say boldly, ‘I’m going to do a case study!’, but,
rather, ‘What do I want to know?’ or ‘What do I want to find out?’ It is the research
question that leads you in the direction you need to go.
Your question does not need to be fixed in concrete right at the beginning. It
proceeds from a prima facie question to one that is a more refined, final question.
When I talk about your ‘prima facie’ question I mean your first not-very-well-
thought-out question. This will nearly always change in the course of your inquiry.
It is quite difficult to get your question right, but you have to start somewhere
when you are doing your research, which is where your prima facie question comes
in. It lets you make a start without unnecessary dithering about whether or not you
have got your question right. It allows you to
A question is the starting point for get going. Go boldly with it; be unafraid, on
your research. Begin with a question, the understanding that it will change. It will
not a presupposition that you are
going to do a case study. A case get better as you think more about the subject
study should follow logically from and do more reading.
your question, or else you should
In Figure 2.1, let’s see where your prima facie
not do one.
and final questions lie in the design trail.
After your literature

review you are able
to review and refine
The questions your prima facie
you pose in your questions to emerge
introduction are with final questions
actually your
prima facie
questions
Literature
Introduction
review
Prima facie questions Refined, final questions
Figure 2.1 Prima facie and final questions
Literature review
I think the best policy is to state your prima facie questions at the outset, in the
introduction to your dissertation or thesis, then refine these in the light of your
literature review and state them again, rethought out and reformulated, at the end
of the literature review.
In any research, the literature review is a crucial element, enabling you to see
what has been done in the area of your interest, and allowing you to focus your
ideas based on others’ work. You can look at the agreements, disagreements,
dilemmas and questions in an area and adapt or redirect your own focus on the
basis of what you have found. You can also look at the methods used by other
researchers and the trials and tribulations they encountered in their work to see
if you can garner any lessons about method and approach for your own research.
However, in a case study, the literature review assumes a particularly important
role. The ideas you encounter and consider while reading the literature will help you
to formulate your own ideas about the object of the study, which I discussed in the
previous chapter. If you remember, I talked about the subject and the object of a case
study, with the object being the theme, the topic of analytical interest that you hope
to illuminate and elucidate by looking through the lens of the subject, the case itself.
So, here, in the literature review you are exploring the topic of analytical interest, the
object, to see what others have thought – to develop a theoretical understanding of
the area which you can test, refine or otherwise cultivate in your own research work,
that is your own practical work of collecting evidence and analysing it.
What, then, should you be concentrating on when you are doing a literature
review? How should you be cultivating this topic of analytical interest? The key, I
think, is to search for what the great psychologist Jerome Bruner (1997: 142) calls
‘the trouble’ in your area of interest. What, in other words, does not make sense?
What itches? What arouses your interest, and why? Elsewhere, Bruner (1991) talks
about what he calls hermeneutic composability: in plain English this means how
does it all fit together? What appears to depend on what? What contradicts? Where
are there paradoxes? (I explore Bruner’s ideas in more detail in Chapter 11.)
As you do this, you should be developing a storyline with lots of connections
between one part and another. It should have a beginning, a middle and an end:
you outline the issues at the beginning; you provide the analysis and synthesis in
the middle, and you conclude by summarising the issues, differences, paradoxes,
dilemmas and questions yet to be resolved.
The aim is to find themes – or, by contrast, discontinuities, breaks, disagreements –
that run through the literature. When you have done a reasonable amount of searching
you will be able to see these emerging and it is useful at this stage to draw a storyboard –
a plan that sums up and brings together the ideas that are emerging from your litera-
ture review. So, let’s look at a storyboard …
Using the literature and a storyboard to

help you design your case study
In conjunction with your review of the literature, a storyboard is invaluable. It
enables you to brainstorm initial ideas and trace a path through the good, the bad
and the ugly in these initial musings. With it, you can sift out the not-so-appealing
ideas and proceed with those that promise most fruit. You can add to your story-
board as you get further into your literature review and gain information in your
initial inquiries.
In the following example, I try to give a flavour of the kind of thinking that needs
to go on as part of developing your research ideas out of the literature – the general
discourse in your area of interest – and how the storyboard can help in this.
Designing a case study: An example with

brainstorms and storyboards
Let’s assume that you are a political science student and interested in the
cultural conditions that existed in Eastern Europe during the 1980s. Your
purpose, then, is to study this.
This may lead you to a prima facie question such as ‘What led to the fall of
the Berlin Wall?’ This is fine as a prima facie question, but it will need some
work to make it interesting as a research question for, at the moment, it is
too broad. It could be answered a thousand different ways. Imagine it. I offer
some starting points in the brainstorm illustrated in Figure 2.2.
Kennedy’s ‘Ich bin ein Powerful

Berliner … We have symbol Border
never had to put a emplacements
Influence wall up to keep our
of USA people in’
Division
and hatred
The Berlin Wall

Levi’s jeans
Fall of Eastern
Europe
Cultural change Gorbachev,

glasnost and
Breaking down
perestroika
with hammers
and pickaxes
Figure 2.2 Brainstorming the fall of the Berlin Wall

Anything can be considered in brainstorming, and this particular session

might lead you to think in more detail about a number of possible reasons
that could be explored more fully:
• the interest of the USA and the West in the survival of West Berlin over
decades
• the bankruptcy of the Soviet Union after four decades of the arms race
• the conspicuous differences in standards of living in West and East
Germany
• the assumption of power by Gorbachev and his liberalisation of politics
• the damage to the economy caused by the black market in Levi Strauss
jeans
• the Soviet political hierarchy’s refusal to support Erich Honecker, the
East German leader at the time
• and so on.
For each of these possible starting points for an answer – each of these ana-
lytical frames – there are many potential ways to proceed to examine things
further. You could look at statistics and official records, examine newspaper
reports, interview those who lived through it, and so on.
For a piece of small-scale research of the kind done as part of an under-
graduate or postgraduate programme of work, these potential avenues will
be narrowed by the options that are open to you. You may have limited time,
money and access to information. More importantly, though, as far as this
book is concerned, the questions will fan out into a broad range of potential
design frames – and the one that we are interested in here is the case study.
So, what marks out the avenue of inquiry that leads to a case study, as
distinct from some other design frame? It will be a question that:
• leads to a study of one thing, one phenomenon, one situation

• focuses on a situation that is marked out as special or interesting in some way
• enables you to look at the relationship of a variety of different factors in
the situation.
Thinking about it, reading about it (for your literature review) and talking with
tutors and fellow students, you may realise that, from the various avenues of
(Continued)
inquiry open to you, what you are really interested in is the role of popular
culture in Eastern Europe at that time and how far this contributed to the fall
of the Wall. This still does not tell you whether or not what you need to do is
a case study, though. What will tell you this?
Remember that the defining characteristic of a case study is the
uniqueness of your subject. So, if you are choosing to do a case study
of the Berlin Wall, you are in some way identifying the Wall as special
(see Figure 2.3). Clearly, we have to be sensible here. It is not unique in
the sense that it is different from your back garden wall or the wall of
your house. (Well, it is, I guess, but not in any interesting way – unless
you are a builder or a bricklayer, in which event you might take a special
interest in its straightness or the quality of the mortar.) It is interesting
perhaps as:
• a definition of territory by an invading power

• a symbol of authority
• a partition
• a national border
• an object on which graffiti are drawn
• an object of state violence.
With any of these phenomena you could take the Berlin Wall as a case. You
would then be looking at the Wall as an especially interesting example of
one of these phenomena – one of these analytical frames.
Taking a slightly different slant, you could choose to look at the fall of the
Berlin Wall as an especially interesting example of a rather different set of
phenomena, such as:
• the collapse of authority

• the victory of the people over autocracy
• the power of bulldozers vis-à-vis the weakness of humans
• the unassailable integrity of the German state
• the success of the Western propaganda machine in the Cold War
• the power of popular culture and mass media in contemporary governance.
Given your interest in popular culture, you decide to take the last of these.
You may be interested in how the fall of the Wall in some way exemplifies
the influence of popular culture on national and supranational affairs. Your

starting point – the germ of your thesis, if you like – may be that politics
as it is practised by politicians operates only as a superficial gloss, as a
kind of leitmotif on the ‘natural politics’ of a nation. Your view may be that
there is always a deeper political current which will make itself felt and
come to the surface to defeat autocracy. (Incidentally, this is not a view
I hold, but it is a perfectly respectable starting point for exploration in a
piece of research and would lead to an interesting study.) With this start-
ing point, you could identify the fall of several such autocracies, pointing
to the role of popular culture in each, then make the point that you are
going to concentrate on the fall of the Wall as an especially interesting
case and one about which there is a great deal of documentary evidence
on which you can draw. Here, then, is the start of your case study.
Figure 2.3 The Berlin Wall: a case study of what? Photo by Luis Diego Hernández
on Unsplash.
Having established your grounds for using the fall of the Wall as an exam-
ple of the influence of popular culture, you need to keep in mind the whole
time that it is the fall of the Wall that is the event on which you are concen-
trating. In other words, don’t slip into a discussion merely of the influence
of culture on the decline of Soviet influence on Eastern Europe. What is
iconic in the fall of this huge symbol of division? How can its fall in some way
exemplify the power of popular culture? You need to make this connection.
Let’s imagine that you brainstorm again and think of connections with
culture and how those connections can be linked with the Wall. How can
(Continued)
a story be made? I hope you can see something clearly emerging from the
brainstorm diagram in Figure 2.4.
Graffiti: the Wall Intellectuals: Wall Celebrities: the Wall

is a lovely space to can stand as an provides a backdrop to be
draw on example of brutal photographed against
oppression
Media: the Wall – these

words alone can be TV: good
used as shorthand for televisual images
the evils of separation The Berlin of bravery or
Wall brutality
Newspapers: human
interest stories about Novels: good
separated families backdrop for
scaling the Wall storylines
Clothes:
black market Painting: the Wall Pop music: machine-
smuggling of provides bold, like, without soul; lends
fashionable iconic statement of itself to particular kinds
clothes over separation for of sounds or lyrics about
the Wall works of art freedom
Figure 2.4 Culture and the Berlin Wall
What I’m hoping you can see is that, from the forms of popular cul-
ture I have identified here – media, painting, graffiti, TV, newspapers,
pop music, novels, intellectuals, clothes, celebrities – there are many ele-
ments to the structure of the Wall, culturally speaking. With a spark of
insight, you may see that these make it especially fertile ground for these
various cultural forms. It’s big; it can’t be ignored; it is screaming out to
be painted on; it stands as a conspicuous symbol of brutality in its scale
by comparison with the buildings on either side. It is bounded by stark,
empty space; it stands also as a symbol of the seeming hatred of human
scale and humanity held by the regime that built it – it’s an icon of a rejec-
tion of individualism.
It is in all of these areas that you can make a special distinction of the Wall –
a reason for it to become a case study that makes it stand out as different from
or similar to other borders.
Now you can draw a storyboard (see Figure 2.5) in which you ‘join the
dots’ about all of this thinking. You connect the relevant and important bits
of your brainstorming.
In the middle of
the city, seen
every day
Features of the Wall
that make it susceptible
Big – can’t be
to cultural influence
ignored
Can be painted on –
looks like an advert for
Characteristics of the evils of oppression
the Wall that make
it different from Scale and concrete
other borders – its solidity makes it a
Exciting stories,
uniqueness symbol of brutality
e.g. about
smuggling jeans,
can be written Politicians and
celebrities can use it
Other border as a photo-op
emplacements
The Berlin Wall
Figure 2.5 Storyboard for the Berlin Wall as a case
The Wall is a symbol of separation and oppression. Here, in Figure 2.5, the
storyboard concentrates on its unique features as far as culture is concerned.
It points to the Wall’s special features vis-à-vis others where separation and/or
oppression may be involved – think of the separation of North and South Korea
or of Lebanon from Israel. What makes them less (or more) iconic cultural sym-
bols than the Berlin Wall? For a start, those barriers are of barbed wire rather
than concrete and are separated from ordinary folk by demilitarised zones that
effectively sterilise them from contact with people and, necessarily, culture.
These latter features automatically make other border emplacements
inert as far as their cultural impact is concerned. By their very nature, they are
insulated from people, whereas the Wall was very much ‘in your face’. People
saw it every minute of every day and it was ripe for use as a cultural icon,
eventually working against the very purpose for which it was constructed. It
was, we might say, its own worst enemy.
The storyboard shown in Figure 2.5 enables you, I hope, to see how, with the
literature, you can go from a loose collection of ideas to a starting point for sin-
gularity and specialness to, eventually, an exploration of that specialness, that
uniqueness, as a case study. I describe the nuts and bolts of drawing a storyboard
in Chapter 11.
This case study in the typology

In this case, we could say that your subject, the Wall, is a key one in the topic
(for it is not one that is within your local knowledge). Your purposes are
intrinsic (since you are interested in this phenomenon in its own right), explor-
atory (since you want to explore something about which you do not know very
much) and explanatory (since you want to explain – to yourself and others).
From here, we can move on to the approach of your study, given these starting
points. Will you be seeking to develop new ideas or test out those of others? Probably
you will be doing both, so the study is both building theory and testing it and this
theorisation comprises your object. It is likely that there will not be much simple
description going on, but you will be seeking to understand the perspectives and posi-
tions of those who lived through the period. In this sense, the study is interpretative.
From here, we could say that, while you are doing a single case study (that is to
say, of one event, namely the falling of the Wall – not the falling of several walls),
you are looking at nested aspects of it, since (let’s imagine) you have chosen to look
at popular music, art and the media as nested elements of popular culture. We can
also say that you are looking at the case retrospectively – you are looking at some-
thing that happened in the past.
The connections between subject, purpose, approach and process for this particular
case study are shown in Figure 2.6. We can say that this case is a key case with purposes
that are intrinsic, explanatory and exploratory, approaches that are testing a theory,
building a theory and interpretative, and processes that are nested and retrospective.
Single
Multiple
Outlier Intrinsic Testing a theory • Nested
Key Instrumental Building a theory • Parallel
Local Evaluative Drawing a picture • Sequential
Explanatory Descriptive • Retrospective
Exploratory Interpretative • Snapshot
• Diachronic
Figure 2.6 Mapping out the design for the case study ‘The role of popular culture in the
fall of the Berlin Wall’
How to go from idea to question to

case study
Race and achievement: A case study of

one school
I draw here on a case study undertaken by the investigative
Idea
journalist Rageh Omaar.
In a Channel 4 film about race and intelligence, Omaar
(2009) started with a set of puzzling and troubling findings
about race and achievement. In essence, his study con-
cerned the significantly worse performance at school of young people of
African American or Hispanic heritage than those who are Caucasian.
It is a well-established phenomenon, but why
should it occur? This was his prima facie question
Prima facie
and was the basis for the analytical frame he would
question
employ in his inquiry.
Omaar first outlined traditional explanations for
this poor level of achievement. Intelligence was (and is often still) taken to
be at the root of achievement – that if you have a high level of intelligence,
it is assumed you will achieve; if you have a low level of intelligence, you
will not do well. He looked back through psychological accounts for dif-
ferences in intelligence, too, and was able to locate a powerful stream
of thought in the early twentieth century that had settled on a suppos-
edly hereditary basis for differences in ability. At that time, there was a
keen desire to calibrate people’s abilities, and this found its expression in
the development of intelligence testing and the IQ
test. All this happened at roughly the same time as Search the
a flowering of interest in Darwin’s ideas, summed literature
up in the idea of ‘the survival of the fittest’. and think
This came to be used by ‘social Darwinists’, in a
terrible distortion of the theory of evolution, as an
argument for sterilising people of lesser ability in a
movement that came to be known as eugenics (Thomas
Preliminary and Loxley, 2007: 34–7).
fieldwork
Although eugenics and the social Darwinism on which
it is based are now considered wholly objectionable,
(Continued)
their legacy resides in a strong undercurrent of belief about the heritability

of intelligence. This is still at the core of many people’s attitudes to success
and failure. In fact, it is probably still the received wisdom. Omaar interviewed
some key psychologists who took this view. The attitude they hold extends to
beliefs in differential intelligence for different ethnic groups, with the assump-
tion that people of black and Hispanic ethnic origin tend to be of inherently
lower intelligence than white and Asian people.
Omaar’s study was now beginning to take shape and harden. He was able
to refine his research questions with the benefit of wider reading and new
ideas. One of the ideas he took forward was that of the so-called ‘Flynn effect’.
James Flynn is a researcher who has noted that intel-
ligence, which is supposed to stay fairly stable in both
individuals and populations, is, in fact, rising. The rise is Wider
disguised by the fact that average IQ scores always stay reading
at 100. This maintenance of the score resolutely at 100
happens because of a standard practice in psychomet-
ric science – a periodic ‘renormalisation’ of the data from which IQ scores
are calculated. This ‘normalisation’ is key to the way our understanding of
intelligence is distorted, and it was normalisation that Omaar used to lever
his way into the debate.
So, what is normalisation? IQ scores are based on the idea that the values
of most things in life (height, weight and so on) are normally distributed. In
other words, there are the most occurrences in the middle of the range (most
people are of average height, for example) with relatively few occurrences at
either end of the range (relatively few people who are very tall or very short).
In between are intermediate numbers of people and the distribution can be
set out in graph form to produce the curve shown in Figure 2.7. This ‘normal
distribution curve’ describes the way in which most attributes in nature are dis-
persed among a population. Following on from this, the assumption among
the early psychologists was that this should also be the case for intelligence.
Number of people
Value of height, weight, etc.
Figure 2.7 A normal distribution curve
This assumption about normal distribution is all very well for height and
weight, which are unproblematically measured, but – as I’ve been at pains
to point out throughout this book – things are rather

Carry on
searching more complicated in the social sciences and, as far
and thinking as intelligence is concerned, we cannot just measure
it with a ruler or weighing scales.
In fact, the idea of intelligence has been constructed – made up – and,
in its construction, because of the assumption that intelligence is a normally
distributed phenomenon, there has had to be a correction of the raw scores
people get on tests to produce revised scores (called ‘standardised scores’)
that conform to the normal distribution curve. The standardised scores make
it look as though what people are achieving is normally distributed and,
because of normalisation, that this distribution is staying constant over time.
What Flynn (1987) showed, in an article called ‘Massive IQ gains in 14
nations: what IQ tests really measure’ (note also his later work: Flynn, 1998,
1999, 2003; Dickens and Flynn, 2001), was that ‘intelligence’ is not stable if
you go back to the prenormalised, raw figures. People’s ability to do ‘think-
ing tasks’ is rising. Amazingly, Flynn was the first
person to make any kind of a song and dance about Keep carrying
this and what it meant for the notion of intelligence. on searching
and thinking
Interest in Flynn’s work has centred on why the rise
should be taking place, but more interesting – certainly
as far as Omaar’s developing research question was concerned – is the light
that it throws on the construct of intelligence. If intelligence is not the stable,
normally distributed phenomenon it was once thought to be, it cannot be used
as the basis for explanations about why people of different ethnicity fare better
or worse at school. It is more satisfactory to use the far simpler explanation that
we get better at the things we practise. If our lives at home give us no chance
to practise the things that are demanded of us at school, then we are almost
bound to do less well there. Given that black and Hispanic children are dispro-
portionately from poorer families where there is likely to be less of that kind of
school-related activity going on, the explanation for their poorer performance
can be explained straightforwardly by background rather than genetics.
The literature review of this theoretical background – this analytical
frame – provided the backdrop for the case that Omaar was making – a
theory-building and picture-drawing case about the
Reformulated influence that school can have on countering expec-
research tations about achievement. He could reformulate
question his prima facie question to ask a completely new
one: ‘What influence can school have on countering
expectations about achievement?’ In answering this, he would need to draw
on the analysis conducted as part of his literature review and he would draw
also on empirical work: the case study.
(Continued)
Omaar’s next step was to go on to identify a school

that had bucked the trend – an outlier case that had Seek out
‘outlier’ case
shown that a catchment of entirely black and Hispanic
children could achieve extraordinary feats given
favourable circumstances at school. The school he
chose was Hostos-Lincoln Academy of Science in the Bronx, New York. This
is one of the poorest and toughest areas in the USA,
with 80 per cent of its students eligible for free or
Begin
fieldwork reduced-price school meals.
Hostos-Lincoln Academy is 1 of 36 schools in its
locale. It is a public school (that is, a state school),
serving 534 students in grades 6–12 (final year of primary school until end of
secondary, staying to do school-leaving exams). All of its students are black (23
per cent), Hispanic (73 per cent) or of Asian/Pacific island (4 per cent) origin.
Despite the problems of poverty and deprivation,
Hostos-Lincoln Academy is getting its students into
Begin to ‘drill
down’ in some of the best colleges in the USA. This achieve-
fieldwork result ment cannot be explained by better resourcing, for
the school has 17 students for every teacher, whereas
the average for New York is 13 students per teacher.
Assessment results show just how well the school is doing with its stu-
dents. Of the school’s students, 52 per cent met or exceeded the standards
in English language, while the average for schools in
the same district of New York is 31 per cent. In maths,
77 per cent of Hostos-Lincoln Academy students met More
or exceeded the standards, while the average in the drilling down
district was 51 per cent. In science, 38 per cent of stu-
dents met or exceeded standards, while the average
for schools in the district was 29 per cent. It was like this
in subjects right across the board – social studies, chemistry, Earth science,
living environment and so on.
This demonstration of difference, however, is not what is at the heart of
this case study. Just showing that the school is different is of no interest
without what follows. The heart of the research lies in the in-depth analysis
of what happens in the school that means it achieves the results it achieves.
Omaar concentrated on what was going on in the school that enabled it
to help its young people so effectively. He interviewed the principal of the
school, Nicholas Paarlberg, visited classrooms, made observations, talked
to students and teachers and took particular interest in one 13-year-old stu-
dent, Kwan, who became a nested case within the larger one of the school
(see page 191 for a discussion of nesting).
Omaar’s conclusions relate high expectations, affirmation and students

being given access to a culture outside their own. When the students gradu-
ated, they were allowed a portion of the main corridor wall to write their
name on to show that they had graduated (see Figure 2.3). In the nested
case study of Kwan, Omaar showed that this student lived, ate and slept
in a tiny one-bed apartment, but at Hostos he was ‘transported to another
world, and he is thriving on it’.
This was a well-designed piece of research. Omaar linked important ele-
ments to produce a nicely integrated case study. He had:
1. noticed something worth investigating – namely, the consistently poor

performance of black and Hispanic students at school
2. read around the area and developed a hypothesis, or theory, as to why
this should be
3. looked to find a counter-example, or outlier case, where the usual (that
is, the poor performance of these students) did not happen
4. undertaken a case study of the outlier case, by means of which he sought
to explain the better performance of the black and Hispanic students in
the context of the theory he had developed on the basis of his reading
5. used the case study as both an explanation and illustration.
Omaar’s work was multilayered, serving different purposes and taking differ-
ent approaches. Its route through the typology I outlined in Chapter 1 is given in
Figure 2.8. Given his starting point – about the provenance and place of the notion
of intelligence – he had to conduct contextualisation via the relevant literature to
build a potential theory which he would then test using his case study. The theo-
retical position which he built, on the basis of the work of Flynn and others, was
that the idea of intelligence had been misleading: it had led to the belief that it was
genetically fixed and that there was little that schools (or other cultural institutions)
could do to counter its effects. His object was to throw light on this. His identifica-
tion of a subject therefore rested on the ability of that subject to contradict standard
explanations resting in intelligence. That identification came in the discovery of
Hostos-Lincoln Academy, an outlier to the general rule that schools succeed in rela-
tion to their student intake. Not only was a theory developed and tested by the case
study, but it also served as a powerful illustration for Omaar’s purposes as a researcher
and a journalist. The study was a single one, taking a snapshot of the Academy as
Omaar visited it and also looking at its history, its mission and its records and at the
official statistics of its environs and the other local schools – looking at it in other
words retrospectively. But there were multiple and nested elements to this in the
particular focus on certain of the school’s students. The methods Omaar chose to
use included interview, observation and examination of local, national and school
records. All of this is summarised in Figure 2.8.

• Retrospective
Single • Snapshot
• Diachronic
Methodological
Intrinsic Testing a theory
choices
Object
Local Instrumental Theory building

Key Evaluative Illustrative/
Outlier Exploratory Descriptive
• Nested
Multiple • Parallel
• Sequential
Figure 2.8 The organisational structure of Omaar’s case study of the Hostos-Lincoln
Academy
Your question, and the analytical frame associated with it, will evolve, as Omaar’s
did. It will progress from a prima facie question to a final question. On the way from
A to B it may become transformed as you read, conduct pilot work and develop new
ideas. This process is not unique to case study – it is there in all investigative work in
any field. Einstein famously did it using his ‘thought experiments’ (Norton, 2004).
Becker et al. (1961/1980: 33) describe the way that one relatively simple question
led to another, more complex and analytical one in their classic case study on the
training of medical students, Boys in White:
We had first the descriptive question: How much effort did students put
forth and in what directions? Then we had the analytic question: Why
were the students’ level and direction of effort what they were and not
any of the other things they might have been?
Questions and different approaches to

research
Beyond understanding that questions develop – they may cultivate ‘forks’ or ‘layers’ –
it is important to understand that there are various kinds of questions, which lead
to different kinds of research. I like to think of these questions being of four basic
kinds:
1. What is the situation? You are describing something.

2. What is going on here? You are trying to understand what is happening in a par-
ticular situation.
3. What happens when …? You introduce a change and look to see its effects.
4. What is related to what? You examine how one thing is related to another.
Typically, people think of these questions leading to different kinds of research

approach and research design – and they do. What is interesting about the case
study, though, is that it can encompass different kinds of approach to research. It is
like an umbrella: it covers a range of ways of doing research. It does this because, as
I have said, it is about a focus rather than an approach. Nevertheless, I like to think
of the case study as a design frame in its own right, even though it does cover this
range of approaches and methods, which I look at in more detail in Chapter 8. I like
to think of it in this way because it has integrity as a way of going about research, as
a way of approaching the singular. Let’s think a little more about the case study as a
design frame and how it interconnects with your research question.
Questions and design frames

Oranges are not the only fruit, and the case study is not the only design frame. It is
one of a large number of research design frames that can be adopted in an inquiry
to answer the four kinds of questions I have just laid out.
So, there are many design frames that can be used and – just to make things con-
fusing for the first-time researcher – they are not wholly separate from each other,
they can nest into and envelop each other. In this sense, it is not very easy to map
kinds of research questions onto kinds of design frames.
With this caution in mind (and I’ll try to sort out the confusion in a moment),
let’s outline a few of the design frames most commonly used in small-scale research:
• action research
• case study
• comparative research
• evaluation
• experiment.
Let’s also consider what they are good for, because they all have their advantages
and disadvantages. Each is better or worse when focusing on different questions. For
example, if you are concerned with one of the ‘What happens when …?’ questions –
that is, if you are looking to see whether or not X causes Y – you will probably want
to use some kind of experiment. If you want to discover whether or not some new
method of management has had an impact on industrial relations in a particular
factory, you will embark on an evaluation. If you want to improve your practice,
doing this through research, you will set off on a piece of action research. If you
want to compare phenomena in different countries, you will embark on comparative
research. It’s horses for courses. Let’s match up some of the design frames to kinds of
questions and kinds of study.
Table 2.1 Questions, purposes and design frames
Design frame Purpose – especially good for … Kinds of questions

Action research Helping to develop practice What is the situation?
What is going on here?
What happens when …?
What is related to what?
Case study Understanding the details of what is What is the situation?
happening What is going on here?
What happens when …?
What is related to what?
Comparative Looking at different situations and making What is the situation?
research comparisons
Evaluation Seeing if something is working What happens when …?
What is the situation?
Experiment Establishing causation – does X cause Y? What happens when …?
You will note from Table 2.1 that certain kinds of design frame, such as the experi-
ment, are limited to one kind of question (does X cause Y?), while others, such as the
evaluation, are appropriate for more than one kind of question. Others, such as the case
study and action research, may cover all of the kinds of questions that I have outlined.
This is because a case study is defined by ‘singleness’ rather than procedure. There
are no procedures that are associated exclusively with the case study, for, in its
examination of the singular, the one thing, it is ‘allowed’ to use whatever proce-
dures and methods it wishes in doing this. The emphasis when doing a case study
is on the singleness. In other words, the case study is not a method, nor is it a set of
procedures. Rather, it is a focus.
Design frames and methods

Inside the design frame, you will be using one or more of a variety of methods to
answer the questions that you pose. As far as this is concerned, the case study has
broad and capacious arms: it loves all methods. It is perhaps going too far to say
that it is promiscuous, but it is, shall we say, generous in its affection for methods –
observation, diaries, questionnaires, tests, statistics, interviews, whatever …, the
options are limitless.
Likewise, with different elements of the case study, you may use other design
frames under the umbrella of the case study – in this sense, we are talking about
frames within frames.
The important thing is that you are using a case study to examine your case in
detail. In doing this, you choose whatever methods and subsidiary design frames
you can think of to help answer questions about your singular case. If you look
at some informative and important case studies – such as Ball’s (1981) Beachside
Comprehensive or, in a completely different field,
the investigation of the variant Creutzfeldt– The case study is like a scrapbook or
Jakob disease (vCJD) cluster in Leicestershire portfolio of sources and information.
following the ‘mad cow’ epidemic (outlined

on page 136) – you will find that a broad range of methods is used and a catholic
attitude is taken to the use of methods. It is not as if just one or a few meth-
ods are ‘allowed’ in the case study. Rather, when we think about methods, it is an
issue of looking to see what is best in terms of answering the questions that arise.
Observation? Questionnaires? Sociograms? The choice is yours.
It may help to think of your case study as a scrapbook or portfolio. The case study
is everything that you choose to put in it.

take this …
• A piece of research is built around a question, it is not built around a method.
Anyway, the case study is not a method – it is a wrapper for different methods. It is
the focus that is special to the case study – a focus on the singular.
• Build your case study around your question – the case study is a means to an end
(that is, answering a question), not an end in itself.
• Do a literature review early on in your research to find out as comprehensively as
possible what research has already been done on your topic. You will then be in a
position to adapt or modify your original questions accordingly.
• Remember that case study is not a method. Rather it is a design frame – a scaf-
fold for your research. Within this frame, this scaffold, you can use a wide range of
research methods.
• When building your case study, always keep in mind the need to establish a storyline.
The key is to draw rich, interconnected information from your singular focus and derive
unique insights from the analysis that follows. You can use storyboards to help you do
this – to move from what might be a loose collection of ideas to a clear narrative.
REFLECTIVE ACTIVITY
Read again the boxed case study in this chapter of race and school achievement
conducted by Rageh Omaar. Omaar conducted his case study by examining the
outlier case of an outstanding school that bucked the trend of worse school per-
formance with black and Hispanic students.
Starting with the same prima facie question as Omaar about the worse
school performance of children from some ethnic minorities, outline a different
case study design that might effectively address the same research question.
Alternatively, outline a case study design about social mobility more generally.
Use the typology from Chapter 1 to help you think about structure. Tips: You
might choose to look at a child, class or school that you know. You might look
at the work of a research organisation such as the Sutton Trust, which focuses
on social mobility, and identify a subject such as the composition of the current
government in terms of the educational histories of its members.
Further Reading
Ball, S. (1981) Beachside Comprehensive: A Case-Study of Secondary Schooling.
Cambridge: Cambridge University Press.
Lacey, C. (1970) Hightown Grammar. Manchester: Manchester University Press.
These are classic examples of the case study. Each gives a detailed analysis of what
goes on in one school over a period of time. They exemplify the kinds of very
diverse data gathering that can be done in a case study and the ways in which the
case study work can be linked intelligently to national policy.
Becker, H.S., Geer, B., Hughes, E.C. and Strauss, A.L. (1961/1980) Boys in White.
Edison, NJ: Transaction Publishers.
Another classic. This is about American medical students and the way that they
work. It is a good read and tells you as much about the process of the case study as
it does about the training of the medical students.
3
THE WHOLE IS MORE
THAN THE SUM OF
THE PARTS: SEEING A
COMPLETE PICTURE
This chapter offers an intellectual history of the case study – a pedigree, showing
its scholarly heritage. This may not help much with the nuts and bolts of doing
your case study research, but it will help, I hope, in understanding where the case
study has come from and where it currently stands as an approach to inquiry. You
need to know this if you want to understand why the case study is appropriate and
meaningful for addressing a research question.
Looking at the whole represents something of a new – or at least a reawakened –
worldview. It is one that emerged during the twentieth century with the purpose of
understanding phenomena not as so many disconnected parts but, rather, as intercon-
nected elements. Anyone wanting to do a case study implicitly shares this worldview,
though it should be said that the case study does not boast an intellectual allegiance
to any one strand of thought but many. I shall review some of these in this chapter.
Break things down or see them as wholes?

That the case study is not superglued to one of these schools of thought may make
it seem like a bit of an intellectual orphan, which is part of its problem as it attempts
to establish its credibility. While it may not adhere to a particular stream of thought
or school of thinking, its emphasis on the whole – the holistic – puts it in some
respected scholarly company. All of this company share a wish to see things as
wholes.
The wish to see things more as wholes does present a number of challenges,
however, since the thrust of most scientific inquiry since the Enlightenment (that
is, since the beginnings of rational and scientific thought in the mid-seventeenth
century) has been about breaking things down. It has been about reducing them
and understanding them in terms of their constituent parts. It has, in other words,
been about reduction, so is known to philosophers as reductionism.
Reductionism has done wonders in
the natural sciences and the technol-
A H E A V Y W E I G H T C O N T E S T ogy that springs from it – people on
REDUCTIONISM
RONNIE ‘THE REDUCER’ the Moon, refrigeration, antibiotics,
computers, to name but … er … four.
Viewing the world from this position,
V phenomena are nothing more than
HO M
HARRY ‘WHOLESOME’ the sum of their parts.
L I S The starting point taken in the case

study, by contrast, is that certain phe-
nomena are more than the sum of
Figure 3.1 Reductionism and holism: each their parts and have to be understood
has its own advantages as a whole, rather than as a set of inter-
relating variables. How has this holistic
position fared next to the super-successful reductionism? Also, given that reduction-
ism has been so successful, why would anyone want to study things differently?
This chapter is really about the attempt to view and analyse phenomena as more
than the sum of their parts and about the intellectual traditions, mainly in the twen-
tieth century rather than before, that have competed against the highly productive
reductionist tradition. This has, of course, been a bit of a David versus Goliath con-
test. There are a number of themed but unrelated elements to this understanding of
things-as-a-whole that I shall look at briefly – each in its own different way about
the importance of wholeness in inquiry.
The issue is more complex than simply seeing the world as indivisibly complete
versus the world as necessarily divisible and reducible. Associated with the assump-
tions being made here are beliefs about the purpose of inquiry. In other words, why
are we doing the research?
Some have said that the purpose of research is to develop laws and theories with
which we can explain the world and predict what is likely to happen next. Others
have said that this is fine and dandy in understanding the things of the natural
sciences (and it has got us to the Moon and so on, as I mentioned before), but it is
less adequate in understanding the things of the social world. Here, reductionism
is of less value.
SEEING A COMPLETE PICTURE 51
The issue about whether or not reductionism is of value is as old as the hills –
the Athenian hills. Plato, Socrates and Aristotle had major disagreements about it.
Plato thought that universal truths developed from generalisation were what we
should all be searching for if we want true knowledge. He says, ‘I am looking for
the simile in multis.’ He is looking, in other words for the essence that captures the
truth evinced by many cases – for example, I’ve only ever seen white swans, so all
swans must be white.
On the other side, Aristotle averred that we progress only by using our practical
reasoning, craft knowledge or tacit knowing – the stuff we know because of our
experience. It is the stuff we learn ‘on the job’ or by simply being alive and it is
explicable only in terms of the particular – the case. It cannot be communicated by
reducing it to general principles. ‘You are wrong, Socrates and Plato,’ says Aristotle,
‘to dismiss the value of cases in the production of knowledge.’
The arguments on this theme continue to the present day. They proceeded
through Base Camp 1 of the Renaissance, with Francis Bacon – the first scientist,
some would say – averring that the generalisers and regularisers, the followers of
Plato, ‘hasted to their theories and dogmaticals, and were imperious and scornful
toward particulars’. Today, the argument has unfortunately come to be known as
the ‘paradigm wars’ (Oakley, 1999) and shows little sign of abating. Any argument
that goes on for 2500 years has clearly got legs and we are not going to resolve it
any time soon.
It was Plato whose ideas won (broadly speaking) and whose thinking took hold
so successfully for the natural sciences. This was a big defeat and a big loss for
our ways of understanding, asserts an astute commentator on these issues, Bent
Flyvbjerg (2001). Flyvbjerg says that the mistake led humankind on two millennia
of false starts in understanding social phenomena. The mistake was in the failure
to distinguish between different kinds of inquiry for different purposes and it leads
us, in extremis, to the absurd position that it is inappropriate to argue or learn from
particular examples for fear that this might be thought ‘anecdotal’ and, therefore,
unscientific.
The distinction, then, has been on a long journey intellectually. In essence, it is
about the relevance of particular individual events to a larger picture and the extent
to which we can use separate events to establish laws or theories by which we can
explain and predict. Some say that it is legitimate to generalise in the natural sciences,
but not in the social sciences. In the former, that is physics and chemistry, the stabil-
ity of the events studied is such that laws and theories can be worked out in a way
that they will prove useful for explanation and prediction. In the human sciences,
however, events follow such random twists and turns that any attempt at establish-
ing stable laws and theories that will reliably explain and predict is meaningless.
The classic argument here has been posed by the philosopher Alasdair MacIntyre
(1985) in his book After Virtue. If you are interested, read Chapter 8 in that book,
‘The character of generalizations in social sci-
Holism is seeing things as a whole. ence and their lack of predictive power’. It is
(Why hasn’t it got a ‘w’? Don’t ask as entertaining as it is enlightening.
me.) There is an assumption that
the whole is more than the sum of
I’ve said that this argument about reduction-
its parts. It is about seeing things in ism versus holism has been going on for 2500
their entirety, their totality. years and reductionism has certainly won as
far as the main method used for the natural
sciences is concerned. In fact, it is only in the last 100 years or so that protagonists of
Aristotle’s views have come to the fore again in thinking about the human sciences.
Earlier, right at the beginning of the book, I mentioned the idiograph – a draw-
ing or figure that stands not for a sound, as do symbols in the Roman alphabet, but
for an idea. It was the philosopher Wilhelm Windelband who, at the cusp of the
nineteenth and twentieth centuries, drew on the notion of the idiograph to sepa-
rate what he called the ‘idiographic’ and the ‘nomothetic’ (from the Greek nomos,
meaning ‘law’) in social inquiry.
I personally think that it would have made life easier if Windelband had called
the ‘ideographic’ the ‘pictographic’ since ‘pictograph’ is a perfectly acceptable alter-
native to ‘idiograph’ and it more straightforwardly explains what he was trying to
get at. A pictogram is a small picture that has become stylised, so that it represents
the thing in written form. The Chinese pictogram for ‘person’, for example, looks
a little like a matchstick person, with a trunk and two ‘legs’. (Interestingly, these
picture representations eventually take on sound-like representations as well, devel-
oping into our modern system of alphabetic writing.)
With the idiographic, the approach is to specify and study individual phenomena
in detail – we have a ‘picture’ in front of us. With the nomothetic, by contrast, the
approach is to generalise from many cases and derive laws from such generalisation.
Windelband was not the first to notice this difference but, as is so often the case
in the history of ideas, he encapsulated something nicely in his separation of the idi-
ographic from the nomothetic and the distinction he drew has stuck. It has proved
enormously helpful in summarising the differences between these kinds of inquiry.
Gestalt psychology
Not long after Windelband, one interesting avenue down which this discussion
travelled was that of Gestalt psychology early in the twentieth century. This was
key for the case study since the essence of Gestalt thinking is that things should be
seen in their totality, as we try to do in a case study. The key idea to emerge here
was that the mind works not by perceiving things in isolation, separately, but,
rather, as wholes – as integrated units. We humans, alone among animals, invest
meaning in what appear to be unrelated phenomena and make patterns, make
sense, out of these. It is almost as if our minds are patternmaking or sensemaking
machines.
The key figures here – psychologists such as Max Wertheimer and Wolfgang
Köhler (and, later, the great social psychologist Kurt Lewin, the inventor of action
research) – suggested that the methods psychologists use to study psychological
phenomena distort our understanding of them. By always reducing, atomising the
subject matter of psychology into constituent variables, the end result is a mis-
understanding of the way the mind works, they said. The mind works not in the
convenient building blocks of psychological science – variables such as gender, age,
reaction speed, time, class – but wholes.
It is certainly true that Homo sapi-
ens is a pattern-finding animal. The
fertility of our brains in divining
patterns is boundless – it requires
only the mildest encouragement
for shape and form to crystallise
out of what may seem at first to be
chaotic stimuli. Look at Figure 3.2.
After looking for a while, you will
see a Dalmatian dog. (Can’t see it?
Its nose is in the middle of the pic-
ture and its backside towards the
top right.) In fact, perhaps danger-
Figure 3.2 We have a wonderful capacity for
ously for certain kinds of inquiry, seeing shapes and making meaning
we even seem to be predisposed to
preferring to see form where there is none rather than not see any form at all.
The philosopher Karl Popper suggested that this facility may be inborn. It may be
due, he proposed (1977: 270), to ‘mechanisms which make us search for regulari-
ties’ and be responsible for what he called ‘the dogmatic [ouch] way of thinking’. In
a similar vein, Barrow (1997) points out that the human brain has an evolutionary
imperative to over-see patterns – that is, occasionally see patterns where they do not
exist (as ‘false positives’, for instance ‘seeing’ slender shadows as tiger stripes) – as
this has historically given us an evolutionary advantage over the opposite state of
not recognising patterns where they in fact exist (where the slender shadows are
actually the stripes of a tiger).
The bottom line is that we must be cautious about this precious ability to see
pattern since, today, harbouring a predilection to make shape and theory out of
that which is shapeless may not be quite so useful. We may overdo generalisation,
tending to generalise from insufficient information. It is this tendency that has to
be guarded against in all of the sampling procedures that have been developed with
what has been called probabilistic research.
Case study’s potential for offering broad generalisations may be limited. However,
in a case study you can use your ability to put things together, to draw from expe-
rience, make informed judgements about cause and effect in this particular case.
It is what the human brain is good at. It is what got us out of the trees. It is what
Newton, Darwin and Einstein did. We should not be ashamed of it. It is about con-
necting ideas, like the ganglia between neurons.
There are a number of superstructures that have been used to accommodate
wholeness, intellectually speaking, and I shall look at these in the sections that
follow.
Dramas, theatres and stages

All the world’s a stage,
And all the men and women merely players;
They have their exits and their entrances;
Shakespeare sums up here several streams of thought about how the world – the
natural as well as the social world – should be seen. So, it is worth saying a little
more about the stage analogy.
In fact, the As You Like It quotation could perhaps be the motto of the case
researcher because the ways of thinking about social situations that I am going to
review in this section are all characterised by seeing the person in context, where
the action is defined by interactions between people and situations, which, of
course, is the sine qua non of case study research. Each person is affected by the
environment around them: no one is an island.
Interpretivists say that there is no There is no better place to begin to explore
‘objective’ social world ‘out there’. this idea than with the work of sociologist
Rather, it is constructed differently
Erving Goffman, who actually called his view
by each person in each situation they
face, so it is useful sometimes to see of social life dramaturgy. In the dictionary defi-
the world as a stage on which we nition, ‘dramaturgy’ is the art of writing and
play out characters.
producing plays and, for Goffman, life is a set
of dramatic performances in which people take on roles and change the way they
behave depending on their interactions with other actors. I am a different person
with different folk. I take on a different character with my sister Adrienne from the
one I take on with my friend Peter, from the one I take on with my boss Edward.
The core perhaps is the same, but if ever I were to be faced with the three of them
in one room (heaven forbid) I would suffer from an acute case of triple personal-
ity implosion syndrome – or PIS3, as it is known in the psychiatric literature (that’s
meant to be a joke, by the way).
Goffman differentiated between the kinds of situations in which such performances
are necessary, noting that there is a ‘front stage’, where the ‘act’ is always used in order
to ‘impression-manage’, and a ‘backstage’, where people can be more ‘themselves’.
Goffman’s dramaturgical perspective was itself an offshoot, an important one, of a
branch of sociology known as symbolic interactionism. Here is how one of symbolic
interactionism’s main proponents, Herbert Blumer (1992: 82), presents what he calls
the ‘basic premises’ and ‘methodological consequences’ of this school of sociology:
human beings interpret or ‘define’ each other’s actions instead of merely

reacting to each other’s actions. Their ‘response’ is not made directly to
the actions of one another but instead is based on the meaning which
they attach to such actions. Thus, human interaction is mediated by the
use of symbols, by interpretation, or by ascertaining the meaning of one
another’s actions.
Symbolic interactionism is, in turn, a tributary of a wider stream of thinking known as

interpretivism. The basic assumption here is that the social world is constructed by each
of us differently, with words and events carrying different meanings for each person
and in each situation. Interpretivism started, broadly speaking, with the American
sociologist George Herbert Mead, and I’ll examine it in more detail in Chapter 8.
Though not directly linked, a forerunner to these lines of thought was the phe-
nomenology of Edmund Husserl, which emerged at the beginning of the twentieth
century. This is generally thought of as a philosophical rather than a sociologi-
cal movement, but it is connected to the thinking of the interpretivists and the
symbolic interactionists in its emphasis on meanings. Actions, Husserl would say,
cannot be broken down into their constituent elements to understand them. They
are understandable only in terms of the meanings that actors impose on them.
They are understandable only, in other words, in terms of the wholeness of the
context in which they occur.
In fact, a clutch of philosophers, and those in the infant disciplines of sociology
and psychology, were saying much the same kind of thing at this time. Wilhelm
Dilthey came up with the idea of Verstehen (meaning ‘understanding’), more ade-
quately to demarcate between the stuff of the natural and the human sciences and
our ways of inquiring into them. We cannot, he said, ignore the historical context
and the context of meaning in all that human beings do and say.
Ecological psychology
Another manifestation of the need for wholeness is seen in ecological psychology.
In its ‘green’ sense, ecology refers to the relationships between living things in the
environments that they inhabit. It is the study of the complex interrelationships
that take place in those environments.
A starting point here is that the lives of organisms inhabiting these environments
are inextricably intertwined with the nature and quality of those environments.
Changes in any small aspect of the environment will have knock-on effects that will
affect the lives of the organisms populating it. An important aspect of this is that
there will be an equilibrium in this environment and it will be maintained. If foxes
become blighted, their main meal, rabbits will breed like … er … rabbits, with few
predators, so the remaining foxes do well, eating the superfluity of rabbits, so then
rabbits decline, causing foxes, in turn, to decline, then rabbits breed … and so the
cycle goes on (see Figure 3.3). Thus, the equilibrium is maintained, and the same hap-
pens in the environments (family, school, factory) that we inhabit as people.
So, ecological psychologists would say, people – that is, humans, persons, Homo
sapiens – live in a multicoloured world. It is an ecology where an increase in one
stimulus invokes not an automatic reflex or a tropism, but a thought-out reaction
or an overreaction.
By using the ecological metaphor you can see the relevance for inquiry in social
science: it is not as if you could set up a meaningful research study to assess the
effect of turning up the volume of music on the volume of people’s conversation.
If you got some volunteers conversing in a room where rock music is playing qui-
etly and start turning up the volume, you could measure the effect on the volume
of the volunteers’ voices. I guess that you would find that the volunteers’ voice
levels rose (much psychological experimentation is of this kind). An ecological
psychologist would say, however, that the real world is not like this. In the real
world, people would just leave the room to talk or someone would go and turn the
volume down. Then someone would go and turn it up again and perhaps there
would be an argument, then maybe a fight, then someone would be taken to hos-
pital, and so on.
Numbers
Time
Rabbits Foxes
Figure 3.3 Equilibrium is maintained in an ecology
The real world, then, is complex, and it is the real world in which we are inter-
ested. The understanding that ecological psychology offers is that the way people
behave in a psychological laboratory, with its tight control of a small number of
variables, does not remotely represent the way people behave in the real world.
The basic idea is the same as the one behind symbolic interactionism, which is
that no behaviour occurs in a vacuum; it occurs in the context of others’ activity, in
the context of the family, place of work, school or community and in the context
of the linguistic, cultural, legal and physical environments. Because of the way that
psychology and its inquiries typically proceed – by isolating variables, controlling
important ones and manipulating one or two – we can get into a rut of thinking
that they, using these controls and manipulations, are valid representations of what
happens in our everyday lives. We come to think that the world in which we live
is really like this – like a psychological laboratory with a simple set of variables that
are all more or less inert in their relationships with one another.
The traditional focus on the individual in psychological endeavour permeates
the applied field, where interventions also often take place in a vacuum, with scant
consideration of the physical and social environments within which individuals
learn and behave. The ecological metaphor used by some psychologists since the
mid-twentieth century enables an understanding that things are not that simple.
Behaviours may emerge in an environment that are a consequence of conditions
outside the narrow ambit of the individual. Changing one or two features of this
system may have unanticipated consequences or may have no effect at all.
Adopting an ecological model is like using a wide-angle lens, enabling an exami-

nation of the wider situation that surrounds the target and encouraging analysis
in the wider situation. In essence, an ecological view enables a recognition that
people inhabit a variety of contexts, each one impinging on the other. Any attempt
to oversimplify the richness of the ways in which these contexts interplay is bound
to have shortcomings. An ecological view says that it is impossible to disentangle
the functioning of a constellation of phenomena simultaneously interrelating in
any social arena.
It is Roger Barker (1968) who has generally
Ecological psychology is about the been credited with introducing the notion
complexity of the psychological of ecological psychology. He suggested that
world. Everything is connected, so
it is difficult to assess the effect of
behaviour settings have both static and
changing one variable. Things tend dynamic attributes. On the static side, the set-
to stay in equilibrium. ting consists of one or more standing patterns
of behaviour and milieu – the latter comprising
elements such as house, classroom, a windy day. On the dynamic side, the set-
ting comprises the interpersonal relationships among people and groups. Science,
Barker says, attempts to keep apart the static and the dynamic, the physical and the
behavioural, but their interaction is the central focus of his attention. He gives the
name synomorphs to the settings in which these interactions occur.
Barker clearly intended his model to be useful in the real world rather than
the laboratory. By taking as a starting point the integration of behaviour and
environment, he intends to be able to predict real-world events and solve real-
world problems. Unfortunately, as you will realise if you follow up his work,
the Byzantine complexity of his model and the impenetrability of his language
render his aims pretty well unachievable.
Amidst Barker’s complex analysis is a continual reiteration of the influence
of the physical environment and its interaction with a range of other influ-
ences, such as the ideologies of the participants, and the procedures, practices or
rituals associated with certain synomorphs. It is notable that Barker and others,
such as Gregory Bateson (1972/1999), who promoted an ecological thrust to
understanding people in their environments, had a multidisciplinary scholarly
background themselves and promoted intellectual interchange with different
fields.
Another important figure in this area is Kurt Lewin, whom I mentioned earlier.
Lewin was a bit of a psychological polymath – brilliant ideas came sparking off
him almost as if he were a human van de Graaff generator. He was just a fizzing
lightning ball of brilliance and, among other things (in his spare time, probably),
he invented action research.
In the current context, he came up with an idea that he called ‘field theory’,
which is linked as much to Gestalt psychology as it is to ecological psychology, but
I’ll mention it here. Lewin (1951; 1997) suggested that the ‘field’ of an individual’s
psyche comprises a range of social and psychological factors and changes with the
person’s experience. An alternative term he uses for fields is ‘lifespaces’ – home, job,
school, sports club – that we occupy at different times and in different ways. All are
joined by ‘vectors’ that operate at different levels of strength.
Lewin, with his mercurial (and maybe slightly manic) mind, perhaps tried to
make things too mathematically precise, but with field theory he offered another
stratum to the developing bundle of ideas about holism.
One of the main proponents of ecological thinking was Urie Bronfenbrenner,
but since he developed his model from a hybrid of ecological and systems thinking
(covered next), I shall deal with the latter before going on to talk about his ideas in
more detail.
Systems thinking
Systems thinking is a way of viewing the world that lends itself naturally to the case
study since it aims to see the system as a whole, in much the same way that you
aim to do in a case inquiry. The difference is that systems theory and its various
branches are bounded by some quite firm frameworks about viewing the interrela-
tionships that exist.
With the case study, there are no methodological prescriptions. By contrast, sys-
tems theory proffers methodological tools and route maps to assist analysis. It is a
particular way of studying the whole, unlike the case study. That is not to say you
cannot use the methods of systems thinking in a case study – you can, and they
may prove to be very useful for structuring your thinking. For this reason, I go into
some detail on a branch of systems thinking in Chapter 11.
Systems theory was developed in a variety of contexts, from biology to engineer-
ing, and thus can be relevant for a range of foci and adapted for use in each of these.
If your interests are in engineering or biology or an applied science, or if you are
interested in the interrelationship between one of these and the social world, it may
well be worth your exploring aspects of systems theory.
The models for understanding that have been advanced in systems thinking
were developed specifically to address the need for holism in getting to grips with
complex dynamics. As with the parallel trains of thought in psychology and sociol-
ogy that I have just looked at, the finger is pointed at the barrenness of attempts
to understand complex behaviour by reductionist methods. Our understanding,
systems theorists say, has to push that complexity to the front: we have to try to
understand the ability of the complex system to adapt, change and ‘learn’ from
experience. Systems theory used ideas from biology, such as negative entropy (the
ability of a system to make sense, re-form itself and maintain an equilibrium), and
from mechanics, such as feedback (the effect of a signal going back into a loop and
magnifying itself each time it returns). These are characteristics of systems that are
not seen in individual parts of the system.
One of the first proponents of systems theory was Ludwig von Bertalanffy (1950) with
his general systems theory. At the risk of getting repetitive, the same theme emerges – one
of complex interactions – though von Bertalanffy did not leave it at that.
He made an important distinction between what he called open systems and
closed systems. A closed system is one such as an engineering system (think of a
diesel engine) that has very little in the way of a change to its movement once it is
working, aside perhaps from a movement of the throttle by the operator. By con-
trast, an open system is one such as a biological system (think of any living thing,
even a bacterium) that seeks to maintain itself in equilibrium by responding to its
environment.
Apart from features such as feedback and negative entropy, von Bertalanffy spoke
about features such as isomorphism (that is, being the same shape) between biologi-
cal and social systems. So, a beehive could be seen (from some points of view) as
being a bit like a town.
His thinking led naturally to interdisciplinarity, with insights from one field pro-
viding clues for thinking about how things work in another. In fact, von Bertalanffy
expected to see more of an integration of the work of science – natural and social.
Sadly, there is not a huge amount of evidence that this is occurring today, despite
attempts to promote it by all and sundry. He said that the human sciences were too
often atomistic, they were a kind of ‘social physics’ of the worst kind.
Beyond all of this, systems theory has been taken forward by others in some
interesting ways that may be quite useful to the case researcher. One of these is the
development of soft systems theory, expounded by Peter Checkland (1981) and his
colleagues at Lancaster University.
Checkland tried to incorporate the basic tenets of systems thinking while offer-
ing a framework for seeing the ways that the elements interconnect. He thus
advanced a model that was simpler than the one promoted by Barker (see page 58),
but still offered a framework of a more holistic understanding of social environ-
ments. Developing his ‘soft systems’ within the context of industrial psychology,
he suggested that the matters with which we are concerned are complex wholes
that maintain themselves in a changing environment. They do this by means of
adaptation and control action.
Although the concepts used in Checkland’s model are borrowed from biology
and engineering, he does not explicitly use the ecological metaphor, but his aims
are congruent with those of the ecological psychologists. He suggests a model that
draws a distinction between the ‘real world’ and ‘systems thinking’ and forces an
analysis of both. He asserts that the ‘soft, messy problems of human beings in their
everyday lives’ are not amenable to the methods of the ‘hard’ systems tradition of
engineering. He suggests that the hard systems tradition can handle natural (that is,
biological) systems and designed systems (such as bicycles, computers, mathemat-
ics), but cannot cope in situations dominated by human perceptions. He calls the
latter ‘human activity systems’.
His soft systems method suggests a sequence of events and processes for ana-
lysing social settings. I describe this in detail in Chapter 11, which is on forms
of analysis.
Ecological systems theory

I’ve already mentioned Bronfenbrenner (1979) in the context of ecological psychol-
ogy, but his work is also interesting as a fusion of ecological psychology and systems
thinking. Being a psychologist rather than a sociologist, Bronfenbrenner was not so
fond of theatres and stages as metaphors (with which sociologists seem fascinated,
if not mildly obsessed). No, rather than seeing human behaviour and activity as a
drama, he saw it as an ecology. The metaphor was biological rather than theatrical
(or ‘performative’, in sociological jargon).
Bronfenbrenner has perhaps contributed most systematically to further-
ing the ways in which the ecological approach can be used practically. It is he
who could also be said to have been the most assertive proponent of ecologi-
cal approaches, with some quite direct attacks on the more narrowly scientific
approaches to psychology of his contemporaries. For him, a parallel track was
not sufficient. He saw psychology’s traditional methods actually doing harm to
our understanding of people and he wanted to force a reappraisal of its tradi-
tions, habits and methods. It could almost be said that Bronfenbrenner was
aiming for a paradigm shift in the way that psychology proceeded as a form of
science.
For him, culture and society provide a set of ‘instructions’ for how social settings
are made and he conjoined all of his ideas about context into a formal framework
that he developed as a form of ‘systems theory’ (Bronfenbrenner and Morris, 1998).
He saw ecological environments as being composed of what he called micro-,
meso-, exo- and macrosystems. A microsystem is a pattern of activities, roles and
interpersonal relations of an individual in a given setting. A mesosystem is a system
of microsystems. An exosystem is a setting where the individual is not involved but

events occur that affect or are affected by the individual’s setting. A macrosystem
comprises the belief systems or ideology that structure other, lower systems (see
Figure 3.4).
So, taking an ecological and Bronfenbrenner-type view, if someone decides not to
get up and go to work today, how might we explain this? Looking at the situation
in just one-dimensional terms, we might frame it simply in terms of depression. She
is depressed. But does this explain very much? Looking more widely and deeply, we
might see her anomie in terms of lack of friendships. Looking even more widely, we
might find a problem that she has been having at work.
Culture Laws Ethics
Community Workplace
Connections between micro- and exosystem and

systems of microsystems
School Neighbours Family
Individual
Microsystem
Mesosystem
Exosystem
Macrosystem
Figure 3.4 Bronfenbrenner’s ecological systems
This kind of understanding is a million miles from traditional psychological

inquiry and is now encouraged by a new wave of clinical psychologists such as
David Smail. Indeed, in what sounds like an echo of Bronfenbrenner, Smail (1993:
13) asserts that: ‘There is certainly no evidence that the wider availability of psycho-
logical theories and techniques is leading to a decrease in psychological distress.’ It
is the oversimplifying models that are used to approach distress that are to blame,
says Smail.
Thinking particularly about children, Bronfenbrenner’s comment on traditional
psychological inquiry was that it sacrificed far too much to gain experimental control
and led to ‘the science of strange behaviour of children in strange situations with
strange adults for the briefest possible periods of time’ (1979: 19). The links with
symbolic interactionism are manifest. There is nothing new in the world.
Present day
Holistic views
Ecological
Dramaturgy systems theory
(Goffman) (Bronfenbrenner)
Systems 1950
Field theory thinking (von
Symbolic (Lewin) Bertalanffy)
interactionism Ecological psychology
(Blumer) (Barker, J.J. Gibson)
Interpretivism Gestalt psychology

(G.H. Mead) (Wertheimer, Köhler)
Phenomenology 1900
(Husserl)
Separating idiographic
Verstehen, or from nomothetic
understanding (Windelband)
(Dilthey)
Figure 3.5 Holism and its schools of thought

take this …
I hope you can see that the stress on the context and the ‘wholeness’ of behaviour
is many-stranded and is particularly suited to the case study. A number of points
came through in this chapter:
• Reductionist science has concentrated on breaking things down to the smallest

units, manipulating variables and looking for the effects of such manipulation.
• Holism, by contrast, assumes that the whole is more than the sum of the parts – that
psychological and social phenomena cannot be understood without an examina-
tion of the whole system.
• At the heart of the rationale for case study is the assumption that holism provides
an effective and fruitful way of addressing many of the questions that are posed in
social inquiry.
• Several streams of thought have emerged from a holist understanding: Gestalt
psychology, dramaturgy, ecological psychology, systems thinking. Throughout the
twentieth century such perspectives gained respectability as it was recognised
that they provided the rationale for highly productive methods of social and psy-
chological inquiry.
• Figure 3.5 summarises some of the schools of thought that have led to today’s
more eclectic attitude to inquiry. It is an attitude that accepts the relevance and
utility of the idiographic and provides ways of addressing wholeness by emphasis-
ing rather than denying the interconnectedness of the strands of psychological
and social life.
REFLECTIVE ACTIVITY
In the discussion of methodological holism, I have given as an example in this
chapter the insights taken by psychology from ecology. I noted that one of the
features of the ecological insight is that in a complex system nothing remains
constant: as one thing increases, others decline. How might this insight prove
useful in examining and analysing …
a. the behaviour of a child ‘acting out’ in class

b. the frequent arguments among patients waiting at reception in a health
centre
c. the resistance to lockdown measures among some members of the public
following the COVID-19 pandemic?
Further Reading
Barker, R.G. (1968) Ecological Psychology. Palo Alto, CA: Stanford University Press.
One of the most important pioneers in the movement to make psychology more
responsive to the whole, Barker insists that behaviour is always situated – in other
words, it does not exist in a vacuum. A useful platform for the case study.
Bateson, G. (1972/1999) Steps to an Ecology of Mind. Chicago: University of Chicago

Press.
A collection of Bateson’s essays and lectures. Really, his thesis is about ideas rather
than the ‘mind’ of the title, and the book is about what he calls ‘an ecology of
ideas’. It is rather mysterious and enigmatic, but good for thinking further about
the ways in which ideas interact – and we must make them interact in a case study.
Becker, H. (2014) What About Mozart? What About Murder? Reasoning from Cases.
Chicago: University of Chicago Press.
An excellent discussion of what case study is, and what it can do.
Burgess, E.W. (1927) ‘Statistics and case studies as methods of sociological research’,
Sociology and Social Research, 12: 103–20.
A classic article that presages many of today’s debates about the distinctions between
research involving small and large numbers.
Checkland, P. (1981) Systems Thinking, Systems Practice. Chichester: Wiley.

Checkland’s original thinking on soft systems.
Flyvbjerg, B. (2001) Making Social Science Matter: Why Social Inquiry Fails and How It
Can Succeed Again. Cambridge: Cambridge University Press.
See especially Chapter 6, ‘The power of example’, for a comparison of case study
with methods that seek to establish generalisation.
MacIntyre, A. (1985) After Virtue: A Study in Moral Theory. London: Duckworth.

Chapter 8 is the key one in this book: ‘The character of generalizations in social
science and their lack of predictive power’. This is taken to be a key critique of the
methods of social science and their attempt to establish generalisation.
Travers, M. (2001) Qualitative Research through Case Studies. London: Sage.

This is essentially about the interpretative tradition and the case study and has a
good chapter (Chapter 3) on Goffman’s dramaturgical analysis.
4
RIGOUR AND QUALITY
IN YOUR CASE STUDY:
WHAT IS IMPORTANT?
Rigour is important in research: judgements about rigour frame much of our assess-
ment of the quality of research. But what do we mean by ‘rigour’ when we are
talking about case study? And, given the flexibility in the form of case study, what
can ‘quality’ mean with this kind of research? If rigour is spoken about, I imagine
the first things people think about are accuracy, precision, thoroughness and strict
attention to detail. While these features of research are always important (for it
would be strange to accept inaccuracy or imprecision in any circumstances) they
are far from being the whole story when it comes to doing case study. And it may
be that some of the chief structural characteristics of social research – the charac-
teristics via which rigour is judged – may distract us or mislead us, for we may be
judging accuracy, for example, using the criteria by which we measure it in very
different kinds of research.
Let’s have a look at some of those structural characteristics of research – sample,
reliability, validity, quality, generalisability – and think about their meaning and rel-
evance in case study. My purpose in this chapter is also to take a look at features of
research such as theory and positionality, which are especially important in this kind
of research. Much has been written about these dimensions of research quality – some
elements of the methodological discussion being quite contradictory – so I intend to
take a critical look at this in the context of case study.
Is the ‘sample’ important in case study?

When thinking of the sample in research, the emphasis should, I believe, be on its
being a sample of something. It is a sample of a wider population (the total figure
RIGOUR AND QUALITY 67
of all the people in whom you might be interested is called the population, and the
sample is a slice or selection drawn from this population).
You might, for example, be interested in alcohol consumption among students
nationally. Clearly, it would be impracticable to issue a questionnaire to each and
every student in the country (the population) so you have to find a way of getting
a representative sample of the students to whom you are able to gain access. This
smaller sample from the whole population is intended to represent this population
and, clearly, the bigger the sample, the better the chance of representing it accurately.
You can, of course, imagine all of the things that can go wrong in sampling of
this kind. A sample from one university is unlikely to be representative of all uni-
versities; a sample from one group, such as your lecture class, even less likely; and
11 students in the bar on Tuesday night even more unlikely accurately to provide
a representative sample of the wider population – especially if you are interested in
students’ more usual drinking habits.
I said ‘the bigger the better’ for a sample, but there are ways of sacrificing size by
doing clever things such as ‘stratifying’, which means making sure that you get the
important parts of the population (by sex, age, type of institution attended, and so
on) properly represented in your small sample.
Now, none of this is relevant to a case study. The reason why I have devoted
several precious paragraphs to it is to try and hammer home the point that those
‘samples’ and all of the assumptions behind them are not necessary for a case study.
In fact, ‘sample’ is the wrong word for what is wanted in a case study. In textbooks,
to separate out the representative kind of sample (of which I have just given an
example concerning alcohol and students) from all others, the term ‘sample’ is
sometimes divided into probabilistic samples and non-probabilistic samples.
I personally think, however, that to make such a distinction is confusing. In a case
study, the choice that you make regarding your subject is nothing like a sample,
probabilistic or non-probabilistic. In fact, with any research, not just a case study, I
would go so far as to say that there is no such thing as a non-probabilistic sample –
or, at least, there might be, but we could just as easily and with less confusion call
it a ‘bad sample’ or, more generously, a ‘quick and dirty sample’. A sample has to be
‘a portion that shows the quality of the whole’,
in the dictionary definition. If it is not that – if
it does not accurately show the quality of the A sample should show the quality of
the whole. A ‘sample’ is not what is
whole to you as a researcher – it is at best a con- wanted in a case study. (Yes, ‘not’.)
venience, purposive or pilot sample.
The point of a case study is not to find a portion that shows the quality of the
whole. You are looking at your selection of subject – a marriage, country or what-
ever, with one, two or a few being focused on – without any expectation that it
represents a wider population. So, it is not a sample; it is a choice, a selection. It is
this selection that is vitally important for your study. In Chapter 6 I examine the way
in which you can make that selection, based on whether it is …
• a key case – a good example of something, a classic or exemplary case,

• an outlier case – showing something interesting because of its difference from the
norm, or
• a local knowledge case – an example from your own personal experience.
I give examples of these on page 111.
Do I have to worry about reliability and

validity in a case study?
As with many topics in research methodology, there is a variety of views on reliabil-
ity and validity, and the views are different for very good reasons. Those different
reasons usually concern the different ‘takes’ that people may have on what research
is about. I explore this in more detail in Chapter 5 of How to Do Your Research Project
(Thomas, 2017). Let me summarise my argument for advising you that reliability
and validity are not your principal concern when doing a case study.
Reliability is a concept imported into research methodology principally from
psychometrics – the branch of applied psychology that is about testing people’s
mental faculties. When giving tests, it is important that those tests are reliable as
instruments. It is important, in other words, that they give accurate data on differ-
ent occasions and in different circumstances, like a ruler or a weighing machine will
do. This is also true for certain kinds of applied social research, where, in some kinds
of inquiry, there are the same expectations of consistency of findings in different
circumstances. This applies only in certain kinds of inquiry, however. In a case study,
where there is one case, expectations about
reliability drop away. They drop away because,
The notions of reliability and validity
have been imported from other with just one case, there can be no assumption
kinds of research. Their meaning in from the outset that, if the inquiry were to be
the case study is far less clear.
repeated by different people at a different time,
similar findings would result.
Now, I should make it clear that this is my personal position and, while it is
shared by many others (such as Smith and Deemer, 2000) it is by no means shared
by all (Silverman, 2010). A common approach is to hold on to the notion of reliabil-
ity as a criterion for the assessment of all research, even research that could never
achieve (and would never want to achieve) findings that are consistent from one
time to another or one researcher to another.
Validity is similar to reliability when it comes to making judgements about its
value in case study research. Everything that I have said about reliability applies
also to validity. In many kinds of research – in those, for example, where probability
samples are used – validity is about the extent to which a piece of research is find-
ing out what the researcher intends it to find out. (It is not quite as simple as this,
and I tease apart some of the ways in which the word validity is used in How to Do
Your Research Project.) In the kind of research we are looking at here, however, where
there is no probability sample and we may have no idea at all about what we expect
to find out from the research, the idea of validity is less meaningful.
This is generally recognised in discussion about the quality of some kinds of
interpretative research, but, despite the concerns about the limited appropriateness
of validity, it is still felt by some people to be terribly important. There is almost a
cult of what Schwandt (1996) calls ‘criteriology’ – in other words, an obsession with
finding criteria. A treatise on methodology is felt to be naked without a discussion
of validity and criteria that can then be ticked off as having been (or not having
been) achieved. Validity is thus wrenched out of its home in normative research
with its samples, variables and statistics and bent and twisted into something quite
different for the purposes of interpretative research. The result is that we end up
bashing our square peg, validity, into the round hole of case study research.
For example, Whittemore et al. (2001), after having reviewed a range of authors
who have written about validity in interpretative research, emerge with a panoply
of criteria for assessing it. These include plausibility, relevance, credibility, com-
prehensiveness, significance, confirmability, positionality, canons of evidence,
generalisability, fittingness and auditability.
Now, I know that people read books like this one having got to the relevant sec-
tion via the index and do not contextualise what they have found. So, if this is you
and you are tempted to copy these words down (plausibility, relevance and so on)
and offer them in your methodology chapters as criteria for validity in case study
research, let me repeat that I do not agree with them. Validity is about that which
makes the research valid. It is not about plausibility, credibility and all the others.
Looking at the criteria with which Whittemore et al. emerge from their review, I
could say that some of the most fascinating and ground-breaking pieces of research
would have failed all of the tests of validity above. Let’s look at one – Einstein’s
work leading up to his theory of relativity. Would it have met many of the criteria
collated by Whittemore et al.? No. Let’s look at some of those criteria. With the
criteria for validity outlined in Table 4.1, an iconic piece of inquiry, key to the
development of late twentieth-century science, fails on nearly all counts. We have
to be wary of assessing work on the basis of these crude, post hoc constructions.
They are the tools of an audit society, not of good research.
I have deliberately chosen Einstein’s ground-breaking work in the natural sci-
ences, because it is a clear case of highly significant work – done, interestingly, via
‘thought experiments’ and case studies rather than the conventional experimenta-
tion of natural science – that conforms to none of the maxims of methodological
correctness. I could have turned in the other direction and chosen work that does
conform to the criteria of Whittemore et al., but which was … well, worthless. I
could quote work such as that of the educational psychologist Cyril Burt, who has
been shown to have fabricated his results (Hearnshaw, 1979: 245–7), but which, at
the time of publication, seemed to conform to all of the conventional measures of
validity.
Table 4.1 Did Einstein’s work have validity? Or 10 ways of not assessing validity
Criterion or  Has the criterion been met?

Plausibility  No. It was highly implausible (which was why no one else had
thought of it)
Credibility  No. It was entirely unbelievable that time slowed down as you
moved faster. In fact, it was incredible
Relevance  No. What possible use could it have had at the time? In fact,
applications were not realised until decades later
Comprehensiveness  How could it be comprehensive? It applied to one thing:
space-time. Why should it, or any piece of research, in the
natural or social sciences be comprehensive?
Significance  Only a few far-sighted souls could have seen its significance
at the time
Confirmability / Much of Einstein’s work was done through ‘thought
experiments’. His general conclusions could only be confirmed
in a piecemeal way
Canons of evidence  I’m not sure what ‘canons of evidence’ are. Einstein’s
breakthrough came from thinking about things – not from
evidence per se. The latter is as dead as a doornail without
intelligent thought
Generalisability  If correct (which it was), Einstein’s findings were generalisable
Fittingness ? What is ‘fittingness’? Why should it be relevant to validity?
Auditability  There was no possibility of ‘auditing’ mass–energy
equivalence at the time this breakthrough was made
In fact, Burt’s work amply meets the expectations of plausibility and credibility that
appear in the list from Whittemore et al. It is plausibility and credibility (the first
two criteria in Table 4.1) that are usually to the fore when validity is discussed in the
social sciences in the context of interpretative work. For me, however, these raise
particular problems as indices of any kind of validity, whether we are thinking in a
technical or even an everyday sense of the word.
Plausibility (in fact, surely much the same as credibility) seems to me a strange way
to assess validity. It is placing responsibility in the hearts and minds of the scholarly
community (you and me) to make that assessment of plausibility, but our views of
what is plausible will depend on the worldviews that have shaped our own thinking.
Our thoughts and ideas are slaves to what we already know and believe (well, mine
are anyway). The case of Burt is especially instructive because it is clear that he lied
and fabricated evidence to match with ideas about the heritability of intelligence,
which he ‘knew’ to be true. Here is the important point: his conclusions were ulti-
mately embraced because they were plausible. That is why they were accepted.
Researchers’ audiences (you and me again) – as judges of plausibility – cannot
be insulated from the prevailing academic, cultural and scientific orthodoxies
in making their judgements. I wonder what is supposed to inoculate us against
employing, in any assessment we make of plausibility, the very prejudices and
expectations that give shape to plausibility. It was this plausibility that made
Burt’s views acceptable to his intellectual peers. Plausibility was the culprit, not
the knight in shining armour.
That was a long way of saying that I don’t think you need to spend a great deal
of time worrying about validity and reliability in case study research as a form of
inquiry. This is partly also because the case study is not one single kind of research,
for it encompasses component methods. In certain of those component methods you
will certainly want to ensure that you pay attention to reliability and validity, but
remember that there is always a choice involved. You have to look at the constituent
bits of your case study and ask yourself whether or not reliability and validity, in fact,
are appropriate issues. Overall, though, the case study, as a study of one thing, is not
the kind of inquiry in which considerations about validity and reliability should be
to the fore since it is the singleness of the subject and the singleness – the peculiarity,
even – of the interpretation and analysis of the evidence that is significant.
Triangulation
The term ‘triangulation’ is used in a metaphorical way in the social sciences, based
on its original use in geometry and surveying. Triangles do not actually need to
be involved, nor do things need to be done in threes. What the term means here
is that viewing from several points is better than viewing from one. Given the
critical awareness that should be the hallmark of good social science researchers,
another viewpoint or another analytical method may make us decide to reject

initial explanations.
Of course, this should be what ‘case studiers’ are all about, looking in from
different angles and vantage points. I mentioned at the beginning of the book
Foucault’s ‘polyhedron of intelligibility’. By this, Foucault meant that we can only
really understand something – it only becomes intelligible – by looking at it from
different directions and using different methods. So, Foucault went one or per-
haps several steps further than calling for triangulation. We could (if we wanted to
invent yet another awful jargon word for the social sciences) say that he was calling
for ‘polyhedronation’. It is this that is the starting point for the case inquirer –
think small but drill deep, using different methods and drilling from different
directions. So, you may use observation in one part of the study, interviews in
another, a focus group in another, shadowing another person elsewhere, at the
same time as keeping a diary. This collation of methods provides the triangulation,
as shown in Figure 4.1.
Interviews
Observation
Case study
subject Focus group
Shadowing
Diaries
Figure 4.1 Methodological triangulation – looking at the subject using a variety of

methods
Research methodologist Norman Denzin (2017) would call this use of different
methods in research methodological triangulation. Denzin argued, though, that tri-
angulation goes beyond the use of different methods, and he outlined several other
types of triangulation, including investigator triangulation where more than one per-
son is involved in interpretation and analysis, and theory triangulation where more
than one theoretical framework might be involved in its interpretation. I would
add to Denzin’s categories design frame triangulation. In other words, you would be
triangulating if you used both a case study and a longitudinal study together in the
same piece of research.
I discuss triangulation further in How to Do Your Research Project (Thomas, 2017).
Positionality
Oftentimes in case study you will be immersed in the subject and the situation of
your research. You become an active, not passive, agent in acquiring knowledge of
the processes, histories, events, language and biographies of the research context.
You yourself will take a central role in interpretation. In doing this, it is important
to recognise that you have an undeniable position, and this position affects the
nature of the observations and the interpretations you make.
Recognition of this involves an inversion of some of the shibboleths of tradi-
tional methodology, such as the supposed imperative to ‘objectivity’. There is no
attempt to be ‘objective’ when working interpretatively, but rather an acceptance
of the importance of researchers themselves – their likes and dislikes, their back-
grounds and their pastimes, their vested interests and expectations – in making
interpretations. It is subjectivity, in other words, rather than objectivity, that comes
to the fore. Because of the importance of the nature of the relation between the
researcher and research participants, the researcher’s biography – including class,
gender, ethnicity, ideas and commitments – needs to be made explicit.
In presenting an interpretative case study you should accept your subjectivity
and not be ashamed or afraid of it. Given that it is central, your case study should
be written up in a way that makes clear your understanding of the underpinning
principles that guide the conduct of this kind of research. An interpretative case
study will therefore be reported in an entirely different way from an experimental
study. You should begin – right at the beginning, in the introduction – with a full
discussion of positionality: of yourself, why you are interested in this topic, what
your personal circumstances are, and so on. You will always write in the first person,
saying, for example, ‘I believe that …’ rather than ‘The researcher believes that …’.
This may seem obvious, but it is a common mistake for students conducting inter-
pretative research – research that demands that they put themselves at the centre of
the analysis – to report their work as if they have just conducted an experiment and
to omit any information about themselves. If you do this kind of research, readers
need to know who you are and where you stand, metaphorically as well as literally.
Generalisation
The case study has often been taken to be deficient in the generalisation department,
as I discussed in Chapter 1. But its ‘poor relation’ status exists, I think, only because
it is conspicuously deficient in its potential for generalisation. Its weaknesses in terms
of generalisation, in other words, are not disguised. Other design frames in social
science seek ways to calibrate and enable generalisation, but these attempts rest on
something of a sleight of hand in the presentation of what generalisation can be.
If, as many have argued (and I shall rehearse the classic case in a moment), little
significant generalisation at all is possible in the human sciences, arguments about
the case study’s deficiencies in this department lose force. My point is that, when
thinking about inference, generalisation in social inquiry is constrained to remain
at the level of what Peirce (Houser et al., 1992) calls ‘abduction’ rather than induc-
tion. I shall explain what abduction is in a moment.
If my argument has validity, then there are forms of interpretation that can
come from case studies which owe their legitimacy and power to the exemplary
knowledge of these studies rather than to their generalisability. For me, seeking
generalisability – seeing generalisation as the first and most important aim of social
science – can inhibit or even extinguish the curiosity and interpretation that should
be at the heart of inquiry.
Generalisation, induction and abduction

Generalisation is at the core of the natural scientific method. From generalisation
comes induction – that is, X happens regularly in certain conditions, so we can
infer that X will happen again in those conditions. The process is attempted across
the social sciences and is seen across the methodological spectrum.
There are differences between the natural and social sciences, though, that
are easily elided when one is thinking about the value of generalisation. The
establishment of regularities, generalisations, laws, universals and theory in the
induction of natural science is governed by exacting expectations and procedures
that enable investigators to deal with exceptions, anomalies and idiosyncrasies
and establish their credentials and significance as limiters to the power of the
generalisation.
The problem is that, outside natural science, such expectations are unrealisable.
Alasdair MacIntyre’s (1985) argument here is the classic one – comparing generali-
sations in the social sciences with those in the natural sciences and noting the role
that counter-examples to the development of law play in each. The attitude in the
social sciences, MacIntyre (1985: 89) says, is ‘tolerant’ of counter-examples. Social
science’s generalisations – all of them, not just those in case studies – are pretty use-
less, MacIntyre argues, for ‘we cannot say of them in any precise way under what
conditions they hold’ (1985: 91).
He proceeds to argue that social scientists have insisted that what is offered in social
science is probabilistic generalisation. Nor is this any good, says MacIntyre (1985: 91),
for ‘if the type of generalisation which I have cited is to be a generalisation at all, it
must be something more than a mere list of instances’. According to MacIntyre, then,
generalisation is possible in the natural sciences, but its status in the social sciences
can never be any better than that of the generalisation of a layperson. Its value will
always be limited by the sheer variability of social life and human agency in all of its
unpredictability.
The problem here is with expectations of watertight (or almost watertight) induc-
tion following generalisation. Perhaps, though, following Peirce (Houser et al.,
1992), our expectations should be moderated concerning the generalisations that
can emerge from social studies. A number of commentators have queried the reli-
ance on a naïve model of scientific induction (for example, Haig, 1995) and others,
such as Miller and Fredericks (1999), point to the lack of ‘specialness’ of induc-
tive reasoning. The thrust of their argument is that common, everyday induction
is better described as ‘inference to the best explanation’, or Peirce’s ‘abduction’.
Abduction is making a judgement concerning the best explanation for the facts you
are collecting. It is what we do in case studies (Thomas, 2010).
Hammersley (2005: 5) suggests that ‘what is good evidence for abduction is differ-
ent from what is good evidence for induction’, proceeding to describe abduction as
‘the development of an explanatory or theoretical idea, this often resulting from close
examination of particular cases’ (see also Hammersley, 2007). It could, perhaps sim-
plistically, be thought of as ‘conclusions drawn from everyday generalisation’, whereas
induction concerns conclusions drawn from a special kind of generalisation – a kind,
if MacIntyre is right, that is impossible in the social sciences.
Thinking about all of this, it seems to me that any argument about the weak-
ness of the case study resting on its lack of generalisability fails to recognise the
limits of induction in the social sciences generally. It also fails, simultaneously,
to acknowledge the significance of abduction. It fails, in other words, to rec-
ognise the offer that can be made in local circumstances by particular kinds of
looser generalisation, whatever one calls these. I have chosen to use Peirce’s term
‘abduction’ for these to foreground the distinctions to be made with induction in
discussing the supposed weaknesses of the case study, but they have been framed
in many other ways – from the ‘common interpretive act’ (Schatzman, 1991),
to ‘bounded rationality’ (Simon, 1983), Bourdieu’s ‘thinking tool’ (in Wacquant,
1989: 50) or Althusser’s Generalities II (Althusser, 1979: 183–90) – all of which
describe much the same kind of process. That process is a fluid understanding
explicitly or tacitly recognising the complexity and frailty of the generalisations
we can make about human interrelationships.
They describe our processes of garnering and organising information to ana-
lyse and deal with our social worlds. Abduction connects all of these, providing
heuristics – that is, ways to analyse complexity that may not provide watertight
guarantees of success in providing for explanation or prediction, but are unpreten-

tious in their assumptions of fallibility and provisionality.
All of these forms of generalising to regularity – these kinds of abduction – seem
to be the appropriate ‘inference form’ for the case study.
Retrospective generalisation and cumulation

The great educator Lawrence Stenhouse crystallised many of the points I have made
about the tentativeness of generalisations in the human sciences when he sug-
gested that the most that could reasonably be expected of generalisation was a
‘gradient of expectation’:
Cases cumulate, as for example in history we cumulate cases of long-

barrow burials or cathedrals or joint-stock companies, but the retrospec-
tive generalizations derived from surveying the cumulated cases do not
provide predictive generalizations in the classic form: they define rather
a gradient of expectation. (Stenhouse, 1979: 6)
Note here that Stenhouse draws upon the idea of cumulation – we draw conclusions
based upon the cumulated evidence available to us … and we surely do this on the
basis of evidence in many and varied forms.
For Stenhouse, then, we build knowledge from the collation of different forms of
case evidence. We build from case evidence an archive of knowledge from what Berger
and Luckmann (1979: 21) have famously called ‘multiple realities’. Researchers build
exemplary knowledge, making connections between another’s experience and their
own, seeing links, having insights from the noticed connections. We create what
Abbott (1992: 12) calls ‘colligations of occurrences’. In discussing the logic of this,
Mitchell (2006) concludes that a good deal of the confusion about case study has arisen
because of a failure to appreciate that the rationale of extrapolation from a statistical
sample to a parent universe involves two different and even unconnected inferential
processes: (i) statistical inference (that which is used in variable-led research) and (ii)
logical or scientific inference, the latter being about the confidence we may have that
the logical connection among the features observed in the case relates also to the par-
ent population. It is this latter that we draw upon in case-based research.
Theory
Developing or testing theory can be thought of as being at the centre of case study,
and I examine how you can develop theory in Chapter 11. Here, though (and in
the context of rigour and quality), I want to take a sideways, critical look at theory
and what it is.
In the traditional account of the process of science, the end point of the inductive
process results in theory. Seen in the way that Peirce presents it above, abduction
generates ideas, tentative ‘theories’, which serve as hypothetical explanatory con-
cepts. This abduction is then followed by deduction (the connecting of ideas),
which is followed in turn by systematic data collection in pursuit of verification of
the initial explanatory concepts. This last is induction, via which the speculative
theory is or is not verified.
The assumption that this is the course things invariably take seems to me
to lead always to the dead-end identified by MacIntyre. In other words, the
assumption of theoretical validation from inductive inference necessarily
implies a firmness to the structure and reliability of theory that is unachievable
in the inquiries of social science. Because of both the contingency, the way-
wardness, of social life and the necessary limitations on the kind and quantity
of confirmatory evidence that can be disclosed, theory, in any kind of technical
sense, is surely unattainable.
There is an elision of expectations here in the discussion of theory. As I have
noted elsewhere (Thomas, 2007; Thomas and James, 2006), a reading of interpreta-
tive inquiry scholarship from Glaser and Strauss (1967: 1) onwards (Mouzelis, 1995)
reveals an insistent theme in the expectations of theoretical explanations and pre-
dictions from induction.
Of course, something is attainable, but the question is, ‘Is it really theory?’ It can
be theory only in the sense that everything is theory. Fish (1989: 14–15) suggests that
much discourse about theory in the academy generally is not really about theory
at all, but, rather, ‘theory-talk’ – that is, ‘any form of talk that has acquired cachet
and prestige’. He also notes:
Am I following or enacting a theory when I stop for a red light, or use

my American Express card, or rise to speak at a conference? Are you now
furiously theorizing as you sit reading what I have to say? And if you
are persuaded by me to alter your understanding of what is and is not a
theory, is your new definition of theory a new theory of theory? Clearly
it is possible to answer yes to all these questions, but just as clearly that
answer will render the notion ‘theory’ and the issue of its consequences
trivial. (1989: 327)
In other words, ‘theory’ has come to mean almost anything – any generalisation,
any thought, any structured reflection (or, indeed, any unstructured reflection) may
be called a theory. It is unhelpful in our deliberations on inquiry, however, for

looseness to exist about the terms we use – in this case, ‘theory’ – particularly when
separating abduction from induction and looking to the end point of induction.
The process of induction is supposed to lead us to a theory of a particular
character – a theory that summarises and generalises in such a way that we can
use it to explain and predict. It confuses our deliberations on inquiry to call any
kind of summative or generalising process ‘theory’. When a theory has been
used as part of the reasoning to validate a case study (Eisenhart, 2009; Vaughan,
1992; Walton, 1992), it has indeed been as part of this summative, inductive or
quasi-inductive, process.
It is worth returning to MacIntyre’s analysis here, since it speaks to the central
problem with generalisation and theory in the social sciences. He notes that there
are a number of reasons for the failure of theory (that is to say, theory as part of
a quasi-inductive process), all resting on the ‘systematic unpredictability’ of the
social world. Because of these, he suggests, unpredictability will always win over
predictability as far as the matters of social science are concerned. It is only the
trivial things that are predictable, and we do not need any kind of sophisticated
methodology to tell us about these.
Theory and phronesis

If this critique has any legitimacy, and it surely does, we have to make a distinction,
in response, between theory and phronesis.
Aristotle’s notion of phronesis is about practical knowledge, craft knowledge, with
a twist of judgement squeezed into the mix. As it has been used more recently –
particularly in the discussion of the applied social sciences (for example, Back,
2002) – it has come to have more of a sense of ‘tacit knowledge’ (Polanyi, 1958)
about it.
Phronesis is practical knowledge. It is a model based on personal experience. It is
personal, and it helps us to make sense of particular situations. Sometimes, confus-
ingly, it is called ‘practical theory’. It is judgement made on the basis of experience
and has no pretensions to lead us to the kind of external guide that theory is sup-
posed to provide. It is in practice that phronesis is developed and in practice that
it comes into play.
Back goes on to provide a contrast between phronesis and theory, explaining
that, in Aristotelian terms, ‘theory’ exists as a way of establishing absolute laws
that can be laid out in an organised framework which can explain. It must be
able to withstand tests as to its validity and it needs to be consistent. By contrast,
in the practical (or tacit) knowledge of phronesis, there are none of these expec-
tations as to consistency – phronesis is about understanding and behaviour in
particular situations.
Fish’s (1989: 317) explication of practical learning analyses the contrast well,
for he explains how practical learning varies with the context of a practice – that
is, as circumstances change, so ‘the very meaning of the rule (the instruc-
tions it is understood to give) changes too’. The phronesis that we acquire and
accumulate is always malleable and corrigible. So there is, in talking about
phronesis, a recognition of the provisional, the tentative in interpretation and
analysis.
There is, in other words, a need to move away from the expectations of generalis-
able knowledge that go alongside the inductive process. Correspondingly, there is a
need a move towards the ‘exemplary knowledge’ of abduction and phronesis. With
the discussion of ‘exemplary knowledge’, I am talking about an example viewed
and heard in the context of another’s experience, but used in the context of one’s
own. The example is taken to be neither representative nor typical, nor is it exem-
plary in the sense of being a model to follow. Rather, it is taken to be a particular
representation given in context and understood in that context. It is interpretable,
however, only in the context of one’s own experience – in the context, in other
words, of one’s phronesis, rather than theory.
If the case study is concerned with phronesis rather than theory – if it is decou-
pled from the inductive frame of theoretical analysis – what are the consequences?
We are left with a view of a case study’s validation coming no longer from refer-
ence to a body of theory or generalised knowledge. Rather, its validation comes
from the connections and insights it offers between another’s experience and your
own. The essence comes in understandability emerging from phronesis – in other
words, from the connection to your own situation.
This is based in what Abbott (1992) calls colligations of occurrences, involving
the making of narrative – the development of stories with which one can connect.
This is a characteristic, he notes, of all social science, intelligently done.
Finding or regularising
Having had this discussion about induction versus abduction and theory ver-
sus phronesis, I think it is worth noting again that this discussion is not new. I
outlined in Chapter 3 the difference between Plato and Aristotle on the desire
(or not) to establish laws. That debate influences even now our attitude to the
ways in which inquiries should be conducted. You could say that the contrast is
between those who could be called the finders and those who could be called the
regularisers.
Another way of putting it would be that it is an intellectual battle between
the narrators and the theorists, and it exists not just in the social sciences. The
great historian R.G. Collingwood (1946/1994: 30), in The Idea of History, com-
pared the approaches to history of Herodotus and Thucydides and concluded that
Herodotus contributed so much more, because ‘what chiefly interests Herodotus
is the events themselves; what chiefly interests Thucydides is the laws according
to which they happen’. Thucydides’ problem, Collingwood (1946/1994: 31) sug-
gests, is that he ‘is constantly being drawn away from the events to some lesson
that lurks behind them’. Collingwood favours Herodotus’s critical questioning
of narrative as a way of doing history. He notes (1946/1994: 29): ‘In reading
Thucydides I ask myself, What is the matter with the man, that he writes like
that?’ He answers (1946/1994: 29) that Thucydides abandoned history for a kind
of pseudo-psychology: ‘It is not history at all, but natural science of a special
kind. It does not narrate facts for the sake of narrating facts. Its chief purpose is
to affirm laws.’
Collingwood seems to be speaking straight to those undertaking case studies
here since he seems to be saying ‘be more like Herodotus’. Herodotus asks us to
look with him at the facts. He may tell us what he believes, but he also invites
readers to make their own judgements. The lawyer Thomas Geoghegan (2007)
puts it this way: ‘we should also spend more funds to get our young people out
of the library where they’re reading Thucydides and get them to start living like
Herodotus – going out and seeing the world’. Again, a motto for the case inquirer.
Quality
I have used the term ‘structural characteristics of research’ for many of the concepts
I have discussed in this chapter, and, as I hope I have indicated, these concepts
(reliability, validity, generalisation and so on) take on different meanings in case
study from those they might assume in other kinds of research. It is difficult, in
other words, to judge the quality of a case study via the criteria that are often used
in judging other kinds of research.
There are characteristics of research – though very general ones – by which we
can unequivocally judge quality. Assuming for a moment that we are concerned
principally with the interpretative aspects of a case study, Hammersley (2005:
3–4) has offered some useful indicators as to this quality, which I have adapted
as follows:
• The clarity of the writing

� Is there consistency in use of terms?
� Are definitions provided where necessary?
� Are sentences well constructed?
• The problem or question being addressed
� Is this clearly outlined?
� Is sufficient rationale provided for its significance?
• The methods used

� To what degree, and in what respects, was each of the methods chosen (as
regards selection of cases for study, data collection and data analysis) likely to
be an effective one?
• The account of the research process and the researcher

� Is there sufficient, and not too much, information about the research process?
� Is there sufficient, and not too much, information about the researcher?
• The formulation of the main claims

� Are the main claims made clear?
� Are the relations between claims and evidence made clear?
� Is the nature of each claim (as description, explanation, theory or evaluation)
indicated?
Hammersley (2005: 6), however, proceeds to give a health warning about the criteria
he has offered, and this chimes with the ‘criteriology’ warning that Schwandt (1996)
issued. Hammersley (2005, emphasis in original) says of the points that he makes
about quality:
these are lists of what we might call issues for threshold judgment: they
remind the reader what to take account of, what they need to make a
judgment about. However, they are not operationalised in such a way
that anyone who did not already know how to make the judgments
concerned would be able to apply them, nor do they tell the reader what
would be sufficient in each case. And they do not tell us whether all of
the thresholds have to be met for a positive conclusion to be reached,
whether high scoring on one can counterbalance a lower ‘score’ on
another, and so on. Moreover, I do not believe that they can be opera-
tionalised in this way, or that attempting to do this would be desirable.
Put differently, how could you know that you have met these criteria? How could
readers apply such criteria if they were not aware of them implicitly in the first
place? Further, it would be impossible for researchers or readers to weigh up the
importance of one criterion versus another if they were not already steeped in the
discourse of the field of study.
For me, the issues boil down to the choice of your case, the choices made con-
cerning analysis and the robustness of your argument. In other words, your answers
to the following questions:
1. How well has the case been chosen? Decisions should have been made appropriately
about the approach and the processes to be adopted in data collection and analysis.
2. How well has the context for the study been explained and justified? Including
justification for decisions about how analysis has been undertaken.
3. How well have arguments been made? Have rival explanations for the same
kind of observation been explored? This is where the storyboards that I have
drawn for some of the examples in this book can help – providing space to think
about alternative storylines for the same plot. Is it justifiable that, on this evi-
dence and with this reasoning, these conclusions have been drawn?

take this …
The quality of a case study depends less on ideas of sample, validity and reliability
and more on the conception, construction and conduct of the study. It depends on
• your initial idea

• the ways that you choose your case
• the thoroughness with which you describe its context
• the care you devote to selecting appropriate methods of analysis
• the nature of the arguments you deploy in drawing your conclusions.
The case study offers the opportunity to bring evidence together from many and
varied sources to support arguments in ways that would not be possible using
other forms of inquiry that are fenced in by different considerations. Concerns
about how far we can generalise from a case study are neutralised when we real-
ise how tentative any generalisation might be in social research. Conclusions
drawn from case study research become less pronounced when we realise that, to
a greater or lesser extent, all forms of inquiry, especially social inquiry, produce
knowledge that is provisional – in other words, good only until we find out some-
thing else which explains things better.
REFLECTIVE ACTIVITY
Quality in social research is assessed differently depending on the kind of
research being undertaken. Quality is sometimes measured using particular cri-
teria. Reliability and positionality are two of these criteria. In this chapter I have
discussed the use of these criteria and how their importance varies depending on
the kind of research being undertaken. In variable-led research, such as experi-
ment, certain criteria are more important than they are in case study research.
In case study research other criteria are more important. With these considera-
tions in mind, rate the importance of reliability and positionality in variable-led
research and in case study research in the table below, stating your reasons for
the rating you have given.
Rate the importance of reliability and Why do you give your rating?
positionality in variable-led research and in
case study (5, very important; 1 not important
at all)
Variable-led research 54321
Reliability
Case study 54321
Variable-led research 54321

Positionality
Case study 54321
Further Reading
ESRC/TLRP Seminar Series: Quality in Educational Research (available at: www.
education.bham.ac.uk/research/seminars1/esrc_4/index.shtml).
This is a series of seminar inputs on quality in educational research, giving tran-
scripts and articles from the series. The seminars cover issues to do with quality of
evidence and how quality differs in quantitative and qualitative research.
Geertz, C. (1977) ‘Deep play: notes on the Balinese cockfight’, in C. Geertz, The
Interpretation of Cultures. New York: Basic Books (available at: www.si.umich.
edu/~rfrost/courses/Matcult/content/Geertz.pdf).
A classic case study that reads more like a story than a piece of ethnography, and
this is a tribute to the quality of Geertz’s writing.
Matthews, B. and Ross, L. (2010) Research Methods: A Practical Guide for the Social
Sciences. Harlow: Pearson Education.
Matthews and Ross have an excellent section on sampling, and go into some very
useful detail also on other aspects of methodology.
Ragin, C.C. and Becker, H.S. (eds) (1992) What Is a Case? Exploring the Foundations
of Social Inquiry. Cambridge: Cambridge University Press.
Schwandt, T.A. (1996) ‘Farewell to criteriology’, Qualitative Inquiry, 2: 58–72.

The authors critique the idea that quality can be assessed by reference to a series of
criteria.
Thomas, G. (2010) ‘Doing case study: abduction not induction; phronesis not
theory’, Qualitative Inquiry, 16 (7): 575–82.
Thomas, G. (2011) ‘The case: generalisation, theory and phronesis in case study’,
Oxford Review of Education, 37 (1): 21–35.
Thomas, G. (2017) How to Do Your Research Project (3rd edn). London: Sage.
In Chapter 6, I go into more detail about issues such as reliability, validity and
samples.
Tobin, J. (2007) ‘An anthropologist’s reflections on defining quality in education

research’, International Journal of Research and Method in Education, 30 (3): 325–38.
Good on the problems of defining quality in social research.
Winter, G. (2000) A comparative discussion of the notion of ‘validity’ in qualitative

and quantitative research. Qualitative Report, 4 (3–4) (available at: www.nova.
edu/ssss/QR/QR43/winter.html).
Winter provides a full and interesting discussion of validity.
5
ETHICS
Ethics are principles of conduct about what is right and what is wrong. But decid-
ing on what is right or wrong can be a complex business in social research. What is
right for me may not be right for you, and what is right for the researcher may not
be right for the participant. The consideration of ethics may encourage us to look at
the nature of what we are doing in the name of furthering knowledge.
Obviously, we all want to do what is right
and avoid what is wrong. Although the mat- It is especially important to consider
ethics in case study research since
ter seems simple, there are many examples in you may be very closely involved
social research where a questionable action may with the research participants.
have been taken in the name of doing right.
Researchers have had as their purpose the idea that they are pushing back the frontiers
of knowledge or helping humankind in some other way and have used this as a reason
for employing intrusive or distressing procedures in their research. Several high-profile
examples, such as the ‘Milgram experiments’, where volunteers were instructed to give
stooges what they were told were life-threatening electric shocks, have caused decades
of debate about the rights and the wrongs of such studies (Milgram, 1963, 1981).
You are most unlikely to be contemplating anything as dramatic as the Milgram
experiments. However, there may be ethical issues of which you were not aware or
which you had not considered. Many of these will present questions of the kind
that confront us now looking back at Milgram, albeit on a smaller scale. Who is the
research benefiting? Do you have the right to take up people’s time and energy? Is
there any possible discomfort that participants will have to experience? Are you
invading their privacy?
Your participants
Your case study may not involve your dealing with people. It may be a paper-based
study, looking at one historical case. However, in most of the studies in which
we become involved we will be working with people. In the bad old days, when
social scientists did research with people, they called those people ‘subjects’. Now,
especially in applied social science, we think of the people with whom we research
more as ‘participants’ or even ‘partners’, rather than ‘subjects’. Thinking with ethics
to the fore, we have to see the people with whom we are involved as participants
in our research and part of it, rather than simply objects from whom data can be
extracted.
In working with our research participants, we have to recognise that these partici-
pants have rights and they should have a stake in the process of research: it should
not be a question of simply ‘using’ people and then walking away. To be genuine
participants they have to be involved to some extent in the planning, execution
and write-up of your work. You will need to think about their particular kinds of
contribution and their particular needs. Especially, you will need to think about the
needs of people who are vulnerable for any reason.
Vulnerable groups
For those who may not understand the ins and outs of consent or who may be sus-
ceptible to pressure to cooperate because of their social or economic position there
will need to be special considerations. These will include:
• children/legal minors (anyone under 16)

• people from non-English-speaking backgrounds
• anyone with a physical disability
• anyone with learning difficulties
• patients, or clients of professionals
• anyone in custody, or for whom a court has responsibility.
Some people are especially prone to become drawn into research projects because of
their ‘ready availability in settings where research is conducted’, as the US govern-
ment’s Belmont Report (1979) put it. Careful thought should be given to questions
about why and how they are being involved,
Children, patients, employees and
and they should be protected against involve-
people with special needs are all
vulnerable and you will need to ment solely for the sake of convenience, or
consider carefully the ethics of because they may be easy to persuade or even
working with them.
manipulate.
ETHICS 87
Lewis and Porter (2004) urge that the following questions should always be asked with
any kind of vulnerable group, including children and those with learning difficulties:
• Has the participant’s ability to give fully informed consent been assessed and dis-
cussed with others such as parents, caregivers or teachers?
• Have ways of checking for understanding of confidentiality/research purposes
been explored?
• Will participants, at appropriate intervals, be reminded of their right to withdraw?
• Have all possible steps been taken to ensure anonymity, given that this may be
particularly difficult to achieve with minority populations?
• In giving feedback, have steps been taken to ensure the intelligibility of the infor-
mation? This can be done through, for example, asking a person who is well known
to the individuals to talk with them, or offering pictures with simplified text or case
study material.
• How is the end of the research relationship with participants to be managed? It is
easy for close relationships to be forged during research, and these should not be
terminated abruptly.
Participants’ activity
When relating with your participants – and we are now thinking about all par-
ticipants, rather than specifically those who may be vulnerable in some way – you
should think about issues such as:
• Administration of any questions or procedures which may cause mental or physi-

cal discomfort during or after the research – these may include questions which
appear at first sight to be quite straightforward. Always think about the effect that
questions may have on your participants.
• Performance of any acts (such as role-play) which might diminish self-esteem or
cause embarrassment. The more extravert among us sometimes, I think, fail to
appreciate how mortifying and disabling procedures such as role-play can be to
those of us who are more introverted. Not only are such procedures questionable
on ethical grounds, but they also may give misleading findings.
• Involvement of participants in any illegal activity. Definitely a no-no.
• Whether the participants will be in receipt of any substance or agent – this is likely
only in health-related research and there will be very strict professional codes of
conduct on issues such as this.
Deception or concealment
The default position on doing research is that you should be honest and open
in all your dealings with research participants. Sometimes, though, as in the
Milgram experiment I mentioned earlier, you will need to be less than 100 per
cent open because you may be researching something where the participants’
knowledge of your aims would invalidate any findings that you may make. If
this is the case, you must be prepared to argue convincingly for the need for
concealment, and you must build into your plans a debriefing session with your
participants after the research has finished wherein you explain what you were
doing in your study and why you could not be fully open at the outset.
The codes of conduct from professional asso-
ciations usually recognise that an element of
Always put respect for your
participants centre-stage. As far as concealment or even deception may sometimes
possible be open about your aims, be necessary. Indeed, the US government’s
and avoid subjecting people to any
kind of discomfort. landmark Belmont Report (1979) recognises it,
but says that:
In all cases of research involving incomplete disclosure, such research is

justified only if it is clear that (1) incomplete disclosure is truly neces-
sary to accomplish the goals of the research, (2) there are no undisclosed
risks to subjects that are more than minimal, and (3) there is an ade-
quate plan for debriefing subjects.
So, if any degree of withholding or misrepresentation is involved, it will be import-

ant for you in gaining ethical clearance to:
• acknowledge explicitly what is being done

• justify fully the reasons why it is being done
• note that you will be explaining to participants following the research the purpose
of the research and why it was necessary to withhold information from them.
Confidentiality and anonymity

You should always treat any information provided to you as confidential, taking
care at all times not to breach or compromise that confidentiality. Maintaining the
anonymity of your participants is a key part of this, in your everyday dealings and
ETHICS 89
conversations with others, in your storage of data and your reporting. Anonymity
can be ensured by changing participants’ names as well as the name of any institu-
tions (such as schools) to which they are affiliated and the regions in which they
are situated. You can either give pseudonyms or code numbers to achieve this. You
will need to be especially careful in working with participants whose identity may
be hard to disguise, for example those in minority groups. Where appropriate, you
should make it clear to participants that your commitment to confidentiality as a
researcher may be overridden given your legal or moral duty to report incidents of
harm. Circumstances such as these will be most unusual, but if you do come across
them you should discuss them with your supervisor and/or the appropriate services
(such as child protection services) immediately.
Confidentiality is not secured simply by paying attention to anonymisa-
tion, important though the latter is. More broadly, it is about a commitment
to respecting the wishes of informants, who may not want the information
which they have given you to become public. While engagement in research
should involve, of course, an understanding that findings will be published,
your informants may expect that you will disguise or hide information which
they take to be confidential – perhaps using it in drawing your conclusions, but
keeping the details off the record.
You should always be sensitive to this possibility and even proactive in asking
about it, moving on to negotiate with your research participants how much – and
how – you intend to use, analyse and disclose your findings. They may ultimately
decide that they do not wish you to publish certain facts or opinions, and this is
a wish you must respect. In cases where it is unclear whether you have secreted or
masked data to the satisfaction of your informants you will need to give them the
chance to read through your final report to allow them to make suggestions about
how to edit it. You can find an excellent discussion of all these issues in Wiles
et al. (2008).
Occasionally, research participants may ask
As a default position, all data should
to be identified and not anonymised in research be treated as confidential and names
outputs (see Grinyer, 2002). You should respect coded or otherwise anonymised.
Data should be kept secure and
any such requests while ensuring that you dis- deleted after analysis.
cuss fully with all those involved, including
caregivers, the potential consequences of disclosure. And remember that if one
person’s name is revealed, this increases the risk of a breach of confidentiality
for other participants who wish to remain anonymous. Having carefully weighed
up all the pros and cons, you should, if you decide to agree to any request to
disclosure, obtain written consent that an individual wishes to waive the right to
confidentiality.
Data security and stewardship

You have a responsibility to keep the data you collect about people or institu-
tions secure, even though you will have taken every effort to anonymise that data.
Various exemptions exist in the UK Data Protection Act 2018 to allow researchers
to collect data without the need to inform the Data Commissioner, but this does
not mean that you can ignore principles of good stewardship of data. You should:
• Only use data for the purposes for which the data were collected – not for other
purposes.
• Keep data for an appropriate length of time. What counts as ‘appropriate’ var-
ies according to the circumstances: for most undergraduate research it will
be appropriate to destroy raw data immediately your project is completed and
marked; however, for some research (for example, that which is part of a clinical
programme, which a postgraduate project may be part of), professional codes of
conduct specify that data must be kept for several years following the close of a
programme. It is worth noting that the UK Information Commissioner’s Office (ICO)
says that, with justification, data can be stored for research indefinitely.
• Keep data secure. You should use passwords on any files that contain pre-
anonymised names. For Microsoft Office documents, this means clicking the Office
button (top left of the screen), clicking Prepare, then clicking Encrypt document.
(If this does not work, click the F1 button for help, and enter ‘password’.) You will be
asked for a password, which you must remember. If you do not remember it, there
is no way to get back your data.
• Do not pass the data on to anyone else – obviously, since the data would not then
be secure.
• Keep the data anonymous (see previous section).
If you are in any doubt about any of this, you should search for the most recent
advice from the ICO on the EU General Data Protection Regulation (GDPR) (2016)
and the UK Data Protection Act 2018, which govern the processing (acquiring,
holding, using, etc.) of personal data in the UK.
Consent
Because of the harm that might be caused during research involving people, an
important issue to think about when considering the ethics of research is that of
consent. Consent is about the agreement of people to take part in the study. More
ETHICS 91
than simple agreement, however, is needed, given the issues I have already con-
sidered. Informed consent is needed. In other words, potential participants should
understand what they are agreeing to. Informed consent is centred around the fol-
lowing points:
(a) The information participants need to know; this will include:

• the nature and purpose of the study, including its methods
• expected benefits of the study
• possible harm that may come from the study
• information about confidentiality, anonymity, how data will be kept and for
how long, with details of when data will be destroyed
• ethics procedures being followed and appeals
• your full name and full contact details
(b) The presentation of (a) in a meaningful and understandable way, explaining any
unusual terms simply, in non-technical language
(c) The option for a potential participant to choose to take part or not.
Opt-in versus implied consent

An important distinction exists between opt-in consent and implied consent:
• With opt-in consent, participants have to make an active choice about becoming
involved and signal their willingness to take part in the research. You would have
to offer an invitation to participants to become involved, preferably in writing,
requesting that they return a form indicating willingness to participate. This could
be done directly and orally if written communication is not possible. Alternatively,
some other active choice of the participant would have to be involved, as would be
the case in returning a questionnaire. It must be made clear that the participant’s
consent may be withdrawn at any time and that if the participant chooses to do
this there will be no further collection of additional data, no further analysis of
data initially collected and removal of existing data from your records.
• With implied (or opt-out) consent, you tell participants about the research and
assume that they give their consent unless they tell you otherwise. Information
about the research can be distributed by a range of means, for example by a letter
from a school to the parents or guardians of all children who would be personally
involved. The letter would include the information you would give as for opt-in
consent, but would be accompanied by a phrase such as ‘I am assuming that unless
I hear from you, you have no objections to Ellie’s participation’. Depending on the
degree of risk assumed to be involved, information might also be communicated

by announcements or leaflets, as long as it can reasonably be assumed that the
intended recipients will receive the information. There should, of course, in every
case be a clear explanation of the provisions for opting out and people should be
given the easiest possible way of doing so – by text, letter, telephone, social media
or email, or face-to-face with you.
There are pros and cons to opt-in versus implied

Always seek the consent of
participants, by either asking them consent. If there is any degree of significant
actively to opt in to the research, risk, then opt-in procedures are clearly to be
or, if there is less risk, giving them
the option to opt out – in other
preferred. Opt-in consent involves more work
words, you assume acceptance of for the researcher than does opt-out or implied
participation unless the participant consent. However, there is some evidence also
actively contacts you to indicate
otherwise. that samples may be skewed by the use of opt-in
procedures (because only really willing people
opt in, and this may exclude important subsections of the population). Junghans
et al. (2005), working in a medical context, therefore suggest that ‘the opt-out approach
should be the default recruitment strategy for studies that pose a low risk to patients’.
Risk
Risk … low risk or high risk
‘Risk’ refers to the chance that someone could be harmed, psychologically or physi-
cally, by involvement in your research. Making judgements about the harm that
could occur involves identifying the potential hazards that might confront the
research participant. Whether a hazard presents a low risk or high risk is a matter of
judgement and if you are unsure about what might constitute high or low risk you
should discuss the matter with your supervisor.
Potential hazards include:
• something that may cause psychological or physical harm to participants or others

• concealing information about the project from participants, or deceiving them
about its objectives
• the possibility of damaging the standing or reputation of participants or others
who may become involved peripherally
• impinging on the privacy of participants or others
• the possibility of breaking the law
ETHICS 93
• the possibility of harming a community in some way (for example, by drawing

attention to differences within it).
If there is judged to be any significant risk concerning any of these matters, you
should consider its extent and duration. You may need to consider also the extent
to which any risk can be mitigated or ‘undone’, either naturally or through steps
taken by you. You will also need to weigh any risks against benefits or scholarly
merit, which may include particular advantages to participants or others or any
potential advance in knowledge. You should be prepared to expand on all of these
and make a good case for your research when you seek ethics clearance.
As a guide, Table 5.1 offers a selection of different kinds of study so that you
can get an idea of the dimensions on which ethics judgements may be assessed
and the conclusions that could be drawn about whether these are high risk or
low risk. Table 5.1 reveals, I hope, that judgements about risk are difficult to
make and that the conclusions are never definitive. For example, while I have
suggested that unintrusive observation in a teacher’s own class may be low risk,
others may disagree, based on the potential obtrusiveness of observation and
the dangers of class students realising what the teacher is doing. However one
chooses to categorise it, it would be necessary in this example to inform parents
and children in broad terms of the intended research, with the opportunity given
to veto the activity.
If, in seeking ethical clearance (see ‘Getting clearance – ethical review’, below),
you are asked about risks and benefits, you should spell out:
• the physical, psychological, social, legal or economic risks that may be associated
with the research and outline the arrangements that you have put in place to man-
age that risk
• if there are likely to be any immediate benefits to participants
• if there are any specific risks to researchers
• if any degree of concealment or deception is felt to be necessary, and if so why
• how potential benefits to participants outweigh any risks
• any arrangements that you have put in place for support after the study.
Issues of consent are particularly important if you are working with children. There
is an unequal power relationship between any adult and any child, and your posi-
tion as a researcher heightens the perception of authority held by the child. Good
discussions of the ethics of working with children are provided in Alderson (2004),
Kellett (2005), Lewis and Lindsay (2000) and Simons (1989, 2009).
Table 5.1 Ethics: low risk or high risk?
Risk level Case study: subject and object Methods used Why this risk level?
No risk Subject: The Cuban Missile Crisis Documentary analysis of contemporary There is no direct involvement with
Object: Understanding the role of media commentary, government records people (that is to say, research
militarism in US foreign policy, 1960–70 and interviews with key individuals participants) in the data collection
Low risk Subject: A child who ‘freezes’ when Observation over a term of the While this could be high risk if any direct
asked to take a test in class circumstances of each occurrence of involvement with the child were to be
Object: Insight into the consequences for ‘freezing’ in the teacher’s own class planned, risk is much reduced providing
school students of a test-led curriculum observation and note-taking are discreet.
This is quite possible, given that the study
is to be done in the teacher’s own class
Medium risk Subject: A week in the life of an art Observing; interviewing staff, guides, Museums and galleries are unlikely to
gallery volunteers, groups and members of the give rise to much embarrassment or
Object: Understanding responsiveness to public attending the gallery discomfort in respondents. However,
the public in museums and galleries with this particular case study
delicate issues may arise because of
emerging information about tensions or
disagreements among staff about the
running of this gallery
High risk Subject: An intensive study of a young A three-day intensive stay with one While much good could come from this
woman with anorexia nervosa young woman in the young people’s unit study, there are serious potential dangers
Object: Understanding the phenomenon of a psychiatric hospital, ‘shadowing’ her, here for a vulnerable young person.
of ‘taking control’ in anorexia nervosa talking with her and talking with staff Handled insensitively, the young woman
and other patients could become upset or disturbed and
treatment could be disrupted
ETHICS 95
Background checking
If you are involved in any prolonged or regular contact with children or vulner-
able people as part of your research, in the UK you may need to undergo a check
under the Disclosure and Barring Scheme (DBS). (For various background screening
checks in countries other than the UK, see https://cbscreening.co.uk/news/post/uk-
vs-world-background-screening-compare/).
Situations in which an individual has incidental, irregular or indirect contact do
not normally give rise to the need for DBS checks. Decisions about whether you
need a DBS disclosure should happen as part of a risk assessment process (see
Table 5.2), taking into account the exact nature of your contact with individuals
and, for example, the nature of any participant vulnerability. If you are in any
doubt, discuss with your supervisor and any stakeholders in the research such as
a child’s teachers. More details on the DBS scheme can be found at www.gov.uk/
request-copy-criminal-record.
Risk assessment
A risk assessment is a formal examination of the potential hazards that could cause
harm to your participants in your research. It is accompanied by an indication of
how serious that harm could be, considering all of the points I have just discussed,
and how you might manage the risk. Normally, given the brief interactions that
researchers often have with participants, a formal risk assessment should not be
necessary, but with certain participants, particularly those who are vulnerable in
any way, it will be worth systematically undertaking a risk assessment, following
the numbered points I give in Table 5.2.
Information sheets and consent forms

When you write or talk to your potential research participants you will have to do
so after having thought about all of the issues above. If it is a complex case study you
will need to produce an information sheet for participants which explains the details
of your project and what you expect to come from it. If the project involves any degree
of discomfort for participants, they should also be
It is helpful to write an information
asked to sign a consent form. Either – information
sheet for participants, explaining
sheet or consent form – should include: what the project is about.
• the title of project and the name(s) of researcher(s) and their institution
• an explanation of what the research is about
• confirmation that involvement in the project is voluntary and that participants are
free to withdraw at any time and/or withdraw data supplied
• an explanation about what involvement will be asked of participants and whether,
for example, audio or visual recording will be involved
• details of arrangements being made to provide for confidentiality, such as anonymisation
• arrangements for data security including when the data will be destroyed
• arrangements for debriefing and feedback.
Table 5.2 Doing a formal risk assessment
Process of risk assessment Examples/issues/commentary

1. Risk identification. What is the nature of • Autistic child exposed to a noisy environment,
the potential hazard (psychological or potentially inducing anxiety
physical)? • Member of staff asked to comment on managerial
style, inducing conflict and stress
• Child with learning difficulties in company of
single adult, raising potential safeguarding issues
2. Exposure assessment. How much Seconds, minutes, hours, days, weeks?
exposure will there be to this hazard? Regularly with the same person?
a. Length of time In the UK, you may need a DBS check
b. Intensity
c. Frequency
Identify any situations where a Disclosure

and Barring Scheme (DBS) check (see
above) might be required
3. Risk potential. Given (1) and (2) how Weigh up the likely impact of the potential hazard,
is this likely to affect the research given the exposure to it. This is a matter for your
participant? own interpretation, taking into account the points
discussed in the ‘Risk … low risk or high risk’ section
above. If in doubt, discuss with your supervisor
4. Risk management. What can be done • Manage research environment to reduce stress
to alleviate the impact of the potential • Carefully consider question wording
hazard?
• Ensure anonymity with explanations of process
for anonymising
• Incorporate ‘member checking’, allowing
participants to scrutinise and comment on your
interpretations of their words
• Seek to minimise occasions where a single
adult is in the company of a lone child or ensure
observation by others
5. Record your risk assessment Keep a record of your initial risk assessment, how the
risks are actually presenting themselves in practice,
and how you are managing them. Share this with
your supervisor and include it in any write-up you
make of your work
ETHICS 97
University College, Badlands

24th October 2016
Dear Parent
My case study in your child’s class

Say something
I am a student at University College, Badlands, and as part of my course I am doing a case about you
study research project on the way in which teaching assistants help children in school. In my spare
time I work as a volunteer helper at Badlands Primary School and Mrs Tumbleweed, Headteacher,
has kindly agreed to allow me to contact parents to ask for their permission to allow their children to be
involved in my project.
We know that teaching assistants provide a great deal of help to teachers. In my research I want to find out Briefly
explain
how teaching assistants work with teachers and children. I hope this will contribute to our understanding of
your project
the ways in which teaching assistants work, and I hope that I will emerge with suggestions about how they
may work more effectively.
I will be observing the work of the teaching assistant in your child’s class and I may be asking children
about their work.
If you would prefer that I avoid observation of your child and that I do not talk to him/her, Say what the
no questions will be asked and there will, of course, be no consequence for participants are
your child. being asked
to do
I will not be sharing any personal information I gather with anyone, even school staff. No names will
be recorded: any information about individuals will have a code on it instead of a name. After my
write-up, all information I have collected will be destroyed.
If you are happy for me to undertake my project as outlined in this letter, you don’t need Explain
to do anything. However, if you would prefer that I do not observe or talk to your son/daughter as anonymity,
part of my research, please complete the tear-off slip below and return it with your daughter or son data security
to give to their teacher. Or, if you have any questions at all, please complete the sheet as and feedback
appropriate.
Many thanks
Lucy Bloggs
Student in educational studies
-------------------------------------------------------------------------------
If you do not wish your son/daughter to be involved in my project or if you have any Consent is
questions at all, please complete the following and return to your child’s teacher: implied
I’d rather my daughter/son is not involved

Delete as appropriate
I’m not sure: I have some questions
My son’s/daughter’s name ___________________________________________
My name _________________________________________________________
 phone me on ____________________________________________________
 text me on ______________________________________________________
 email me on ____________________________________________________
Figure 5.1 An example of a letter seeking participants’ involvement via implied consent
Where involvement is less complicated, a simple letter may be all that is needed, as
long as it covers these issues in an intelligible way. Figure 5.1 shows an example of
a letter which doubles up as an information sheet and a seeking of implied consent.
Other examples are given in the online resources, see: https://study.sagepub.com/
thomascasestudy3e.
Contacting participants
Remember that each situation is different. You will need to assess how best to make
contact with potential participants and how to explain to them the nuts and bolts
of your research. In written communications with participants, always use straight-
forward language. Imagine that you are in a spoken conversation with them. So, for
example, write ‘my study starts on …’ rather than ‘my study commences on …’, or
‘Please read through this’ instead of ‘Please peruse this’. (When has anyone actually
said the words ‘commence’ or ‘peruse’? I’ve no idea why people who devise forms are
so fond of words like this.) Use everyday terms such as ‘information’ rather than ‘data’.
A very good discussion on ways of seeking and gaining consent is given by the World
Health Organization at www.who.int/ethics/review-committee/informed-decision-
making.pdf?ua=1. This gives templates of letters, forms and information sheets.
Ethics and social media

Increasingly, researchers are employing social media in their research, either to con-
nect with other researchers researching on the same topic, or actually to gather
information. Moreno et al. (2013), discussing this recent trend, note that there are
particular risks associated with the use of social media or other online methods.
They recommend that, as a researcher, you should:
• Always present an accurate portrayal of yourself on social media websites.

• Provide contact information for questions during the consent process.
• Avoid presenting participants’ personal information in ways that could identify
them within their institutions, schools or communities.
• Only collect essential data needed to answer the research question, and present
these data carefully to avoid participant identification.
• Develop a Facebook, Twitter or WhatsApp page as a researcher identity, separate
from your own personal page. Participants can then friend you in a professional
rather than personal context.
ETHICS 99
• Consider listing a privacy policy on your webpages, as well as a page that describes
what data you are collecting and how they are used.
Care for your participants – and yourself

I have discussed the many things you will need to consider in order to avoid
causing harm or discomfort to your research participants. If, despite your best
endeavours, though, it happens that you uncover something distressing for your
participants, or you trigger a response which is clearly painful for them, you
should think about whether it is appropriate to offer information to them about
relevant support services. And you should consider, too, the potential for harm
to yourself in doing social research. While most situations will not, of course,
present any danger, some will. If you are meeting people who are strangers, for
example, always take a mobile phone with you and let a friend or family member
know where you will be going, whom you are meeting, and when you expect to
be back. You might suggest that you will phone at a particular time during or after
your fieldwork.
Where do I put discussion about ethics in

my case study write-up?
This is a moot point. Your discussion of ethics can come right at the beginning
in your introduction, but if you put it there it can rather unbalance the narrative
about the purpose of your study. It can have a chapter all to itself, but this seems
almost to treat matters about ethics as a separate, special concern, bolted on, and
distanced from the main body of your work. I think the best place for your discus-
sion about ethics is in a separate section in your design and methodology chapter.
Here, you can discuss ethics alongside your deliberations about design and expla-
nations about how you have gone about your work. You may also want to discuss
ethics briefly as you actually report your findings and analysis, if this seems appro-
priate. Put sample copies of any forms, information sheets, letters to participants,
guidelines, and so on, in an appendix.
Getting clearance – ethical review

A key element in starting your case study is in getting ethical or institutional clear-
ance. This goes under different names in different places, but may be referred to as
ethical review or institutional review. You will have to write an outline of your pro-
posed research which will be looked at by a group known as the ‘ethics committee’
or, more commonly in the USA, the ‘institutional review board’ (IRB).
The use of these formal procedures by universities for ensuring ethical practice
in social science research projects is relatively recent and still regarded by many
researchers, both student and professional, as a bit of a chore. However, this is
not the right attitude, and it stems – I speak only for myself – from a less than
self-critical belief that I could not possibly be doing anything that would harm any-
one with whom I researched. It stems from a belief that my own judgement about
the balance of rights and wrongs is naturally the best judgement and that it is an
attack on my integrity to question this. But the passage of time and experience have
shown me how delicately balanced these matters can be and how necessary it is to
have ethical concerns explicitly articulated and systematically checked. In fact, the
systematic checking of these concerns with another person often will add insights
beyond ethics to the design and conduct of the research.
Your university will have a code of conduct for research, and it will have a helpful
webpage outlining procedures for gaining ethical clearance. The relevant profes-
sional organisations also have policies and guidelines on ethical research. Those
for the British Educational Research Association (BERA), for example, can be found
by entering ‘BERA ethical guidelines’ into your favourite search engine. Others are
given in the box below. All are very helpful, though some (for example, that of
the American Psychological Association) are
You will need to go through your
so long that you may lose the will to live if
institution’s ethics clearance process
before you start work. Professional you embark on them. A very good, well-
bodies have codes of conduct that written one is the Social Policy Association’s
may be helpful in doing this.
Guidelines on Research Ethics.
Codes of conduct and other guidelines

For the codes of conduct or guidelines issued by these professional bod-
ies or government agencies, just enter the words below into your favourite
search engine:
• American Psychological Association Ethical Principles of Psychologists

and Code of Conduct
• American Sociological Association Code of Ethics
ETHICS 101
• British Educational Research Association Ethical Guidelines

• British Psychological Society Code of Ethics
• ESRC The Research Ethics Guidebook
• General Medical Council Good Practice in Research and Consent to
Research
• Social Policy Association Guidelines on Research Ethics
• US Department of Health and Human Services Office for Human
Research protections
• World Health Organization guidelines on submitting research proposals
for ethics review
The key thing to remember in looking at any of these protocols is that it is the
consideration of ethics that is important – not the protocols themselves, helpful
though these can be. A reflection on ethics should be integral, and not seen as a
bolt-on or a chore to be squeezed into your timeframe.
Undergraduate-level ethical clearance

For undergraduate research, ethical review usually happens at school or department
level. Getting clearance will involve completing a form which gives in a nutshell
details of your plans. You will need to think about this right at the beginning of
your project, as soon as you have decided what to focus on and have discussed the
topic with your supervisor. Do not be intimidated by this form or by the procedure
surrounding it. It should not take too long and it may help you to get down your
early thoughts on paper to give structure to your work. In most universities, this will
usually be looked at in-house – that is, in your department or faculty – for approval.
Postgraduate-level ethical clearance

For postgraduate projects there are more complex arrangements. These usually
entail a process that will involve completing an online form first in which you
are asked various questions about your research. Your answers to these determine
whether the research is ‘low risk’ or otherwise. If you have been able to answer ‘no’
to a series of questions, the project is taken to be ‘low risk’ – and that, for you, is
then happily the end of the procedure. You can proceed with your plans.
However, if you have not been able to answer ‘no’ to all of the questions you
will have to proceed to the second stage, which involves scrutiny of your plans by
a university-wide committee. The process of ethical review must happen before the
start of any active empirical work, including any pilot studies.
For your own university’s exact procedure, take guidance from your department,
or just enter ‘University of [name of your university] ethical review’ into your search
engine, or, in the USA, ‘University of [name of your university] IRB review’.
At this level, ethical approval may need to be sought from multiple stakehold-
ers (for example, your university and a health service such as the NHS), certain of
whom may require annual updates, reporting of any adverse events, etc.
Ethics are a specialised and increasingly important topic, and all universi-
ties will have detailed procedures for ensuring that you consider ethical issues
appropriately (my university’s procedures – accessible to anyone – are at: www.
birmingham.ac.uk/Documents/university/legal/research.pdf). If, in your planned
work, ethical issues seem to loom large and you wish to think about them in more
depth, then you may wish to look at the further reading listed at the end of the
chapter.

take this …
Ethics are more than just a process to be gone through. It is essential to consider
carefully the ethical dimensions of a case study, given the intensity of involvement
that may happen in this kind of research. You have to consider especially …
• Your participants. Make sure that they know what is involved and that you give
them ample opportunity to participate or withdraw, as they wish. Consider espe-
cially the needs of vulnerable groups – in other words, those (such as children)
who may not fully understand what is being asked of them, or those who feel that
they have to accede to a request because of their social position (for example,
employees).
• Gaining consent from participants. In doing this, you will need to provide as much
information as possible about the project.
• Openness and sharing of information. Some withholding of full details may be nec-
essary in some circumstances, but this will shift a project into ‘high risk’ and will
need to be explained fully to an ethics committee.
• Confidentiality. There may be exceptions, but the default position is to treat all
data as confidential and anonymise names.
ETHICS 103
• Data security. Keep data secure and destroy after your analysis.
• Getting ethics clearance. All universities and large organisations will have proce-
dures to be gone through. Ensure you do this before your case study starts.
REFLECTIVE ACTIVITY
In Table 5.1 I offered a number of research scenarios and suggested the level of
risk that might be associated with each, from low risk to high risk.
I reproduce the empty shell of the table below. Think of a case study topic, or
take one from the examples I have given already (such as Amelia in Chapter 1),
and outline the methods that might be used in pursuing that study. Then ‘fit’ it to
one of the risk levels in the table and account for why you have decided on this
risk level in the final column of the table.
Case study: subject

Risk level and object Methods used Why this risk level?
No risk
Low risk
Medium risk
High risk
Further Reading
Alderson, P. and Morrow, V. (2020) The Ethics of Research with Children and Young
People. London: Sage.
Good on the ethics of working with children and young people.
BERA (2018) Ethical Guidelines for Educational Research (4th edition). London: BERA
(available at: www.bera.ac.uk/publication/ethical-guidelines-for-educational-
research-2018-online).
These are the ethics guidelines for educational researchers from the professional
association of educational researchers. Similar guidelines can be found for research-
ers in other areas of applied science from the websites of those associations.
Butterwick, S., Head, G., Madalinska-Michalak, J., Raykov, M., Taylor, A., Taylor-
Neu, R. and Zgaga, P. (2020) ‘Introduction: ethical issues in educational research’,
European Educational Research Journal, 19 (1): 3–9.
This is the introduction from a special issue of the European Educational Research
Journal (EERJ), which addresses the issues surrounding current concerns about the
ethical conduct of research in education.
Grinyer, A. (2002) ‘The anonymity of research participants: assumptions, ethics and

practicalities’, Social Research Update, 36: 1–6.
Discusses the value of anonymity and whether in some cases disclosure is to be
preferred.
Honan, E., Hamid, M.O., Alhamdan, B., Phommalangsy, P. and Lingard, B. (2013)
‘Ethical issues in cross-cultural research’, International Journal of Research & Method
in Education, 36 (4): 386–99.
Looks at some important issues in cross-cultural research and negotiating ethics
committees with them.
Malone, S. (2003) ‘Ethics at home: informed consent in your own backyard’,

Qualitative Studies in Education, 16 (6): 797–815.
A critical account of research and its potential impact on participants.
Murphy, E. and Dingwall, R. (2001) ‘The ethics of ethnography’, in P. Atkinson and

A. Coffey (eds), Handbook of Ethnography. London: Sage.
A comprehensive review and discussion of research ethics in ethnography.
Shaw, I. (2003) ‘Ethics in qualitative research and evaluation’, Journal of Social Work,
3 (1): 9–29.
Good on ethics and social justice.
Sikes, P. and Piper, H. (2010) ‘Ethical research, academic freedom and the role of
ethics committees and review procedures in educational research’, International
Journal of Research and Method in Education, 33 (3): 205–13.
This is one article in a special issue of IJRME edited by Sikes and Piper which
addressed a range of issues concerning ethics in educational research and which
looked in particular at the role of the university ethics committee.

See especially Simons’ reflections on principles and procedures on pp. 103–10.
Simons, H. and Usher, R. (2000) (eds) Situated Ethics in Educational Research. London:
Routledge.
See especially Glen’s paper discussing ‘complex integrity’.
ETHICS 105
Social Research Association (2003) Ethical Guidelines (available at: https://the-sra.

org.uk/common/Uploaded%20files/ethical%20guidelines%202003.pdf).
A useful introduction and bibliography.
Wiles, R., Crow, G., Heath, S. and Charles, V. (2008) ‘The management of confiden-
tiality and anonymity in social research’, International Journal of Social Research
Methodology, 11 (5): 417–28.
An excellent discussion of issues around confidentiality and anonymity.
STAGE 2
CHOOSING A CASE AND

Chapters 1 to 5


Chapters 6 to 9


Chapters 10 to 12

In this second part of the book, I look at:
• h
ow cases can be categorised
• h
ow you can identify a case
• t he purpose of your case study
• t he approach you may take in your case study
• t he kinds of processes that you may follow

when conducting a case study
• a
nd how all of these interrelate to create your
unique study.
6
DIFFERENT KINDS
OF CASE STUDIES:
SELECTING A SUBJECT
FOR YOUR CASE STUDY
If you remember, at the end of Chapter 1 I introduced a framework for thinking
about the structure of case studies. It is a classification system – or ‘typology’ –
which splits thinking about case studies into six elements, which translate to
six steps in the construction of a case study. I repeat that structure, those steps,
here …
1. The subject of the study – what is the case you are actually looking at? Is it a
classroom, a child, an institution, a war, a country, a political event, or what? Why
have you identified this subject? It may be one that you know a lot about from your
own local knowledge and experience. Or it may be a key case in the literature. Or
it may be one that is interesting because it is an outlier – different from the norm.
Whichever, this is the lens through which you are examining the topic of theoretical
interest that you are focusing on, namely the object of the study.
2. The object of the study – what is the theoretical topic at the heart of your ques-
tion that the subject of your study is allowing you to explore in detail?
3. The purpose of the study – why are you doing it? Is it to make an evaluation? Is it
out of curiosity, or is it with a particular understanding in mind?
4. The approach of the study – is it to test a theory that you or someone else has
developed about a topic? Or is it to develop a theory from new about that topic?
5. The methods you will use to do the case study. What data collection tools might
you use, and how might you analyse the data that you gather?
6. The process you will use to conduct the case study. Will it be a single or multiple
study, and will you look at the case during one particular moment in time, or over
a longer period?
The choices you make about your subject, object, purpose, approach, methodology
and process will interact in various ways to produce one unique study out of an
array of possible permutations.
In this chapter, then, let’s begin at the beginning with step number 1 and focus
on how you go about choosing your case study.
How do you select your case study subject?

Your research question – if it leads to a case study – will be about something in its
completeness, looking at a process within your choice of focus and at the how and
why of the process. Once you have established this, your choice will be determined
by a number of questions. I summarise these in Figure 6.1.
A local
Yes knowledge
Is it your own case
special
knowledge
that leads you
to this case? Is it a case that reveals
something by virtue of A special or
No Yes outlier case
its ‘specialness’, its
‘differentness’?
No
Is it a classic or
A key case
exemplary case that
Yes
reveals something
from in-depth study?
Figure 6.1 Where is the study coming from? Your subject
Your subject, your choice of focus, whether it be an interesting or unusual or

revealing example, will be critical since it will lead you on to questions about the
kind of study that you ultimately intend to conduct. This is particularly important
for the student researcher, since, as Figure 6.1 indicates, there are two principal
routes that can be followed when making this decision.
SELECTING A SUBJECT 111
The first route may be taken because of your familiarity with it. I have thus called
this a local knowledge case. Is it, for example, your own work or domestic situation
or another situation with which you are familiar? Here, in your own place of work,
your placement or even your home, you have intimate knowledge. You know and
can ‘read’ the people who inhabit the arena. This is a ready-made strength for con-
ducting a case study.
Another aspect of this is that you may be obliged to undertake a research
project as part of your university studies and have limited access to anything
other than your own situation – or, to put this more positively, it may be that, in
your own situation, you can gain access to the richness and depth that would
be unavailable to you otherwise. This is a perfectly valid reason for choosing
your own situation in which to do your research and a good one for doing a
case study. Do consider, though, if you are in this situation, that there are other
forms of research also well suited to small-scale inquiry (I outline some of these
in How to Do Your Research Project, Thomas, 2017).
The second route is taken if you are not choosing your study because of your own
local or special knowledge. Are you focusing on this subject because of its inherent
interest – is it a key case? In the examples that follow in this chapter (summarised
in Table 6.1), Hurricane Katrina is a particularly interesting key case by virtue of
its size, the political context when it hit, and the extraordinary amount of damage
that it wreaked on New Orleans and its environs.
Table 6.1 The origin of your case: a summary of the boxed examples in this chapter
Possible choice –
Kinds of case examples used in
study subject … which means this chapter How to justify your choice
Local An example of 1. Ioan, from a 1. A case of good

knowledge something in your Romanian adjustment
case personal experience orphanage 2. Brandon’s sudden fall-off
about which you want to 2. Brandon in progress in reading and
find out more intimate knowledge of
Brandon’s situation
Key case A good example of 1. Hurricane Katrina 1. A key case of hurricane

something; a classic or damage
exemplary case
Outlier case An ‘outlier,’ showing Kerala A state in India with unusually

something interesting low infant mortality
because of its difference
from the norm
Alternatively, are you examining the case because it is in some way special – in
other words, because it is an outlier that shows something by virtue of its difference
from other examples? An outlier may occur because it is different from the rest, as
in the case of Kerala, the state in India that has extreme poverty yet unusually low
infant mortality rates.
Let us look at examples of a local knowledge case, a key case and an outlier case,
remembering that a key issue in each case will be to identify an analytical frame
that the case is a case of.
A local knowledge case – your knowledge

Research projects often start with some special knowledge, noticing something
interesting or unusual, putting two and two together and, with a spark of curiosity,
a research project is fired and ready to fly. How might this happen?
Ioan: Adopted as an infant: A local knowledge

case
Let’s imagine that you are undertaking an undergraduate degree in psychol-
ogy for which you are required to complete a small research project that you
intend to do in your special area of interest – child psychology.
Down the road from your home, there is a family who adopted a baby,
Ioan, from a Romanian orphanage in 1990, after the fall of the Romanian
dictator, Ceausescu.
˛ Many families in the West adopted Romanian children
at this time. Everyone in your neighbourhood knows about Ioan’s adoption
since the family have made no secret of it.
You know Ioan, now grown up, and his family well – Ioan’s older sister
went to school with you. The starting point is that you have some special
knowledge that you feel may enable you, with the family’s consent, to do
a useful case study. From reading around, you know the results of research
undertaken by Rutter and the English and Romanian Adoptees (ERA) Study
Team (1998) that showed these children were typically severely develop-
mentally delayed on arrival in their new families, but caught up impressively
thereafter. You know also that those who were adopted before 6 months of
age did better than those who were adopted after.
Clearly, you are not able to do a piece of research remotely similar to
that of Rutter in his comparison of 111 Romanian adopted children with 52
native adopted children. You are not going to be able to make meaningful
comparisons between Ioan and his peers in your neighbourhood. What you
can do, however, with your special knowledge and the knowledge you have
from your reading and wider research, is construct a case study that con-
nects what you know about Ioan – he seems to be a happy, fulfilled young
man – with some of the findings of the research: that is, some of the adopted
children proved to have serious emotional difficulties.
How can you construct a study here? You can try to interconnect your own
information with that from your reading to emerge with a topic that focuses
on Ioan’s success, noting that success has not always been the picture for
some of these young people.
How has success been achieved? In answering this question, you could
look at the age at which Ioan was adopted, the nature of the family into
which he was absorbed (the number of siblings and so on), Ioan’s parents’
reflections and Ioan’s own reflections, including his experiences at school.
You will be using all kinds of data for evidence and analysing and inter-
preting the data in a variety of ways – see Chapters 9 and 10.
The point is that the case study will enable you to ‘drill down’ into this
one case – it is as if it were one of Rutter’s sample of 111. Where Rutter, in
his huge piece of research, was able to offer some generalisable findings
about the progress of children adopted from Romanian orphanages, he was
not able to do the kind of study that yours promises to be with your local
knowledge.
Brandon: Not progressing in reading: A local

knowledge case
Kareena completed her postgraduate certificate in education (PGCE) two
years ago and teaches a class of 9–10-year-old children in a large inner-city
school.
One of her pupils, Brandon, is a bright boy and has always done well
in class, but has recently begun to be truculent and his work has dete-
riorated in quality. He is laggardly in wanting to engage in classroom
activity and, most importantly as far as Kareena is concerned, his read-
ing has failed to progress at the same rate as that of the other children
in the class.
(Continued)
Kareena has decided to do an MA in teaching studies at her local uni-

versity. The main element of this is a thesis, which will involve a piece of
small-scale research. From the research methods course that occupies the
first part of the MA, she understands the potential of the case study to illu-
minate puzzling states of affairs in need of in-depth analysis. So Kareena
decides to do a case study, with Brandon as her subject.
Kareena realises that the essence of a case study is to ‘drill deep’ and
look at the situation from as many angles as possible – to take an eco-
logical stance (see Chapter 3). She understands, therefore, that she has
to see Brandon’s behaviour not simply as a function of what is happening
in the classroom, but also as a manifestation of everything happening in
his life.
She decides to frame the analysis that she undertakes in the models pro-
vided by systems psychology (see page 238), wherein she is able to draw a
picture, literally, of the interconnecting forces that appear to be having an
impact on Brandon’s life.
She arranges to see Brandon’s mother, who arrives at school with her new
partner. It transpires that the relationship Brandon’s mum has with her new
partner is far more intense than was the case in her previous relationships.
Brandon’s mum realises that this is having an unusually serious impact on her
strong relationship with Brandon.
Kareena interviews the school’s inclusion coordinator (INCO). She also
talks to Brandon after class and observes him informally in class, taking notes
as she does this. She completes a diary every day about interrelationships in
the class and draws a sociogram of the class.
It does not take systems psychology to understand that Brandon is feel-
ing squeezed out by his mum’s new relationship. The analysis that it offers,
however, placing his life at school in the context of his life at home, is a useful
counterpoint to the predominant way of tackling difficulties at school, which
tends to seek pedagogical (that is, instructional) alterations that might help
to ease the problems.
Kareena’s analysis confirms her original suspicions – that directing extra
instructional resources to Brandon’s new difficulties with reading is likely to
be unsuccessful as he is confused, depressed and increasingly switched off
from everything that school has to offer. He will also, surely, reject new forms
of teaching or additional help organised by the INCO.
The important point about Kareena’s choice of this case study is that it
is one about which she has intimate knowledge. She is part of the situation
that she is researching. She could be a participant observer. She has easy
access to all of the people involved, with some degree of authority and con-
trol over the situation.
Hurricane Katrina: A key case

Having completed a degree in geography, Jake is now studying for a
master’s degree in environmental management and has to do a piece of
research that examines the consequences of a major disaster in which
environmental management has played a part. Jake has an uncle and aunt
and a brood of grown-up cousins in the New Orleans area, which makes
him think that he might be able to get some real-life accounts of what
happened in the Hurricane Katrina disaster. So, he decides to focus on
Hurricane Katrina and the part environmental management played in the
ensuing disaster.
Could Jake simply entitle his project ‘Environmental management and
the Hurricane Katrina disaster’ with a question such as ‘What effects did
environmental management have on the disaster that followed Hurricane
Katrina in New Orleans?’
Well, he could, but it would not make for a very interesting project.
It misses the analytical frame (see page 115), and without this it would
be nearer to a piece of reportage than a research project. Some research
would clearly be involved, but it would not be directed at anything very
much other than a description, ranging widely around the event, with some
superficial speculation thrown in. It could very easily become an account
that showed merely straightforward links: the sea level rose, the levees
became overwhelmed, the city was flooded.
‘Front end’ • Hurricane Katrina as an example of the kind of ‘storm surge’

that is typical of hurricanes. It causes the sea level to rise
in the area of a hurricane and is especially dangerous in
low-lying areas
• New Orleans as a coastal city with large areas below sea level
protected by levees (wall defences)
‘Consequence end’ • Gangs were roaming the city, stealing from homes and looting
from shops
• Around 3 million people were left for over a week without
electricity
• Much of the country’s oil-refining capacity was disrupted
• Many people were made refugees by the disaster, having to
move to temporary homes elsewhere in the country. Some of
these would never return
(Continued)
What, then, is needed to pursue this as a more penetrating piece of case

study research? First, a tighter focus is needed. To narrow down the focus,
Jake would have to do some brainstorming to come up with ideas about the
disaster and what constituted it, both at the physical geography level and
at the social level, and how these interrelated. After the brainstorming, Jake
might emerge with the following as possible elements. Two of these foci
stress the ‘front end’ of the weather event in itself and the consequences
of this; others focus on the ‘consequence end’, looking at the nature of the
disaster, how it became a disaster and whether or not anything could have
been done to avert it.
Each of the points here relates to a specific aspect of the Hurricane Katrina
disaster. For each, Hurricane Katrina can act as a case in its completeness
for Jake’s examination. Figure 6.2 shows a simple storyboard that takes the
‘front end’ ideas forward. It links the disaster that Hurricane Katrina inflicted
on New Orleans to a set of cultural expectations and prejudices about plan-
ning and the role of government – local and national – in planning. Could it
be that these prejudices in some way affected the political climate in which
planning and environmental management took place?
How far did planning plan

for the unexpected?
Could it have been badly planned

How far was all this due to
urban development below sea
cultural and political
level?
prejudices about planning?
What has made Were there wrangles

New Orleans over who should pay for
particularly better levees?
susceptible to storm
surge?
Hurricane Katrina as an example of a

storm surge that causes sea level to
rise in the area of a hurricane –
especially dangerous in low-lying areas
such as New Orleans
Figure 6.2 Storyboard for the consequences of Hurricane Katrina

This is the basis for a thorough case study. Some pilot exploration, per-
haps by examining online local planning records in the local papers or
discussions with relatives, could confirm this as a basis for a plausible prima
facie hypothesis. If it does indeed seem to be a reasonable narrative that
could be explored in depth, then further data can be collected – from news
footage, interviews, books, original documents and archives about building
plans locally and objections made to them.
Alternatively, Jake could focus on the ‘consequence end’ and look at
the huge upheaval caused by Hurricane Katrina – the looting, lawlessness,
displaced people and disrupted power supplies – locally and nationally.
He may wish to examine the extent to which this was the consequence
particularly of a hurricane or how likely it would be for it to happen else-
where and with different environmental events. Here he could look at the
particular features that mark out Hurricane Katrina and New Orleans to
see if they interacted in some peculiar way that made for this collection
of social consequences. A different storyboard would follow, like that in
Figure 6.3.
Was any difference due to the

scale of the disaster?
Was it
Was the outcome
‘head in
worse or better?
the sand’
mentality?
Other kinds of Other places where

natural disaster, floods have
such as forest fires happened Was it ‘refusal
to plan’
politically?
New Orleans and Hurricane Katrina provided

ideal circumstances for maximum social
disruption and upheaval?
Figure 6.3 Second storyboard for the consequences of Hurricane Katrina
The two storyboards show how different narratives are possible from very
similar starting points.
An outlier case
An outlier case is a puzzle. It is a case that is conspicuously different from the norm,
and the question is why. You may have an idea as to a reason (or reasons) and this
will offer a ‘way in’. It may also offer a possible analytical frame (see page 115).
Kerala – Puzzling longevity: An outlier case

An interesting outlier case comes from the work of Caldwell (1986), who
examined the correlation between gross national product per head in a
country and the average life expectancy of a person in that country. In other
words, he was looking at the association of average income with the age
a person could be expected to live. A high correlation was expected and
revealed – that is, as income goes up, so life expectancy goes up. Caldwell’s
great inspiration was to focus on the exceptions – countries (or states within
countries) that bucked the trend. He looked for the outliers.
A shining outlier is the Indian state of Kerala. The state has extremely
low levels of income – at the time of Caldwell’s study, it had a gross domestic
product per person of $160, which placed it higher than only four other
countries in the entire world. Despite this, it had a life expectancy of
66 years – well up with the best performers, including industrialised
states in the West. How could this be explained? What analytical frame
could be put around this phenomenon? Caldwell’s focus was on how
Kerala and two other similar outlier states – Sri Lanka and Costa Rica –
achieved such good life expectancy and if similar routes could be followed
by others.
Looking back at the original statistics, which came from a Rockefeller
Foundation inquiry, Caldwell noted that its author had suggested Kerala’s
achievements had resulted from a combination of ‘political and social will’.
He was left with a series of questions, which included the following:
• In what ways do countries with exceptionally good health records con-

trast with those with exceptionally bad ones?
• Under what circumstances can political and social will be exercised? To
what extent has the political leadership been uniquely shaped by the
history and nature of its society?
• For countries with vastly different histories from those with success sto-
ries, are there lessons to be learned or policies that can be put in place
to accelerate mortality reduction?
His research design route in answering these questions was to conduct a

case study. He had to use available statistics and knowledge of culture and
history to arrive at potential explanations for how Kerala, Sri Lanka and Costa
Rica had been able to manage such extraordinary accomplishments in the
health and longevity of their citizenry.
The first job he set himself was to isolate from the statistics not just the
best but also the worst, to see if this offered any opportunity for contrast
(see Figure 6.4). In doing this, he was struck by what appeared to be a cul-
tural difference between the best achievers and the worst achievers. The
worst achievers happened to be Oman, Saudi Arabia and Iran, with incomes
15 times higher than the poor states with good life expectancy, yet life
expectancy down in the 50s. What could the differences be between the
achievements of Kerala, Sri Lanka and Costa Rica on the one hand, and
Oman, Saudi Arabia and Iran on the other? Caldwell hypothesised that
cultural and religious differences were at the root of many of the contrasts
between these two groups of states. In the cases of Kerala and Sri Lanka,
the Buddhist tradition stresses ‘enlightenment’, which can be interpreted
in Western terms as education. By contrast, the wealthy poor achievers –
Oman, Saudi Arabia and Iran – were characterised by cultural and religious
traditions that separated women, giving them limited access to education.
There were not only low levels of female schooling in these countries, but
also low levels of family planning and limited access to employment for
women outside the household.
2000
1800
1600 Most states in the
1400 scattergram conform
Mean income
1200 to a pattern that shows

a correlation between
1000
income and life
800 expectancy
600
400
200
0 Kerala is a ‘good’
0 20 40 60 80 outlier: it has high(ish)
Life expectancy life expectancy despite
serious poverty
Figure 6.4 Life expectancy in years against mean income ($)
(Continued)
The link with female education is striking, notes Caldwell (1986: 177): ‘For
most countries in both lists the 1982 infant mortality rankings are very close
to the 1960 female primary school enrolment rankings’ – as education of
females goes up, infant mortality goes down.
He concludes that there are some remarkable parallels between these
states. These parallels include a substantial degree of female autonomy,
dedication to education, open political system, largely civilian society
without a rigid class structure, history of egalitarianism and radicalism and
national consensus arising from political contest.
In a fascinating and wide-ranging analysis, Caldwell suggests reasons for
the importance of the position of women in society, from the likelihood of
girls becoming nurses to the position of children in society being higher
where that of women is higher.
He reports that an early anthropologist, John Davy, noted the Sinhalese
treated their children with ‘extraordinary affection’, attending to them when
sick in the way that other groups he had studied did not.
Caldwell’s conclusion was that women’s education may be important not
just because women may know more about, say, nutrition and development
to help their babies, but also because then women are valued and included
– something more intangible but nonetheless seemingly just as significant.
Thus, the provision of education for all seems to be operating at two levels: it is
providing for individual knowledge but is also symbolic of society’s affirmation
of once-marginalised groups and its endorsement of those groups’ partici-
pation and inclusion. So, in some way, this provision seems to be endowing
added value. Perhaps this added value comes in the form of a boost for iden-
tity, status, belonging and self-belief. Inclusion, and the status it brings, breeds
not only health but also the conditions for learning and growth to occur.
He looks also at historical and cultural investment in health and hygiene,
noting that the high performers had a record of valuing these matters and
investing powerfully in them since the nineteenth century. Education is
important, too, but not just as education per se. It is linked with an active
and participatory political system that in some way enables the hearing of
the people’s ‘voice’.
He compares this with what happens elsewhere in India: ‘The ineffi-
ciency of much of the health infrastructure in rural India is explained by
both the low educational levels and the political passivity of the poor’
(Caldwell, 1986: 203). There appears to be a symbiosis between education
and health.
Although Caldwell at no point explicitly calls his work a case study, it
emphatically is one – of this limited set of countries. In doing work with such
a small number, he cannot definitively assert that factor X or factor Y is, in
fact, the cause of the good life expectancy figures in these outlier countries,
but he can make a good, informed deduction. His initial speculations and
conjectures led to conclusions informed by evidence from historical records,
other statistics, anthropological studies, and reasoning about cultural differ-
ence. In fact, while Caldwell’s study starts off as an epidemiological inquiry
based largely on statistics, it ends more as a comparative history. The inclu-
sion of all the forms of data, argument, reasoning and analysis is a form of
triangulation that should not be downplayed in its importance.
Caldwell’s informed hypothesising is the foundation stone of good anal-
ysis. As long as it involves the intelligent putting together of evidence, it
provides the basis for finding solutions and solving problems.
Always, though, we should have at the front of our minds the importance
of the need to be critical in this or any other kind of thinking when conduct-
ing a research inquiry. We should keep to the fore the need to be our own
most strident critics of the explanations we propose.
If you were doing Caldwell’s study today, you would be able to draw on
a wider range of studies that have used Kerala as an example of something
extraordinary. For example, the Nobel-Prize-winning economist Amartya
Sen (1999: 45–8) describes the process operating in Kerala as a ‘support-
led’ approach to improvement in life conditions: ‘The support-led process
does not operate through fast economic growth, but works through a pro-
gram of skilful social support of health care, education and other relevant
social arrangements.’ You could draw on Sen’s notion of ‘capability’ among
a nation’s citizenry, which he takes to be more important than imposed top-
down measures designed for the good of the people. In other words, this is
about giving the people the opportunity to be involved – a concept similar
to the one Caldwell identified in the participatory political system.
Nowadays, if doing this kind of study, you can draw on some extraordi-
nary banks of data that are bundled together with software that enables
you easily to draw comparisons of the kind that Caldwell would have had to
spend weeks putting together. For example, Gapminder (http://gapminder.
org) can instantly draw a sophisticated set of international comparisons of
everything from national per capita income to life expectancy, oil reserves
and quality of teeth. It is not a problem getting data nowadays – it just needs
leavening with your imagination.
If I were doing a ‘Caldwell-type’ study now and looking at Gapminder, I
might choose to focus on the ‘best’ and the ‘worst’ African countries using
the same measures as Caldwell. These turn out to be Cape Verde and
Swaziland, respectively: Cape Verde has a life expectancy of 72 years with
a per capita income of $3283, while Swaziland’s life expectancy is 40 and its
per capita income $4595. There will be different forces at play for these two
(Continued)
countries, but ones that will reveal fascinating outlier case studies for com-
parison. A storyboard might include factors such as:
• insulation from continental Africa, in Cape Verde’s case

• stability of post-colonial government and its nature
• consequences of AIDS
• use of international aid for large capital projects
• support from émigré communities
• spending on education and health
• availability of new industries, such as oil, solar power, desalination or
tourism.
Same starting points, different paths –

there is no single right way
In Jake’s case study of Hurricane Katrina (a key study), I indicated how rather differ-
ent case studies might follow from the same starting point, and this is a natural and
inevitable feature of the case study method. A case’s uniqueness is defined not only
by the core topic, but also, more importantly, by your own interests and circum-
stances. It is about what you want to find out and the limits of what you can do.
This is one of the challenges of the case study, when its holistic (holistic means
‘to do with the whole’) quality is talked about. Because completeness is being
emphasised, this does not mean that everything has to be covered. The trick is to
get a line of inquiry running through the study. This should either be determined
at the beginning (in the theory-testing phase), as in the case of Hurricane Katrina,
or develop as the study progresses (in the theory-generating phase). (I shall look
more at testing a theory versus generating one in Chapter 8.) The quality of the case
study will be determined by the integrity and the coherence of this line of inquiry
and the extent to which you can weave together data from a range of sources in
addressing the narrative that you develop.
There are no rules, then, about the scope of your study – about how big or small
the focus is. Its quality is determined by the line of inquiry that you choose to take.
As White (1992) points out, the scope can be as broad as Lenin’s analysis of peasant
social formations or as narrow as one of Goffman’s smiles. It is up to you how you
relate the subject to the analytical frame or object.
As I tried to make clear in the examples

There is no formula or right way to
in this chapter, the choice you make regard- decide how your study begins or
ing what kind of case study you do has to proceeds. As White (1992) suggests,
the scope can be as broad as Lenin’s
be for a relevant reason that concerns your
analysis of peasant social formations
research question. It is not something that is or as narrow as one of Goffman’s
imposed on the inquiry at the outset, but has smiles.
to be arrived at following your thought about

your purposes in doing an inquiry, your approach to addressing these purposes
and the processes you will use with that approach. I shall come to this in
Chapters 7, 8 and 9.
To close this section, I should reiterate that it’s not worth getting into a tizzy
if you can’t work out whether your case study is, for example, a ‘key’ case or an
‘outlier’ case. These are categories into which a study is put for the benefit of under-
standing what is going on and what alternatives are open.
The categorisation of case studies merely provides you with a set of options – a
menu through which to look so that you can see what is available and choose from
alternative ways of proceeding. You should not feel straitjacketed by it.
The typology in practice

As I’ve done in previous chapters, let me map one of the case study boxed examples
from this chapter onto the typology. I’m going to use the outlier case of the people
of Kerala and their puzzling longevity (see page 118). Here, the subject is one cho-
sen because of its outlier status. Its purpose is intrinsic and explanatory. Its approach
is both to build ideas and test them (building a theory and testing a theory), and its
process is through a single study with multiple elements, taking Kerala as the focal
point, but also using contrasting and similar cases in parallel (see Figure 6.5).
Single
Multiple
Explanatory Descriptive • Retrospective
• Diachronic
Figure 6.5 Mapping out the design for the case study ‘Kerala: puzzling longevity’
In this chapter, we have primarily looked at your subject, as indicated in Figure 6.6.
As you proceed you have to make a number of further decisions here and these
define the kind of case study that you undertake. These choices about purpose,
approach and process interact variously to produce an array of permutations and,
ultimately, each unique study.

Retrospective
Snapshot
Outlier Intrinsic Testing a theory Diachronic
Key Instrumental Building a theory Single
Local Evaluative Drawing a picture Multiple
Explanatory Descriptive Nested
Exploratory Interpretative Parallel
Sequential
Figure 6.6 First, think about your subject

take this …
There are three main routes you might take in selecting a particular subject for your
case study. You may choose it because:
• you know a great deal about the case in question and you want to understand
some aspect of it – this is a local knowledge case
• it provides a particularly good example of something – this is a key case
• it reveals something interesting because it is different from the norm – this is an
outlier case.
Once you have established your reason for doing a case study, there are a number
of kinds of case studies for you to choose from. The classification in Figure 6.6
enables you to think about structure in such a way that you can review and select
from different design routes for case studies – routes that involve consideration of
purposes, approaches and processes. So, in the next three chapters I shall offer ideas
on how to think about:
• purposes behind doing a study

• approaches to take when you do it

• processes to adopt to achieve the most fruitful crop of findings.
REFLECTIVE ACTIVITY
In Figure 6.5 I mapped out the design for the case study ‘Kerala: puzzling longev-
ity’, using the typology to show how Caldwell developed this case study from its
outlier beginnings.
Choose one of the local knowledge cases – Ioan or Brandon – and trace a
path through the typology as I did in Figure 6.5 for the Kerala case study. Then
explain in a paragraph or two how the person conducting the case study is able
to use their intimate knowledge of the situation in question to offer insights from
various different angles and with different kinds of knowledge, in one of these
local knowledge cases. You may choose to inform your reflections by reading
Lawrence Stenhouse’s advice on the use of case study in research at https://
tinyurl.com/yca9y37m
Further Reading
Bassey, M. (1999) Case Study Research in Educational Settings. Maidenhead: Open
University Press.
Bassey discusses the case study in the context, principally, of educational action
research in classrooms. He expounds his idea of the case study producing a kind of
generalisation that he calls ‘fuzzy’. We can, in other words, be pointed in the right
direction by the findings of a case study without being certain about whether or not
a generalisation will apply in another setting.
de Vaus, D.A. (2001) Research Design in Social Research. Thousand Oaks, CA: Sage.
Useful for putting the case study as a design frame in the context of many other
design frames.
Elman, C., Gerring, J. and Mahoney, J. (2016) ‘Case study research: putting the
quant into the qual’, Sociological Methods & Research, 45 (3): 375–91. https://doi.
org/10.1177/0049124116644273
A useful ‘left field’ view of case study and in particular study identification, recog-
nising that the selection of cases may be viewed through ‘a quantitative template’.
George, A.L. and Bennett, A. (2005) Case Studies and Theory Development in the Social
Sciences. Cambridge, MA: MIT Press.
A good combination of theoretical and hands-on advice on the construction of a
case study.
Goodrick, D. (2019) ‘Comparative case studies’, in P. Atkinson, S. Delamont,

A. Cernat, J.W. Sakshaug and R.A. Williams (eds), SAGE Research Methods
Foundations. doi: 10.4135/9781526421036849021
Goodrick here offers an excellent discussion of the theory and practice of compar-
ative case studies.
Thomas, G. (2019) ‘Case study’, in P. Atkinson, S. Delamont, A. Cernat, J.W.

Sakshaug and R.A. Williams (eds), SAGE Research Methods Foundations. doi:
10.4135/9781526421036812890
Here, I unpack some of the details of case study construction and identification
outlined in this chapter.
7
YOUR PURPOSE:
THINKING ABOUT
THE OBJECT OF
YOUR STUDY
Having considered your subject, we can now move on to the purpose for your case
study. So the focus now shifts to the next part of the diagram, as shown in Figure 7.1.
This represents a move to the theoretical or analytical side of your work. What is
the case – the subject that you have identified – being used to explicate or under-
stand? If you remember, in Chapter 1 I offered the simile of the case, the subject,
being used like a lens through which to view and examine some theoretical theme,
which I called the object. It is this object to which we now turn, in order to under-
stand its potential dimensions.

Retrospective
Snapshot
Explanatory Descriptive Nested
Sequential
Figure 7.1 Looking at your purpose

Why are you doing a case study? What is your object? Is it intrinsic interest or
is it as a means to another end – perhaps a way of getting a more rounded picture
of something, such as a medical procedure? Are you doing it to evaluate a process?
Maybe you want to conduct the study to explore the field before doing a different
kind of study. Maybe you are trying to explain what is going on in a situation.
Stake (2005) uses the words intrinsic and instrumental to separate the different pur-
poses you may have in your mind at the beginning of a study. Let’s look at these first.
Intrinsic
In an intrinsic case study, the subject is being studied not with a secondary purpose in
mind but out of interest, pure and simple. Of course, if we went on to try and analyse
our interest, it could come from any one of a number of angles, but we will put that to
one side for a moment. Sometimes, when a study is undertaken simply for its interest,
it is called ‘blue sky’ research or ‘curiosity-driven’ research – blue sky, because you are
free to think just of the sky; there are no obsta-
With an intrinsic study, you are
cles or barriers to block the view because there is
inquiring for the purpose of
inquiring. It is ‘blue sky’ research. no ulterior motive – that is, you are finding out
for the sake of finding out.
Stake (2005: 445) suggests that the case study is intrinsic
if the study is undertaken because, first and last, one wants better under-
standing of this particular case. It is not undertaken primarily because
the case represents other cases or because it illustrates a particular trait
or problem, but instead because, in all its particularity and ordinariness,
the case itself is of interest.
The intrinsic label is used to separate a study such as this from an instrumental study.
Instrumental
An instrument is a tool and an instrumental study is one that is done with a purpose
in mind. You are using the study as a tool; you are not doing it purely for the love
of knowledge. Is your research aimed at evaluating something? (More on this in a
moment.) Do you have a purpose in trying to understand with a view to making
things better? For example, if your case study is about Joshua’s persistent difficulty
with reading, your ultimate aim is probably going to be something to do with help-
ing Joshua. The case study in this sense is instrumental; it is a means to an end.
YOUR PURPOSE 129
Stake (2005: 445) makes the distinction between the intrinsic and the instrumen-
tal to emphasise the point that some case studies are
examined mainly to provide insight into an issue or to redraw a gen-

eralisation. The case is of secondary interest, it plays a supportive role,
and it facilitates the understanding of something else.
You may notice Stake’s use of the word ‘generalisation’ in this quotation. If so,
you may be saying to yourself, ‘Hang on. Didn’t the bloke who wrote this book
say on the first page that you can’t generalise from a case study?’ Well, yes, I did
say that and this is where it begins to become a little tricky (strokes beard mean-
ingfully), thinking about what the case study can and cannot offer. There may be
generalisations that each of us holds in our heads – not scientific generalisations,
but everyday generalisations – that a case study
With an instrumental case study,
may help to confirm or refute. You may remem-
the inquiry is serving a particular
ber that I discussed this property of case studies purpose. So, the case study is acting
in Chapter 4, where I noted their importance as an instrument – a tool.
for abduction and phronesis.

Beyond distinctions between the intrinsic and the instrumental, you can divide
studies according to whether their purposes are evaluative, explanatory and/or
exploratory – or any combination of these. Let’s look at each in turn.
Evaluative
Evaluation is framed by the expectation that you are doing the research to see how
well something is working or has worked. Something has been changed or a new
idea introduced, and evaluative research is carried out to find out what the change
has led to. Have things got better or worse – or just stayed the same?
There are, of course, many ways to answer questions such as these, and a case
study provides just one avenue into evaluative research. The main thrust of evalu-
ative research will usually involve counting something before the change has been
introduced, introducing the innovation, then counting again at the end. Whatever
the consequence – an improvement, a steady state or a deterioration – we have
grounds for believing that it may have been caused by the innovation. Mind you,
I should stress that the key words here are ‘we have grounds for believing’ since
there are many reasons for such changes to have happened. Indeed, much discus-
sion concerning social science method is about the confidence we can have in the
beliefs that we hold after research has taken place.
The management team: An evaluative

case study
If, for example, you are part of a management team that introduces a new
rota scheme in your finance office and you find that productivity – the
amount of work being got through by each member of staff – drops after
the introduction of the new scheme, rather than rises as you expected, you
will want to pursue this finding further.
You might do this by conducting some interviews to ask those who work
in the finance office what they think is happening. This would be a perfectly
reasonable follow-on from your first finding, but it would not, in itself, be
a case study. A case study would involve an in-depth examination – the
emphasis here being on depth rather than breadth. The aim would be to
gain fine-grained detail about what is going on, looking from many angles
at possibly only one person’s experience. In looking at one person, you will
want to explain their personal and professional circumstances in as much
detail as you feel necessary, and amplify this with all of the situational detail
that you assess to be relevant.
A case study here will contribute to the evaluation, but cannot form the
major part of it. By using an outcome measure of some kind, the main part
of the evaluation can establish that there has been a drop-off in perfor-
mance. Then, and only then, can the case study provide insight into why
that drop-off might have occurred. Given that the case study will be sacrific-
ing breadth for depth, you will probably focus on a limited number of staff
to follow up.
You might choose to focus on only one person – say, Jennifer, who has
been a member of the finance department for 18 months – to get her
impressions of what has happened. You might talk to her in an unstruc-
tured interview, asking for her opinions on what happened before and
after the change, how she felt about it, what its effects were for her. You
might look at Jennifer’s Microsoft Outlook diary and her ‘to do’ lists before
and after the change. You might question her about these (always remem-
bering the need for complete protection of your volunteer informant in a
situation such as this, and being fully aware of the need to consider eth-
ics fully – see Chapter 5). All the time, you will be looking for Jennifer’s
attitudes in this, watching for the nuances in her behaviour that give clues
to what she really means, since this was a finding (the deterioration in pro-
ductivity) that you did not expect and was not expected by the company.
You should be alert for subtle messages that Jennifer may be reluctant to
articulate.
YOUR PURPOSE 131
It is in these subtle clues that you may be able to build a story about what
has happened in this case. The key advantage of doing a case study is that you
are empowered to do this building job – building from varied sources and using
your own knowledge of the world, its intricacies, meanings and webs of under-
standings. With these, you construct an explanation, a story that knits all of the
threads together. The philosopher–educator Jerome Bruner (1991: 2) calls this
‘The narrative construction of reality’. In his famous article in Critical Inquiry,
Bruner (1991: 2) says:
we organize our experience and our memory of human happenings

mainly in the form of narrative – stories, excuses, myths, reasons for
doing and not doing, and so on … Unlike the constructions gener-
ated by logical and scientific procedures that can be weeded out by
falsification, narrative constructions can only achieve ‘verisimilitude’.
Narratives, then, are a version of reality whose acceptability is governed
by convention and ‘narrative necessity’ rather than by verification.
This is important for understanding what the case study offers. Bruner is making
the point that there are different ways of understanding what people do and, if we
are assuming that we can use only in a limited way the methods of natural science
in social science, we must treat as different the knowledge that comes via building
the narrative.
Explanatory
Explaining is probably the most common purpose of a case study. Remember that,
in a case study, you are trading breadth of coverage for depth of understanding,
and potential explanations based on depth of understanding are what a case study
does best.
These explanations may be tentative or context-specific, but it is in the mul-
tifaceted nature of a case study that you get the opportunity to relate one bit to
another and offer explanations based on the interrelationships between these bits.
In being able unselfconsciously to look at these interrelationships, a case study
is, thus, the most powerful engine of potential explanations. We must remember,
however, that the explanations it offers may be limited to the background pro-
vided by the case study’s circumstances. Always be aware that you are looking at
one among many.
Changing patterns of adult employment:

An explanatory case study
Let’s suppose that you are interested in the changing patterns of adult
employment in education. In the USA and Europe, these changes have been
occurring over the last few decades, with far more teacher aides (teaching
assistants in the UK) employed than was the case before the 1980s. Despite
the improvement in teacher–pupil ratios provided by these additional adults,
there is little evidence that the new pattern of employment actually does the
children any good. The outcomes – in, for example, attainment tests – for
children in classes with aides or assistants show little or no improvement
over children in similar classes who have not had the benefit of this addi-
tional resource (Vincett et al., 2005).
Now, this is a finding that merits some thought and, indeed, concern. Why
should it be that attainment does not improve despite the better teacher–
pupil ratios? We can also ask ourselves what the existing research may not
be telling us, since it seems obvious that these better ratios ought to be
doing something.
Storyboarding may help here – it may offer some avenues to pursue, with
the aim of developing an analytical frame.
Figure 7.2 shows a very small storyboard. You will see that there are two
directions of travel in the questions that stem from the question at the bot-
tom. Those on the right are about the things that might be going wrong
due to the advent of additional people in the classroom – things that might
offer an explanation for the findings of research about children’s attainment.
We tend to think that extra people can bring only benefits, but maybe this
brings problems just as big as or bigger than the benefits.
The questions on the left, by contrast, are about our focus in the research.
Is the focus on attainment curtaining off some aspect of improvement in the
classroom that is unrelated to attainment?
A case study might offer an explanation here. Maybe, though, I should be
more tentative and say instead that it may offer some clues to an explana-
tion. How would you construct a case study that might begin to examine this
teacher–pupil ratio paradox?
First, we should draw a line around the situation that we wish to study –
we should define its parameters. We should be aware that what we are
doing is looking for pointers to an explanation, not provide definitive
answers. So we will not seek to define this or that variable based on a
hunch that it may be important and then test its importance with a study
or a series of studies. Rather, we will systematically seek hunches, seek
potential explanations.
YOUR PURPOSE 133
How can you start on this? Imagine that you are throwing a rope, a lasso,
around one situation that in some way exemplifies the area of your interest.
By defining the area of interest in this way, you give yourself the opportu-
nity to explain in this small situation. Here, you are thinking of classrooms
with extra people working alongside the class teacher, so choose one such
classroom. In this classroom, your attention is focused single-mindedly on
the problem: why is it the case that research shows additional adults do not
seem to help student attainment?
Once the question has been framed and the rope thrown round the situa-
tion, you can then begin to think around the issue, using a storyboard of the
kind shown in Figure 7.2. The boundary of the study here is this class – this
teacher with her 29 children in this room in this school. It is not, as would be
the case in other kinds of research, this variable isolated and manipulated in
this way, judging its effects on that variable. Your aim is explanation – albeit
tentative explanation, but explanation all the same – and this involves taking
a holistic view.
To get this holistic view with explanation in mind, you will be looking from
a number of different directions, and those directions will be determined by
the ideas that have emerged from your first storyboard.
For the first view, going back to Figure 7.2 and looking at the left-hand side
of the diagram, you might decide to look not at attainment but instead at
something that seemed not to be examined in the literature on this subject –
the children’s attention.
Might other things be

improving – such as
student happiness?
Is the teacher now faced

with difficult staff
Are we looking at the right management issues?
things (attainments)?
Could these people be

Why aren’t attainment getting in each other’s
figures improving with way?
better teacher–pupil
ratios?
Figure 7.2 Storyboard of questions about the effects of having extra help in the
classroom
(Continued)
So, you set up two conditions for observing the children: first, on Monday
morning when Mrs Speed, the teaching assistant, is present in the class; and,
second, on Tuesday morning when the children are doing the same kind of
work but Mrs Speed is not present. You look at each child in each situation,
video recording their behaviour. Then you compare the two situations. From
this, you are able to say whether or not there was a significant difference in
the children’s behaviour in the two situations. This on its own, though, is not
enough.
For a second view, going back to the left-hand side of the diagram again,
moving up, we come to happiness. Even if attainment cannot be shown to
change, might the children be happier with an aide or assistant present?
How could you examine this? Could you look at how much the children are
smiling? Probably not a good idea since, even if you could find a good way to
measure a smile (impossible, surely), you could not be sure that smiling was
an accurate representation of happiness. Maybe children who appeared to
be smiling were merely trying to contain excess wind. Maybe they were smil-
ing because they were laughing and joking in a classroom with little control.
Would this be a measure of happiness? We know that children do not like
being in uncontrolled classrooms, so, even though they may be laughing,
they may not be happy. (Social research – it’s not easy, is it?)
Perhaps it would be better to use some other indicator of happiness, such
as a self-report. You could, in other words, simply ask the children, using
charts with smileys (), how happy they are in the different situations.
Moving to the right-hand side of Figure 7.2, we come to a different set
of issues entirely. Here are some quite profound questions about the ways
in which people work together, and some interesting ideas have emerged
from your storyboarding about the effects that putting two people together
in a working environment can have. In fact, academics in economics and
management science know these effects well, summarising them in ‘the law
of diminishing returns’. This law is that each unit of extra resource – whether
it be people or machines – will produce less than the one before it. It may
not be 100 per cent appropriate to this situation, but it is a clue as to what
might be happening when there are extra people in the classroom. More
people equals more confusion if the extra people do not know what they
are doing. (They may not know what they are doing since teachers are unfa-
miliar with having to manage and work with other people in their territory.)
So, how could you, in your case study, try to examine whether or not this
law of diminishing returns might be operating? You could do it in a number
of ways – perhaps actually observing to see if there are any signs of confusion
or irritation. You might ask the teaching assistants themselves, by interview or
questionnaire, how they feel about being in this situation. Do they feel unsure?
YOUR PURPOSE 135
Do they feel directionless? You might ask the teacher if she feels unconfident
about how to handle and manage the additional staff. You might watch the
children for signs of them taking advantage of confusion between the teacher
and the assistant. Are they capitalising on ambiguity?
So, with the case study you would be doing the ‘drilling down’ into this situation
to provide the explanation for the kind of paradoxical finding about attainment.
Exploratory
An exploratory case study will be done where you are faced with a problem or an
issue that perplexes you. You need to know more: what is happening and why?
You may have little preliminary knowledge of the issue – or, at least, little rounded
knowledge of it. You may have some familiarity with it, but what you know may be
one-dimensional since you only ever see the situation as one person – in your role
as a teacher or as a visiting parent or the school’s headteacher or a midday supervi-
sor who helps out as a volunteer during lessons.
Let’s take an example.
A child who freezes: An exploratory case study

Imagine that you are a teacher and there is a child in your class, Joel, who
freezes when it is time for the annual SATs to be given. Exploring this in a
case study, your initial brainstorming will be around the potential reasons for
this, which you can follow up with your multifaceted exploration.
Your ideas are likely to be far more tentative and less well formed than
they would be with an explanatory study, where you have some clear ideas
that you wish to pursue. An exploratory study will therefore be more open-
ended. You will be seeking ideas from others using a variety of methods,
watching and making notes and conducting interviews. These will give you
leads, corridors to follow – some of which may lead to an open door, others
to a dead-end.
(Continued)
Suppose that the first part of your study is to ask a range of different people
about their impressions of Joel. You ask each of them – his teacher, parents,
headteacher, the midday supervisor – why they think that Joel freezes when
the time for SATs is approaching. Given that the study is exploratory, you will
principally be listening for ideas, rather than presenting ideas to test against
your respondents’ views. It would be inappropriate to say, for example, ‘Do
you think that Joel has dyslexia?’ This would be very much framing the initial
ideas of the respondents according to constructs you have suggested –
putting words into their mouths.
You might set aside an hour each morning for a week to watch Joel in
class at a time when he and his peers are doing more formal work. You would
likely observe in an unstructured way, trying to pick up any clues about Joel’s
behaviour from the contextual antecedents (that is, the things that hap-
pened before) and consequences (the things that happened after) of formal
work. Did he seem to use avoidance strategies? Did he look anxious? Did he
try to remove himself from situations with other students? Did he start mess-
ing around when structured work was imminent? Did his teacher or teaching
assistant seem to be expressing any anxiety through their behaviour, and
could this be transferring itself to Joel?
It is with such exploratory questions, followed by observations and open-
ended interviews, that you would begin to conduct an exploratory case
study for Joel’s situation. As ‘leads’ appear, you follow them up. Some might
seem to have substance, and it is these that you then test further.
Here is another example, from real life. It is a much more complex one, in the
arena of public health. It shows, I hope, how useful the case study can be when you
have a seemingly amorphous problem. In fact, it shows not just how the case study
can be valuable in an exploratory sense, but also how these explorations can be
taken further, narrowing down the area of search as more data are gathered.
A public health mystery: An exploratory

case study
After the arrival of ‘mad cow disease’ (properly called bovine spongiform
encephalopathy, or BSE) in the UK in the 1980s, it was realised, after labora-
tory work, that the disease had been transferred to the cows from ground-up
sheep (including their brains), which had been fed to the cattle as bonemeal.
YOUR PURPOSE 137
Because cows are, by nature, vegetarians, the eventual assumption about

the infection was that they had no evolutionary protection from the diseases
that might be transferred to them by eating another creature.
A collateral assumption was that the new disease, found to be caused
by an agent called a ‘prion’ in the cows’ brains, could not be transferred to
humans (we being omnivores and, therefore, evolved to eat other animals).
The UK agriculture minister at the time, John Gummer, even fed a burger
to his daughter to prove to the nation that there was no danger to humans.
Then, worryingly, it was noted that the incidence of a human brain dis-
ease called Creutzfeldt–Jakob disease was rising. It was a particular kind of
Creutzfeldt–Jakob disease – a variant – so the new illness came to be called
variant Creutzfeldt–Jakob disease (vCJD). Eventually, the link between BSE
and vCJD was established – BSE seemed to be causing vCJD – but it was
expected that the risk of contracting the disease was tiny. In 2000, though, it
was noted that there was a cluster of people with vCJD in the Leicestershire
village of Queniborough. Between August 1996 and January 1999, five peo-
ple contracted the disease in the village, all of whom subsequently died.
This was now a huge concern for public health. Could it be that this was
the beginning of an epidemic? Was it possible that everyone in the UK who
had eaten beef was susceptible to contracting vCJD? The cluster had to be
investigated quickly to explore if there were particular circumstances that
were causing this cluster and if many more clusters were likely to be found
across the nation.
The only form of investigation possible here was the case study. The case,
defined by this cluster, had to be explored for the circumstances that sur-
rounded it. Were there peculiar circumstances in Queniborough that led to
the vCJD cluster? Two researchers, Gerry Bryant and Philip Monk (2001), led
the investigation for Leicestershire Health Authority, and their research pro-
vides for us a classic example of an exploratory case study.
A number of lines of thought were immediately open to the investigators.
Aside from the unwelcome possibility that this was the first of many such
clusters to appear, there were also competing explanatory themes that had
to be explored.
The disease had a long incubation and the young adults affected had
probably contracted it several years previously, at primary school. Might
hygiene in the primary school kitchen be to blame? Might a particularly
badly infected batch of cattle be responsible? Was it something to do with
how the cattle were slaughtered or the meat transported? Yet more poten-
tial explanations open to Bryant and Monk included:
(Continued)
• infection via blood transfusions, dental surgery, injections, body piercing

and so on in this locality
• baby foods consumed locally, given the youth of the vCJD patients
• lack of manganese in the diet, since preliminary investigation showed
that this area of Leicestershire has a particularly low level of manganese
in the soil, so manganese elsewhere might be providing some kind of
protection.
Any of these lines of thought could provide a promising avenue to pursue …

or lead to a dead-end (see Figure 7.3).
Test the hypothesis further
Which now seems most/least likely? Start from here
How are they fed and kept here?

Are there
signs of Explore. Are there Explore. Do big
clusters other signs of high differences exist in
elsewhere? Explore. If the
dental-related slaughtering and
answer is ‘yes’ …
infection? butchery practices?
More Other
infected sources of
cattle infection,
The start here? such as
Local meat-
of a national from
epidemic? handling
dentistry?
practices?
What caused the cluster of

vCJD in Queniborough?
Figure 7.3 Exploring and generating ideas is a haphazard process, so follow the
ones that seem most likely first
The first rule is to think. The investigators had to think about these alterna-
tives and start with the most likely. They started by looking into cattle-rearing
practices locally. They found that:
YOUR PURPOSE 139
Local beef cattle were raised alongside dairy cattle. This meant that
beef cattle were fed meat and bonemeal supplements from the age
of 6 days rather than 6 months, which is the case for pure beef herds.
(Bryant and Monk, 2001: 2)
The upshot was that the cattle raised here had a higher chance of incubating
BSE than did cattle elsewhere, having had a longer exposure to the prion in
bonemeal than in many places elsewhere, and this area of Leicestershire did
indeed prove to have a high incidence of BSE.
A strong additional candidate for culpability was the nature of the meat
eaten in this particular place, which led the investigative trail to butchers’
shops. At this point the story gets rather gory, so, if you are squeamish, it is
best to avoid the next few paragraphs.
Butchers in the area obtained their carcasses in one of three ways:
• they slaughtered the animals themselves

• they bought them from local, small slaughterhouses
• they bought them from large, wholesale suppliers.
So, the investigators also looked in more detail at the differing cattle-slaugh-
tering practices in each place.
They discovered that the local practice in small slaughterhouses included
the insertion of a ‘pithing rod’ after the cattle were killed with a bolt. The pith-
ing rod is pushed through the bolt-hole in the brain of the cow to prevent the
animal kicking, which can occur after the use of the bolt. This practice is not
used in large slaughterhouses (and the practice has now been made illegal).
There were two other major differences:
• in large, ‘industrial’ slaughterhouses, carcasses were hosed down after

killing, whereas in the smaller ones they were wiped, because the lore of
traditional butchery was that hosing made the meat ‘go sour’
• the traditional practice, employed by the small, local slaughterhouses
and butchers, had been to split the head to remove the brain, offering
further opportunity for cross-infection; this practice was not used in the
large slaughterhouses.
Local killing and butchering practices, then, provided a likely candidate for
infection. They could actually be responsible for spilling the infective agent
(Continued)
out of the brains of the cattle and on to the raw meat. Using the traditional
butchers’ methods, this material would not be hosed away. In fact, the use of
cloths to wipe the carcasses could actually make matters worse, spreading
the infective agent around.
Following their initial investigations, Bryant and Monk refined their origi-
nal ‘thought bubble’ ideas (see Figure 7.3) to arrive at a hypothesis that:
1. local cattle were particularly predisposed to being infected because, in

local rearing practice, consumption of bonemeal happened at an early
age;
2. then, given the local slaughtering and butchering practices, the infec-
tive agent was spread over the carcasses and, from there, the infective
agent was transferred to butchers’ local customers (it was known from
experimental work that the infective agent – the prion – is not destroyed
by cooking in the same way that other pathogens are).
Bryant and Monk tested the second part of this hypothesis by comparing
the source of meat consumed by vCJD patients with that of non-patients.
Was there a pattern, with the families of patients with vCJD sourcing their
meat from traditional local butchers? Using a structured questionnaire, they
first of all asked relatives of the infected patients about diet and purchasing
patterns for meat during the 1980s. The responses here were compared with
those of a set of ‘controls’, these being comparable people who had not
contracted vCJD.
The findings were, in short, that four of the five vCJD casualties had regu-
larly consumed meat from one of two local butchers, who had sourced their
meat from cattle killed using traditional slaughtering methods. The meat
bought by the controls was from a far wider range of retailers (supermarkets,
freezer stores and other butchers). The latter – the ‘safe’ retailers – were
questioned about the sources of their meat, which turned out, as expected,
to have come from further afield and not butchered using the traditional
methods.
Using simple statistics, it was then possible to show that the likelihood
of this pattern occurring by chance was very slim indeed. The point I keep
making, about the impossibility of generalising from case study, however, is
still valid and Bryant and Monk (2001: 5) were wisely cautious in their conclu-
sions. They say that the study:
provides a biologically plausible explanation … [but] On a national

basis, it is unlikely to explain how all of the people who have devel-
oped this disease were exposed to the BSE agent.
YOUR PURPOSE 141
Enormously helpful information was provided by this case study, however.

An unexpected additional piece of information also came from this study:
Analysis of the exposure of our cases to this butchering practice points

to an incubation period for the development of vCJD of between ten
and sixteen years. This is the first time that it has been possible to pro-
vide an estimate of the incubation period. (2001: 5)
This case study is from the field of public health, but the steps taken by these inves-
tigators when doing this exploratory work are common to all such investigations,
in any field. They involve:
• initial fieldwork or a reconnoitre to gather facts

• the posing of potential explanations or solutions
• exploratory work to examine the likelihood of any of those potential explanations
having substance – having ‘legs’
• the testing of those potential explanations.

take this …
While the subject, the case itself, is the focus of your direct interest in a case study,
it is, in a sense, only a tool you use to examine or explicate some theme or topic.
You are using the subject to understand this theme or topic. In this chapter we have
looked at the potential shape and dimensions of this theme, this analytical frame,
which can be thought of as the purpose or the object of your study. We have noted
that the purpose of your study may be:
• intrinsic – looking at a subject purely out of interest, to identify or illuminate a

theoretical topic
• instrumental – using the case study as a means to an end, better to understand
some theme, process or idea.
Beyond this, we can divide case studies into:
• evaluative – where the aim is to find out how well something is working or whether
it has worked as expected
• explanatory – where the phenomenon in which you are interested needs ‘unpack-
ing’, the connections between different parts of the issue need unravelling and the
case study offers a route to explanation
• exploratory – where little is known and the principal purpose is to establish the
‘shape’ of the problem or issue.
The issue is about the way in which the analytical frame for your study will take
form.
REFLECTIVE ACTIVITY
In two of the boxed examples of case studies in this chapter, I have used story-
boards (I outline how to use them in Chapter 2) to help identify an object or
purpose for the case study.
Now look at the boxed case study under the ‘Evaluative’ subheading in this
chapter. I have based it around the paradoxical finding that a new office rota
scheme has resulted in a fall-off in productivity in the office, and have moved
straight into a theory-building case study route which allows ideas to emerge as
data collection proceeds. In other words, I have not used a storyboard: I wanted
ideas to emerge from research participants. An alternative to this would have
been to begin with a set of ideas about why productivity might fall in a sce-
nario such as this. How would you gather such a set of ideas? A good way to
start would be to draw a storyboard. Draw a storyboard that outlines the various
hypotheses and directions you might follow in this inquiry.
Further Reading
Acemoglu, D. and Robinson, J.A. (2012) Why Nations Fail. London: Profile Books.
The authors of this book ask why some nations are more prosperous than others.
They set out a series of key cases which are instrumental in that they aim to demon-
strate the validity of an already-developed thesis.
Becker, H.S. (1996) ‘The epistemology of qualitative research’, in R. Jessor, A. Colby

and R. Schweder (eds), Essays on Ethnography and Human Development. Chicago:
University of Chicago Press, pp. 53–71.
This is a fascinating discussion, essentially, of what we mean by ‘empirical’. It plays
down the distinctions between qualitative and quantitative and makes the case
YOUR PURPOSE 143
that different kinds of research find common cause in being empirical – in other
words, in finding out from (our own) experience and analysis, as distinct from what
some very important authority tells us.
Booth, W.C., Colomb, G.C. and Williams, J.M. (2016) The Craft of Research (4th
edn). Chicago: University of Chicago Press.
Booth and his colleagues give one of the best overviews I have come across of why it
is important to look at your purpose. Chapter 14 is especially helpful for establish-
ing the rationale for your work. The book looks at research as a holistic, integrated
process that, although it is integrated, does not just ‘happen’. You have to think
about your purpose, what has been done before and what you have to offer to the
area with your proposed research. Although they are not looking at the case study
in particular, the stance they take on research is very relevant to the case study,
stressing, as it does, the integrity of the complete work. You have to work at making
the integrity real for the reader, though, and this is what the book helps you to do.
Thomas, G. (ed.) (2013) Case Study Methods in Education: Volume 4: The Case Study in
Practice – General Issues and Specific Examples. London: Sage.
Certain examples are given in this chapter (that is, in this chapter of this book), but
the volume I refer to here contains many examples of published case studies con-
ducted for different purposes in different contexts.
Here, Kevin Myers and I give detailed examples of different kinds of case study and
how they can be analysed according to the typology I have outlined.
Tight, M. (2017) Understanding Case Study Research: Small-Scale Research with Meaning.
London: Sage.
A well-organised book on the nature and use of case study research, with much
good, practical advice.
8
YOUR APPROACH:
THEORY TESTING OR
THEORY BUILDING;
INTERPRETATION
OR ILLUSTRATION
Retrospective
Snapshot
Explanatory Experimental Nested
Sequential
Figure 8.1 Looking at your approach
Having looked at your purpose in Chapter 7, you can think now about the approach
you are going to take in your study (see Figure 8.1). Are you, for your analytical
YOUR APPROACH 145
frame, or object, going to try to build your own set of ideas from scratch (build-
ing a theory) or are you going to test ideas that you have come across elsewhere
(testing a theory)? Are you going to do both of these, as in the vCJD case study in
Chapter 7? Is your approach in the case study to illustrate something, perhaps a
process that you need to explicate in more detail for some purpose? Will you be
seeking to uncover and interpret what is going on in the area of your searchlight
beam?
Theory testing or theory building

Let’s begin with case studies where the object is to build or test a theory. Presenting
these as contrasting is perhaps a little confusing since they are better thought of as
being at opposite ends of a continuum of kinds of inquiry. At one end, you are try-
ing to develop shape from data – your mind is open to the shape that it will take,
rather like the bits in a kaleidoscope when you twist the barrel.
Building a theory Testing a theory
At the other end of the continuum, the shape is already there, like a template,
and the job you set yourself is to see whether or not the template, the theory, in
fact provides satisfactory explanations for the data you are collecting. Remember,
though, that all research – a case study or otherwise – begins with an inspiration or
a hunch and it is this that leads to the journey of your research. You cannot begin
without the hunch stage, whether you are building or testing a theory.
Building a theory
‘Theory’ is a tricky word in the world of research since it has so many meanings. I’ll
come to this and discuss it a little more in Chapter 11, but, for now, let’s assume that
theory means an explanatory model. Building a theory is therefore about develop-
ing, almost from scratch, a framework of ideas, a model, that somehow explains the
subject you are researching.
So, in a theory-building case study, you are aiming to develop ideas, starting off
in much the same way that you did for an exploratory case study, but taking the
process much further. This is much more than exploration. You build a framework
of ideas that has no overt connection to preformulated notions about what is
important here. Perhaps that is putting it too strongly, since it is impossible not to
connect to pre-existing ideas. What I mean is, you have no allegiance to pre-existing
ideas and no presuppositions. Your pre-existing ideas will always be there – you
cannot erase them from your mind, after all – but you must have no commitment
to them as potential answers. Your mind should be open to new interpretations
suggested by your data.
Global terrorism: Building a theory with a

case study
Let’s suppose that you are a student of forensic psychology and interested
in the way that global terrorism is affecting people’s day-to-day lives. Is it
influencing the way that people go about their workaday existence? Are
there ways in which a fear of terrorism changes what people do? What are
the changes imposed by government and institutions in response to global
terrorism that have an impact on people’s lives? These are the starting points
for your research and might lead to a case study building a theory that in
some way explains people’s behaviour.
Let’s also suppose that you have begun your inquiries by interviewing
some people who are about to go into a London Underground station – say,
Oxford Circus. Most of the people you talk to claim that global terrorism has
no effect on them at all, but, among the group of people you interview, you
come across a few who say that they would be suspicious of a young man
wearing a large rucksack and would try to get into a different train carriage
from him.
Here is the basis for a theory-building case study. Given that we are
focusing on building rather than testing a theory, the emphasis will be on
generating ideas as the research progresses, rather than testing out pre-
formulated hypotheses. While your focus may be on your participants’
constructions about danger, race, religion, youth and their interconnection
(since you have to have a starting point), you will not have preformulated
ideas about how important each of these is, how they were conceived or the
nature of any interrelationship. Your aim will be to build these as your study
progresses.
Your initial interviews give you some pointers. From here you will need to
make decisions about the focus for the case study? Will it be:
YOUR APPROACH 147
• the station
• a group of people in a station
• one person who has expressed a fear of a terrorist incident?
Whichever you choose, your focus must be the site of something that is of
interest. Remember again Wieviorka’s (1992: 160) comment that a case ‘is
significant only if an observer … can refer it to an analytical category’. It can-
not simply be plucked out of the air. You cannot simply think, ‘I’ll do a case
study on this station’, with no special reason for choosing that station. If you
do choose to focus on a station, then there must be a reason for focusing
on the ‘station-ness’ of this particular station. So, for the case study, there
are two issues:
• Why do a case study on a station, in the context of fear about terrorism?

• Why do a case study on this particular station?
What do I mean by the ‘station-ness’ of a station? Well, if you were to do any

other kind of research, you might choose this station because it is in some
way typical of a station. You would choose to select participants at this station
because it in some way represents stations in central London. Alternatively,
you might choose a sample of similar stations or representative groups of
contrasting stations. Not so with a case study. With a case study you would
choose this station to look at the station environment in the context of poten-
tial fear. So you would look at a whole range of matters. You would look at the
number of people on the platform and how they enter and leave. You would
look at jostling and the distance people stand from each other. You would
look at the height of the roof, at the noise made when the train enters and
leaves the station. Do the advertising hoardings make people feel uncom-
fortable or put them at their ease?
So, to choose a station as a focus for your case study, you would have to
have a reason for doing so, not just because it is a convenient site for your
study. There would be no point in simply choosing a station as a place to
interview people since you could just as easily interview them anywhere else.
Why would this make a good case study?
This raises an issue as far as building a theory is concerned, since here you
are supposed to be starting from scratch. You are, in a sense, expected to
clear your mind of expectations and preconceived notions. Is this possible?
As I have already noted, probably not, but this is the expectation if you are
building a theory.
(Continued)
How, then, can you proceed? The answer is to make a broad assumption
at the outset about the importance of your chosen focus – about the station
and its ‘station-ness’ – then follow the usual steps that I have been talking
about until now, getting as many different sides of the situation as far as this
focus is concerned. These might include starting points about noise, prox-
imity of people, movement, places to sit, presence of concealed entrances,
presence of bins, and so on. You can then find ways to explore these issues
and let the information that you discover speak for itself. You do not go
into a theory-building study expecting to confirm the significance of these
features of the situation – rather, you explore their importance.
So, our theory-building case study might be called ‘The station: a site of fear
about global terrorism’. The weight is immediately being thrown on the station
as the core of your inquiry. You are making it clear in the first two words that it
is the station that is the focus of your interest. It is this that provides the kernel
for your inquiry into terror and why people should feel it. Does the station have
any features that exacerbate (or perhaps lessen) people’s fears? Your methods
need to be geared towards finding an answer to this basic question; then the
theory will grow around the data you collect as a result of your inquiries.
Proximity of Places to sit

people Presence of
rubbish bins
Presence of
Noise concealed
entrances
Fear in stations
Figure 8.2 Your first storyboard for building your theory
I like to think of theory growing, becoming thicker, around productive ideas,

while it stays thin or atrophies away completely around the starting points
that seem to be going nowhere. So, you might start off with a storyboard like
the one in Figure 8.2, which maps out your basic ideas for building explana-
tions for fear in stations. When you have collected your data – completed
your interviews, taken photographs, made videos, taken sound recordings,
and so on – you will find that the evidence seems to grow ‘thicker’ around
certain themes, so that you are able to redraw your storyboard, showing the
YOUR APPROACH 149
themes that are developing, the theory building, as in Figure 8.3. Here, in
this redrawn figure, you can see that the explanation, the theory, is building
around the people who are present in stations, rather than the physical struc-
tures that are there (such as bins and seats).
Proximity of
people
Presence of
bins
Noise Places to sit
Presence of
concealed Fear in stations
entrances
Figure 8.3 Your redrawn storyboard with the beginnings of a ‘theory’
So, in the redrawn storyboard, I have moved the boxes around so that
the important features are closer together and represented by heavier lines.
There are, moreover, connections between the boxes.
Testing a theory
The assumption here is that there is already some sort of explanatory framework
available for the phenomenon or situation on which you are focusing. Your case
study, then, is being undertaken to test this explanatory framework, this ‘theory’.
Infection control: Testing a theory with

case study
Imagine that you are a senior nurse. As a ‘clinical specialist’ you are an
expert in one area of clinical management, so let’s imagine that your
specialty is infection control and that, in your hospital, infection from the
(Continued)
COVID-19 virus has been particularly high. You want to examine why this
might be the case.
There will already be a great deal of existing ‘theory’ to explain why lev-
els of the COVID-19 infection are higher in particular places. It is known,
for example, that rates are higher in hospitals that have more vulnerable
patients and undertake more invasive specialist care, and lower in paediatric
units. It is known also that the frequency of handwashing by staff is nega-
tively correlated with rates of other infection – such as MRSA infection – that
is, the more handwashing, the better. Furthermore, figures can be affected
by the number of patients transferring from one hospital to another, and
between care homes and hospitals.
All of this pre-existing knowledge and theory provides a substantial back-
drop for the construction of your own theoretical formulation about the high
level of infection by COVID-19 in your hospital. In fact, it is important to stress
that all of this contextual information is essential to the conduct of a case
study and the ‘shape’ of your own emerging theory. The fact that you are
doing a case study does not absolve you from the task of setting the study
in context and finding all that there is to know about this particular situation.
In doing your case study, your first port of call will be this pre-existing
theory. Does your hospital fall into one of the categories which make it par-
ticularly susceptible to infection (for example, more vulnerable patients,
such as those who are elderly)? If not, you can move on and assume that
there are some other features of the situation that are causing your high
rates of infection, and it is these features that may form the basis for your
theory-testing case study.
So, the theory that you test will be based on existing knowledge and your
own knowledge of the situation. As I noted earlier, all research starts with a
hunch, and your own hunch will be based on some key observations that you
have made yourself as a member of the hospital staff. Imagine that you have
heard comments such as:
• ‘Everyone knows it’s there and everyone in the know realises that we
can’t get rid of Covid.’
• ‘The politicians, public and media are ignorant – we know that Covid
can’t be got rid of.’
• ‘It’s about more than washing your hands.’
• ‘It’s an excuse, a smokescreen, for too little money being spent on the
health service on better PPE equipment, better cleaning. They are trying
to blame frontline service providers for lack of equipment and shortcuts
on cleaning.’
YOUR APPROACH 151
Let’s assume that your experience includes working in several hospitals and,
alongside these comments, your casual observations suggest that there
seems to be less concern about handwashing in this one. You may feel that
there is a culture of complacency about this aspect of hygiene, which goes
right the way through the system – from the cleaners and ward orderlies
right up to senior management.
At the heart of your ‘theory’ here is an implied comparison with other
places and their culture of hygiene, but you have nothing on which to base
this comparison other than your own experience. To test this comparison
out empirically, you would have to visit other hospitals and collect data to
compare with data from your own place of employment. This might involve
a comparative case study, but time precludes this.
Initially, you might like to look at the kinds of features of this situation
that constitute the culture in general, and a culture of complacency about
hygiene in particular.
The first question that you need to ask here is about the supposed culture
of complacency and resistance. Your focus would be on the presence of this
culture in a wider culture – a highly visible and highly audible one – among
politicians, health service managers, the media and the public about the
importance of hygiene. No one could claim not to have heard about the
need for not just hygiene but much better hygiene – all the time, the mes-
sage is coming through loud and clear.
What kind of narrative could be built around this mismatch between
knowledge and action? In other words, what can we infer from the fact that
there might be a culture of complacency in an environment of excellent
knowledge about the danger of this superbug? What does it tell us?
Surely, it says that there is something else going on in the social environ-
ment of the hospital that is subverting the message about the danger of lack
of hygiene. No one could claim that the message is unimportant, so, if it is
being undermined, what is going on?
Your theory – the theory you want to test – is that the ‘undermining factor’
concerns a resentment about management and new initiatives. This resent-
ment is finding its expression in an oppositional attitude to innovation in this
situation. The storyboard for this might be as shown in Figure 8.4.
For the theory testing that you are doing in this case study, then, you
have made a number of assumptions based on the literature and your own
knowledge, with a twist of your own analysis. From here, you can go on to
test the theory in a number of different settings, as shown in Figure 8.4 –
that is, by discussing the matter with staff, watching people’s behaviour,
(Continued)
listening to them talking and hearing the official and unofficial messages
that are articulated during committee, working group and other meetings.
Recording
Talking to Observing Listening to comments
staff behaviour conversations made in
meetings
Testing the theory
Theory: an undermining
factor – resentment about Internal
External management contextual
contextual information
information
Why is COVID-19 infection so

high in this hospital?
Figure 8.4 Testing the theory
The theory you are testing, in other words, has emerged, has ‘thickened
around’, an idea that you had at the outset, which has been reinforced as
you have looked into the issue further.
Drawing a picture – illustrative–

demonstrative
Some case studies aim first and foremost to illustrate a phenomenon. In the same
way that an illustration in a book brings the text to life, so an illustrative case study
makes a topic more real for the reader. The subject comes to life.
Clearly, there is more involved than simply drawing a picture, but the picture
analogy serves well, for it shows the possibility of making a major difference under-
standable. The point is made well, I think, in the line drawings here of a cat and
dog in Figure 8.5. They are easily identifiable as such, but it would be difficult to put
your finger on why the dog is ‘doggyish’ and the cat is ‘kittyish’.
YOUR APPROACH 153
Is it the tongue sticking out on the dog?

My cat, Archie, sometimes leaves his tongue
sticking out. It makes him look silly, but it
doesn’t make him look like a dog. Why does
the cat look like a cat? Is it the pointy ears? It
can’t be – many dogs have pointed ears.
We know from our experience of the
world (encapsulated in the word ‘phrone-
sis’, which we looked at in Chapter 4) what
these domestic animals look like. We have
some kind of model in our brains that ena-
bles us to say ‘that’s a dog’ when our eyes Figure 8.5 Pictures say more than
words alone
present it with all of the squiggly lines in
the dog picture and ‘that’s a cat’ when our
optic nerves deliver the cat squiggles. That this is so does not bear deconstructing. I
recently learned that the best artificial intelligence programs are still highly unreli-
able when it comes to establishing this kind of distinction. Yet our brains can do it.
We have some kind of amazing picture-making facility in our brains that lets us fit
a new picture with our existing perceptions – here, the cat and the dog. It works for
more than just cats and dogs, too. It is ubiquitous in our sensemaking apparatus as
the pictures, literal and metaphorical – about cats and dogs, organisations, notions
like democracy – are there all the time in our daily sensemaking. A case study, by
presenting those ‘pictures’, enables the connection to be made. It enables readers to
share an experience and, drawing from their own models (their own reservoirs of
knowledge and experience), to make sense of the new image that is offered.
It is not simply a question of providing illustrations, though. There has to be
analysis accompanying it. The great sociologist Erving Goffman (1956), in his
book The Presentation of Self in Everyday Life, unpacks one illustration after another
involving analyses of the ways in which people behave in social life. He notes from
these illustrations that people behave as if they are on a stage, trying to convey
something with their outward behaviour (see also the discussion of Goffman and
dramaturgy on page 254).
Goffman describes a man walking down the street who trips on a loose kerb-
stone. We already have the picture in our heads, involving not just another
person but also ourselves. The man tripping, observes Goffman, will invariably
turn round to look at the offending stone. He does this, Goffman says, not
out of any deep interest in the kind of phenomena that trip people up, but in
order to convey something to any onlooker. The look has a meaning. With the
turn and the look, the man is saying to any onlooker who might have seen
him trip and humiliate himself, ‘I am not the

The case study enables you to offer
metaphors by which you can ‘get sort of dummy who is always tripping over
inside’ a problem, thinking about it things. Nor am I drunk. What kind of extraor-
and connecting with the characters
dinarily disguised danger could have led me
of the story. It enables readers and
inquirers to share experience, using to stumble?’ The turn and the look have sig-
their own reserves of knowledge and nificance and meaning. By presenting the
experience to make sense.
picture, Goffman lets us share his picture and
his analysis.
In his book, Goffman provides a series of annotated illustrations – each one
unique – performing work beyond that of mere illustration. Goffman works on
the illustrations and asks himself what they are saying. It is the reflexive and self-
questioning attitude that the well-chosen illustrations bring that is of particular
service. Flyvbjerg (2006: 223) offers a good account of what is going on:
The highest levels in the learning process, that is, virtuosity and true
expertise, are reached only via a person’s own experiences as practi-
tioner of the relevant skills. Therefore, beyond using the case method
and other experiential methods for teaching, the best that teachers can
do for students in professional programs is to help them achieve real
practical experience; for example, via placement arrangements, intern-
ships, summer jobs, and the like.
Importantly, Flyvbjerg connects inquiry and its illustrative capabilities with

practical learning here. The case study is not a proxy, an alternative for real expe-
rience, but incorporates its ingredients. It illustrates and provides metaphors by
which the learner can ‘get inside’ the problem, thinking about it and empathising
with the characters of the story being told. It enables readers or inquirers to share
the experience, using their own reserves of knowledge and experience to make
sense of its structure and its lineaments. It offers everything that is beneficial in
the ‘show, don’t tell’ that Sennett (2009) promotes (I discuss Sennett’s ideas further
in Chapter 9).
I can give an example from some research I did with some newly qualified den-
tists who were required to conduct a case study as part of their first year in practice.
During their first year, dentists are carefully monitored by both the senior prac-
titioners in the dental practice that employs them and the dental faculty of the
university covering the geographical region in which they work. As part of this
postgraduate professional development, they attend university one day a week and
are supervised in their own dental practice by a mentor (one of the practice den-
tists), who, in turn, has recourse to one of the university tutors if necessary.
YOUR APPROACH 155
The newly qualified dentists thus have to engage in a whole range of further
training and supervised practice. Each element of this postgraduate work has to
demonstrate that important activities have been covered and shows how profes-
sional development will continue so that the dentists in question can prove that
they are able to keep up to date as their careers progress. The case study has to illus-
trate an aspect of this professional development.
The case study is a form of inquiry that is especially well suited to these needs at
this stage in the dentists’ careers. They have all done the important basics in their
undergraduate work at university and shown their competence in these basics. As
they start real work on real people, their needs become more individual, with a
more individually tailored programme of postgraduate work necessary for each.
So, a starting point was that each case study in the year’s cohort of young den-
tists was always likely to be different from the next one: there is idiosyncrasy in the
work of each one. The study could take one of many forms – the use of a particular
piece of equipment, a week in the working life of the practice receptionist, the
use of X-ray equipment by the practice, a particular patient’s experience, and so
on. The only proviso was that the case study should in some way exemplify and
demonstrate the learning that had happened on the course in the postgraduate pro-
fessional year. Clearly not everything could be covered, so it was specified that the
case study should focus on and illustrate two of the four main areas covered in the
postgraduate work – areas such as business skills, clinical skills, patient psychology
and control of infection. It may not be Goffmanesque in the beauty of its analysis
or in the depth of its insight, but this illustrative case study did lead to better under-
standing and development of the dental practice.
A patient’s experience: An illustrative–

demonstrative case study
Nav was one of the cohort of newly qualified dentists. He and his field tutor
decided that, in his circumstances, the case study would focus on one new
patient’s experience following her acceptance by the practice. The pur-
pose of the case study would be to concentrate especially on two aspects
of professional practice, as outlined by the expectations of the course.
These two aspects would be business skills and patient psychology.
(Continued)
Nav first had to make absolutely sure that his case study of his patient
‘Hannah’ was ethical, with Hannah giving her informed consent to the use
of the information relating to her case (this is discussed more fully on pages
85–103). He planned for the case study both to demonstrate knowledge that
had been acquired and to illustrate how this knowledge had been used with
Hannah. To draw the maximum benefit from this (and to gain the highest
marks) Nav would attempt to draw out problematic aspects of the case to
exemplify how the problem had been overcome or indicate where there was
a systemic problem – that is to say, a problem to do with the system that had
been set up in the practice.
Nav started by looking for ways in which he could gain purchase on
Hannah’s experience. Available to him were the practice records of appoint-
ments, letters and emails and his own records as a new dentist. He planned
to supplement these records with two sets of interviews – one with Hannah
herself and another with his mentor, Jayne, who happened to be a senior
partner in the practice. As an important part of patient communication, Nav
also planned to look at the practice leaflets and other publicity and how
these were dovetailed (or were not) with phone calls and emails. With all
of these Nav hoped to secure leverage on Hannah’s experience and ‘drill
down’ (no pun intended) into this. With the information, he would be able
not only to illustrate his own connection with and understanding of the prac-
tice, but also, if possible, to offer signposts for improvement in the practice’s
work and his own.
Nav started by drawing a storyboard, as shown in Figure 8.6. On the
left-hand side of the storyboard are some points Nav planned to address
concerning his business planning module, and on the right are some fea-
tures relating to patient psychology.
Given that he was new to the practice and Hannah was also new, he felt it
appropriate to look at the practice’s ‘front of house’ habits, as well as its meth-
ods for acquiring new patients and finding out about their needs. Much of this
questioning could be done in a structured interview. Richer information – for
example, about the demeanour of staff and their willingness to help – could
be gathered in an unstructured rather than a structured interview.
On the right-hand side of the diagram are the issues to do with patient
psychology and whether or not the practice adequately handles this, offering
help and support where necessary, both in the person-to-person contacts
that happen immediately on coming through the door and in the general
ambience of the place. Are there flowers around? Are there comforting pic-
tures (rather than posters of molars, exposed nerves and adverts for teeth
whitening)? Are there large notices about the dire consequences of not turn-
ing up to appointments? Is Mozart playing calmingly in the background or
YOUR APPROACH 157
can you hear the mosquito-like whine of drills behind doors? That kind of
thing.
For both sides of the diagram, the case study is an ideal form of inquiry,
for it provides, if it is done properly, the kinds of data that will not be gained
from other kinds of inquiry. It provides:
• a useful means of addressing the needs of the professional develop-

ment activity that Nav is undertaking
• a source of rich and valuable information, telling of a patient’s experi-
ence at the practice in depth and providing temporal continuity (how it
was over time) and enabling a connection between different sides of the
phenomenon under study (note the dashed line in Figure 8.6)
What do our rooms do to

Ability to speak to a dentist calm our patients? Are
or is reception like
Patient psychology
they homely or clinical?

gatekeeper?
Reception
Friendliness of staff Did I offer Hannah any

calming measures?
st appointment Is there training for

receptionists in
Hannah’s dealing with
First acceptance frightened patients?
experience at the
practice
Figure 8.6 Storyboard of Nav’s illustrative case study
• a source of readily understandable feedback so people will be able

to use the information to improve the service – the case study is on a
‘human scale’ and, for those working with people, provides access to
the kinds of complex emotions and their relationship with each other
that other kinds of research often fail to communicate.
A much simpler example of an illustrative case study is that used in medical

training or the reporting of unusual medical occurrences. An unusual feature may
provide a stimulus for further research or an instructive example for the training of
medical staff. Here is a brief example of the latter.
Medical traning: An illustrative–demonstrative

case study
Mrs Smith is a 58-year-old woman who fell and broke her hip. She was admit-
ted to hospital and given the usual treatment for this, which is an operation
to install something called a dynamic hip screw. This is, essentially, a long
metal bolt that has been bent so that the top part can be screwed into the
top of the hip (the ball part), while the end of the bolt is, in turn, screwed into
the top of the femur (the long leg bone) with three, four or five screws. The
dynamic hip screw thus holds together the broken bone at the hip. Because
it is made to be ‘squashable’ (technical term), the dynamic hip screw allows
the weight of the person’s body, when standing (which is encouraged almost
straight away after the operation), to compress the fracture, while holding
the whole arrangement in place. The compression aids healing, and the
dynamic hip screw has been shown to be a very effective means of mending
a broken hip or a top-of-the-femur fracture.
While it is very effective in most circumstances, there are times when it is
not the best treatment, which include when the hip is broken high up, beyond
the neck of the femur. Mrs Smith had such a fracture, yet the dynamic hip
screw was still used. The problem in such cases, where the fracture occurs
just below the ball of the ‘ball and socket’ of the hip joint, is that the ‘ball’
is separated from the main part of the femur, with the result that the blood
supply to the ball is disrupted, making healing less likely.
Unfortunately the blood supply was indeed disrupted in Mrs Smith’s case.
After 10 months, she reported to her doctor that she was experiencing pain
in her hip. What had happened was that the head of the hip had collapsed
due to lack of blood (or ‘avascularity’, as doctors like to call it) and the end
of the screw had pushed through the ball, damaging the inside part of the
pelvis – the concave part, known as the acetabulum, into which the ball fits.
Because of this, Mrs Smith suffered a great deal of pain and had to have
another operation to fix the problems arising from the initial decision to
install a dynamic hip screw. The second operation involved something called
a hemiarthroplasty – half hip replacement, in which just the ball part of the
joint is replaced, not the socket (in a total hip replacement, the socket is also
replaced). With hindsight, a hemiarthroplasty would have resulted in a much
more favourable outcome.
This case study illustrates why it is necessary to take into account the location of
the fracture when decisions are made about the nature of the remedial action that
should be taken to fix the problem. This is, if you like, the analytic frame or object
YOUR APPROACH 159
(see pages 15–19). It demonstrates what can go wrong and why, as well as being
instructive for practitioners, bringing a theoretical set of issues to life.
In medicine, illustrative case studies can also be used to reveal a new or unfamil-
iar set of issues, a new syndrome or a promising new technique, yet to be trialled
for its efficacy with a larger number of people. With any kind of inquiry, it may also
be added to a larger piece of research to exemplify a theme or illustrate an aspect
of the analysis.
Case studies are also often used in the teaching or explication of the law. As
in medicine, they also tend to be illustrative. A good website for viewing some
US legal cases is Stanford Law School’s Case Studies Collection (available at: www.
law.stanford.edu/organizations/programs-and-centers/environmental-and-natural-
resources-law-policy-program-enrlp/case-studies).
Interpretative
Often, when a case study is written about in the academic literature, what is being
discussed is an interpretative case study. This is, if you like, the ‘classic’ approach to
doing a case study. This is not to deny the importance and significance of the other
approaches I have talked about so far, but the aims of a case study and the style of
interpretative inquiry dovetail together very nicely. For this reason, I shall go into
this approach in rather more detail than the previous ones.
Interpretative inquiry is a form of inquiry that employs a particular approach to
answering questions – an approach that assumes an in-depth understanding and
deep immersion in the environment of the subject. You can probably see why,
then, I have said that the case study and interpretative inquiry are natural bedfel-
lows, since each calls for rich, intensive understanding. I should go further: they are
not only natural bedfellows but also obvious marriage partners. They were made for
each other: it is love and marriage – and they go together like a horse and carriage.
Each demands a deep understanding of the multifaceted nature of social situations,
so they complement each other and seem natural with each other.
This kind of approach is often called ethnographic, so a case study using such an
approach may be called an ethnography. The word ‘ethnography’ is from the Greek
ethnos, meaning folk or people, so ethnography literally means ‘study of folk’. It
was used first in a field of social research that emerged in the early part of the twen-
tieth century as a branch of anthropology.
Before the new ethnographers, anthropologists had done their work by treating
people as objects of study, as dispassionately as a biologist might study an insect.
Displacing this style of work, the new ethnography aimed to get to the heart of
people’s understandings of life by doing fieldwork with them rather than suppos-
edly objective study of them. One could, interestingly in relation to the case study,
say that they aimed to reach what is summed up by the Latin phrase casus anima,
which roughly translates as ‘the soul of the case’ or ‘the heart of the matter’; they
aimed to reach an understanding of what makes the person or people tick.
So the ethnographers aimed to get right into the centre of the cultures with
which they worked by becoming members of those cultures. They would actually
live with the people and try to understand their culture from within. It was as if the
earlier anthropologists were taking callipers and rulers to measure the people they
were looking at, while the ethnographers, by contrast, thought that the best tool
for the study of other human beings was themselves, as fellow human beings.
The ethnographers, therefore – people such as Bronisław Malinowski, who in
1914 went to live among the Trobriand Islanders of the Western Pacific – suggested
that, instead of trying to distance yourself from your subjects of study, you should,
as a social scientist, get as close as possible by becoming a participant in their cul-
tures. Try to understand them as a person yourself, by being a participant observer.
Forget about any pretence at objectiveness; do not deny that you are a person,
pretending that you can see things objectively. This is never possible, said the eth-
nographers. Not only is it impossible, but it simultaneously throws away one of
your ready-made strengths as an inquirer among people. That strength is you –
yourself – which has to be your watchword in doing the kind of case study that I
am talking about here.
The ethnographers influenced, and in turn were influenced by, the thought of a
group of social scientists in the 1920s and 1930s who said that the world in which
we are interested as social scientists is not straightforwardly perceivable in the way
that the world of chemicals is perceived by chemists. We cannot view it like this
because it is constructed by each of us in a different way. It is not simply ‘out there’;
it is different for each of us. It cannot therefore be adequately studied using the
methods of the natural scientists (such as physicists and chemists), with talk of
variables and quantification – a wholly different mindset and set of procedures is
needed for inquiring into it. This view came to be called interpretivism.
This approach to social study caught on like wildfire and spawned a range
of fascinating, in-depth ethnographies in the middle of the twentieth century,
such as Street Corner Society by William Foote Whyte (1955/1993). For more than
three years, Whyte lived in a slum district of Boston, populated largely by Italian
immigrants. He lived among and with the people and documented how young
men became either ‘corner boys’ or ‘college boys’. It is interesting to look at the
way in which Whyte reflected on theory in doing this, because it might be said
that this kind of interpretative case study is inherently about building a theory.
YOUR APPROACH 161
By interpreting people’s words and behaviour, the ethnographer is building the-

ory out of the naked, raw data that are available.
At the time of collection, of data gathering, these data have no ‘theory’: it is
rough and crude, without shape or form. What the interpreter has to do is build a
theory and an object from it. Whyte (1985: 21), in his own reflection on the pro-
cess, is quite guarded about this building of theory or ‘theorisation’, putting it this
way: ‘what kind of theory can one develop – or to what existing body of theory can
one contribute? I have found myself focusing on the concrete behavior of individu-
als, groups, and organizations’ (emphasis added).
By talking about ‘concrete behaviour’, Whyte is drawing a distinction between this
and theory, for the latter may seem removed from the real, practical world. He clari-
fies by going on to say that he seeks a ‘conceptual framework that will strengthen my
power of understanding and interpreting behavior and organizations across a wide
range of social situations’ (1985: 21).
It is important to keep this ‘framework’
Interpretative researchers assume
notion in mind when thinking about the that the social world is indivisible.
theory that you develop as part of an interpre- It is complex and we should study
it in its completeness. In this sense,
tative inquiry. This ‘theory’ is something that interpretative research marries easily
is used for the purposes of your study. It is not with case study, which also prioritises
looking at the whole.
an immovable and immutable thing that you
are establishing, forever to be embedded in the
canon of social science inquiry. Rather, it is what the sociologist Pierre Bourdieu
(in Wacquant, 1989, cited in Jenkins, 1992: 67 (emphasis in original)), called a
‘thinking tool’. As he put it:
There is no doubt a theory in my work, or, better, a set of thinking tools

visible through the results they yield, but it is not built as such … It is a
temporary construct which takes shape for and by empirical work.
So, the kind of theory that we are looking for is more a temporary ‘conceptual frame-
work’ or ‘thinking tool’ than it is an abstraction in its own right and for its own
purpose. It should not be an intellectual barricade, to be defended against all-comers.
Rather, it should help you to think about and understand the subject in hand. It can
be held, tested and then discarded or retained depending on its usefulness.
It is important to remember that some social scientists, in the bad old days,
wished to develop what has come to be called ‘grand theory’. In the words of the
science fiction comedy The Hitchhiker’s Guide to the Galaxy, these are theories of
‘life, the universe and everything’. It is now generally accepted that that kind of
theory is unattainable and it is not useful to try and establish it. In fact, the attempt
to establish it has probably done social science a good deal of harm, as the grand
theory that is set up is often wildly misleading.
I can give as an example of an interpretative case study a piece of work that I
conducted while I was a parent governor of the secondary school that my daughters
attended.
School governorship: An interpretative

case study
I was able to be a participant in the situation of being a governor (since I
actually was one) and didn’t feel the need to try and stand separately from it,
observing from outside. I was a governor, feeling what a governor felt, see-
ing what a governor saw and hearing what a governor heard.
I recorded my reflections on this process in a diary that I kept over a
period of several months in the role. Although I also made observations and
conducted interviews, it was this diary that was the principal instrument of
my data gathering. Here is a sample of my diary, to give a flavour of the kinds
of reflections that I was making and the use I made of them in subsequent
‘theorisation’.
The Chair pushes us on to the question of minutes – how they should

be written and how they should be presented to the full governors’
meeting. I think this has been raised because Walter [one of the gover-
nors] insists at each full meeting on reading out his minutes in full and
adding long, drawn-out embellishments about things like the lintel
over the door in the boys’ toilets needing replacing. It’s extraordinary
that he never seems to detect any humour in the absurdity of the trivia
he raises. Chair presents a solution to the problem of needing to get
through the subcommittee reports more quickly. He suggests that we
have items reported in the minutes that have either one star or two
stars, relating to whether they need a full discussion at the full meet-
ing or are simply reported on. Walter seems to get the wrong end of
the stick and keeps trying to find examples of one-star and two-star
items, delivering them portentously to the group as though he’s just
discovered the meaning of life. Chair looks at him a little limply as he
does this, clearly not sure how to respond: whether to put him straight,
tell him to shut up or simply let him carry on until he stops. He decides
on the last of these options and we’re all therefore forced to listen to
Walter droning on with more examples of one-star and two-star items
delivered at about one word every five seconds. He obviously thinks
YOUR APPROACH 163
that the slower he speaks, the more gravitas he has. Chairing for bor-
ing, inconsequential, dull or plain stupid people is clearly a problem
in this kind of forum (or in most kinds of forum). The trouble is that
we’re all such nice people, we don’t want to offend the feelings of
others – and this is especially so when we are fully aware in this kind of
setting (that is, as governors) that we’re all working voluntarily and out
of the goodness of our hearts. You don’t want to discourage people
or seem to be denying or downplaying the quality or value of their
contribution. It’s doubly a problem, though, when the body is taking on
quasi-management roles and there is the need to be a little more disci-
plined about the machinery that makes things happen. If that discipline
doesn’t exist (or exists badly) there’s the possibility the machinery will
become completely gummed up.
How can this be turned into interpretation and theory? In other words, how
can an analytical frame be established? It is important to note that the diary
I was keeping was not simply ‘this happened … then that happened … then
this happened … and so on’. Rather, it was my reflections on the topic in
hand, with a critical edge to the commentary. It is not, as the great anthro-
pologist Clifford Geertz (1975: 9) put it, as if I was merely a ‘cipher clerk’ in
making these recordings. Although you cannot perhaps appreciate it from
this diary entry on its own, my attempt in doing this case study was to look
at the ability of governors to fulfil a role that the government at the time
had set them, which was to act almost as managerial overseers – with a slice
of ‘voice of the people’ twisted in. In this particular diary entry, I was saying
something about the ability of school governors to fulfil this role. The key
idea coming through in this entry is of amateurism, of not being able to fulfil
this role. This can perhaps be seen better in the next excerpt from the diary:
The premises subcommittee report is another matter. The report is

prefaced by Walter saying that he will be extremely brief – and obvi-
ously he intends to be and starts talking quite fast (for him). This only
lasts about 15 seconds, though, and eventually he’s back on form,
talking in his slow overly loud drawl – he obviously thinks that public
people should talk like this, as though they are addressing a council
chamber. Citizen Kane syndrome. We spiral down into the depths of
the lintel over G31 and worse. There is some discussion of the costs
of allowing the free use of the hall and gym by the community and
Walter produces some figures that purport to show how much it is all
costing us. Another member questions the figures, saying that Walter
(Continued)
has completely miscalculated. He’s quite right – the calculations are

based on the cost of heating and lighting for the whole year rather
than the very limited times of community use. This sends Walter into a
paroxysm of waffle and as usual there is no decision made.
It isn’t just that these are well-meaning amateurs dabbling and mak-
ing ill-informed decisions. It isn’t simply a random collection of
members of the community. People join the governing body for a
variety of reasons – sometimes because they want to try to improve
things since their own children are in the school and no doubt for
other good reasons, but sometimes it’s because they are self-impor-
tant and pompous and, wanting to get into public life, they end up on
governing bodies. It’s these people who can do the damage.
There are some blatant clues here as to the constructs that I was building.
For a start, I use the phrase ‘well-meaning amateurs’, which is hardly a covert
or disguised expression of my views. It needs little interpretation. There are,
however, more interesting comments about the nature of the enterprise in
which governors are occupied.
First, I make a comment about the sort of language and kind of deliv-
ery that Walter is using – it is of an innocent amateur playing at being a
professional. Then, there is his error in calculating the costs of heating and
lighting. The error is serious, not only for the consequences it might have
had on policy (excluding community use of the school facilities) but also for
the effects it had on the work of senior management in the school. My sub-
sequent thinking – my ‘theorisation’ – about the diary commentary was that
government policy, by mistrusting professionals, had produced a ‘double
whammy’. It had not only displaced the professionals from their proper role
but also wasted the professionals’ time as they then had to intervene and
sort out the mess that the amateurs like Walter were creating.
A stronger narrative around trust, professionalism and the audit culture
could be made of this. It could even be linked with an idea such as that
of Richard Sennett’s (2009) in his book The Craftsman, where he suggests
that much of the lack of purpose, the anomie, of modern life is about the
dismissal of the craft of the tradesperson and the professional – whether the
craft is that of an architect, teacher, X-ray technician or zookeeper. Too many
of today’s crafts are made insipid and trivial, argues Sennett, not only as a
result of the intervention of machines and computers but also, even more
importantly, because of the procedures that remove the care, skill, pride and
dedication from our craft.
Using this insight, I could – for theorising my argument – ‘step sideways’
and look at what enfeebles and belittles the craft of teachers, removes their
YOUR APPROACH 165
professional judgement, their trust in themselves to be making the right

decision. I could try to show in this case study how the simple policy decision
taken at government level filters down and expresses itself on the ground,
in the staffroom and the governing body. Thus, a naïve feeling about woolly-
minded, progressive, liberal teachers from a politician travels from cabinet
room to civil servants’ meeting, to regulations implemented at local authority
level, to the meeting of a governing body where a pompous and unquali-
fied governor behaves naïvely and endangers community involvement and
wastes professionals’ time. Thus, a nice circle could be drawn showing the
connections from initial high-level naïvety to ultimate ground-level naïvety.
The case study in this circumstance could be either building or testing a theory,
depending on how I had framed it at the beginning, but the important point, as
far as this section is concerned, is that interpretation is being used. The germ of the
idea at the outset could be about policy decisions and their journey from idea (the
idea of the politician or whoever) to action. You collect data about this journey and
make interpretations about what is happening along the way.
Experimental
When people talk about case studies, the experiment is probably the last thing that
comes into their minds. We assume that a case study is about singularity and depth,
whereas an experiment is about large numbers, control and comparing this group
with that. This assumption, however, while partly correct when talking of experi-
ments, is only partly so.
Certainly, an experiment is a particular kind of research design where ideas are
being tested under controlled conditions, but, in everyday language, ‘to experiment’
has a much looser meaning. It can just mean to try something out, and so ‘an experi-
ment’ could be simply a little trial of some kind. So, I might say ‘I’m going to do a little
experiment’ if I choose one day to try putting my milk in my tea before rather than
after the hot water to see if the tea tastes better the new way. (I did try it and it doesn’t.)
When approaching certain kinds of questions for case studies, we can take the
systematicity of the experiment and graft it onto our expectations of a case study.
This requires a particular kind of experiment, though. Before going into this – into
the kind of experiment that might be used in case study – let me say a little more
about experiments and how we can think about them.
In the natural sciences (chemistry and physics, for example), to experiment

means something more precise than it does in the vernacular. It means to test an
idea under controlled conditions to prove or falsify an idea, a conjecture, a hypoth-
esis. Robert Hooke in 1676 had an idea about elasticity in springs and he tested
this idea systematically under controlled conditions, stretching the springs with
weights and recording the consequences. He was able to emerge from his experi-
ments with what became known as Hooke’s Law, which says that how far a spring
extends is in direct proportion to the load added to it.
Ordinarily in the social sciences, experiments are rather different from Hooke’s in
the natural sciences. Usually, social scientists are trying to establish whether or not
something causes something else to happen. Does X cause Y? Does the introduc-
tion of a new science curriculum cause an improvement in students’ understanding
of science, for example? The only way to find out with any sort of validity in social
science is to do an experiment.
This has to be done using special procedures and having a particular kind of
mindset about the social world that is rather different from that adopted when
doing interpretative research of the kind described in the previous section. This
mindset is that we can look at this world as comprising a set of variables (that is,
things which vary), such as age, sex, amount of pocket money, income, credit his-
tory, educational background and so on.
This may not seem too contentious, for we have all become familiar with the
methods of social science and we have all got used to these assumptions about
variables. We realise that they may be related to one another. We understand that
clusters of variables tend to vary together, so, for example, income is related to edu-
cational background, which is related to educational attainment.
The mere acknowledgement and observa-
Though case study is not normally tion of variables for some kinds of social science
associated with experimental work,
you may choose to include some inquiry is different, though, from actually trying
kind of experiment as part of your to do things with those variables systematically –
study. If so, you will probably choose
a design that enables you to use the
manipulating them – in the way that natural
subject as their own ‘control’ – an scientists do. The essential trouble is that the
‘n = 1’ design. social world comprises such a ‘blooming buzzing
confusion’ – as the nineteenth-century psychol-
ogist William James (1890/1981) put it – that it is sometimes impossible to know
what the relevant variables might be.
Imagine that you are looking at a classroom full of children, a teacher and an
assistant. What are the relevant variables here that you would need to be aware of
if you wanted to observe the classroom under controlled conditions (that is to say,
controlling for the important conditions)? If I tried, I could think of hundreds of
YOUR APPROACH 167
potentially relevant variables: the variation in the ages of the children, the colour of
the walls, the direction of the light, the heating in the room, the presence or not of
a certain naughty boy, the hungriness of the teacher (depending on whether or not
she had to skip her breakfast that morning), the kind of work the children are doing,
the educational background of the teaching assistant, and so on.
We can potentially change any of these things and, in a perfect world, control for all
of the others. The trouble is, though, that we are not in a perfect world and we cannot
control for all of them (or even most of them), and we do not even know which ones we
should be controlling. We can guess, certainly, but there is a strong chance that we may
be wrong. There is also the problem, as the psychologist Jacob Kounin (1970) noted,
that the classroom is like an ecology: you cannot change one aspect of it without some
unexpected consequence. So, if you take out the naughty boy – let’s call him Kyle – you
may find that another member of the class takes over Kyle’s role. Playing with variables
in a social situation is a bit like squeezing a balloon – it will bulge somewhere else.
So a preliminary, cautionary note is needed when we are talking about a manip-
ulation of the variables of the classroom – and an experiment demands such
manipulation. It makes the assumption that all of the variables can be manipulated
and, when we manipulate them, we can look to see how this manipulation has had
an effect on other variables.
In its simplest form, an experiment seeks to isolate two variables – the one that you
assume may cause change (such as the science curriculum) and the one you are assum-
ing may change as a result (the understanding of science, in the example that we have
been using) – from the myriad that might be at play in a situation. Beyond this, there
is a vast superstructure of kinds of experiments that are built on this very simple basic
assumption, but we do not need to go into them here as it would be most unusual, if
not impossible, to find some of these designs incorporated into a case study.
What kinds of experiment, then, can be done in a case study? Well, the case
study has to conform to the basic idea of what a case study is, which is a study of
one thing. Interestingly, research specialists who are clever with numbers have a
special way of saying ‘one thing’, which is ‘n = 1’. I can already feel the adrenaline
beginning to flow into the veins of some of my readers as I include an equals sign
in the text. I know that you might be saying to yourself, ‘I thought by doing a case
study that I was going to be able to avoid equations and statistics.’
Well, you should not think that way. The method you choose should be the serv-
ant of your research question and if your research question demands an inquiry
that uses numbers and simple statistics, it is these you should use. People’s fear of
statistics is often completely groundless, and the kinds of statistics that you need
for this kind of study are actually quite simple. I look briefly at these in How to Do
Your Research Project (Thomas, 2017).
The usual form, the classic experiment (see Table 8.1), in the social science situa-
tion, involves two or more groups being treated in exactly the same way, except for
the manipulation of one variable. So, the procedure is to bring in an extra group,
as alike as possible to the first group, and give them everything that the first group
had so we can eliminate sources of variation between the groups – every source of
variation except the one we are deliberately varying. Any differences that then exist
between the two groups after the manipulation is made to happen by the experi-
menter are taken to exist because of the experimental treatment.
Table 8.1 The classic experiment
Pre-test Treatment Post-test

Experiment group  Take first measure  Give treatment  Take second measure
Control group  Take first measure  Don’t give treatment  Take second measure
There is another form of experiment, however, that is especially appropriate for

case studies since, in this form, you can look experimentally at change within one
situation, such as the members of a classroom. This form is called the repeated meas-
ures design.
Whereas with the classic form of experiment you are comparing the experimen-
tal group with the control group, in the repeated measures (or crossover) design
there is no second group used. Instead, the control comes from the group itself,
with the ‘change’ being imposed by the difference in one of the variables. I will
not go into the technicalities of the design here, since the explanation of how a
repeated measures design is used can be quite complicated and you can find more
details elsewhere (see, for example, Field, 2013). I shall simply focus on how and
why this procedure is appropriate for a case study. You will have noticed that I said
‘the control comes from the group itself’, so we are back to our one thing. The main
condition of the case study is therefore satisfied.
Let me give an example (loosely based on Cremin et al., 2005) of how a case study
might incorporate a little experiment of this kind.
Additional adults in the class: A case study

using experiment
Alyssa is interested in the way that additional adults – teaching assistants
(or aides), parents and volunteers – are working in her classroom because
it seems to her that something is going wrong. The adults are not working
YOUR APPROACH 169
effectively with each other, sometimes getting in the way of each other, and
certainly not communicating effectively.
Her first task is to set down a question and draw a storyboard that might begin
to sketch out some lines of inquiry that could be addressed (see Figure 8.7).
Alyssa has one permanent teaching assistant and two parent volunteers
who come in every morning. The management and organisation of this kind
of assistance were not part of her teacher-training curriculum, so Alyssa has
read the literature and found that there are a number of ways of thinking
about how extra assistance of this kind can be deployed. One method is
called ‘room management’ (Cremin et al., 2005) and involves giving each
adult a set of very specific tasks to do (such as predominantly helping indi-
viduals or groups), those tasks constituting a named role. The roles can be
changed from session to session, but the main thing is that each individual
knows which role he or she is supposed to be fulfilling.
Are they getting Am I not

Other managing
in the way of
problems – them
each other?
status, correctly?
territory?
Lack of
communication?
With each other? Not knowing
With me? what to do?
What is going wrong with the

extra people in my class?
Figure 8.7 Storyboard leading to an n = 1 experiment
Alyssa thinks that this method of organisation holds some promise and
hopes it will provide some structure for her work with other adults in her class.
Rather than simply trying it out and proceeding according to a gut feeling
about whether or not the new system of organisation seems to work, she does
a small experiment. This involves a repeated measures design where the one
group (Alyssa’s class) is examined under two circumstances, or conditions –
the only change being to the system of organisation in her class. In the first
condition Alyssa and the other adults work normally; in the second, the only
change is that they will use the room management system of organisation.
(Continued)
To assess the effect of this change, Alyssa decides to measure the chil-
dren’s engagement with their work – in other words, to look at whether or
not they do what they are supposed to do, so certain behaviours, such as
sitting and writing, will count as ‘on-task’, while others, such as wandering
around the class, will count as ‘off-task’. By drawing up a list of on-task and
off-task behaviours and observing for these systematically, Alyssa can meas-
ure what happens in the class under the two conditions.
The experiment requires organisation that will take many features of
the situation into consideration. Alyssa first has to plan when the observa-
tions will take place under the two conditions. Because the two conditions
must be identical apart from the change in the system of organisation, she
has to ensure that the same kind of work is being done by the children and
the same people are there with her class for both periods of observation.
They also have to be at the same time of day and even, if possible, on the
same days of the week, for these are also sources of variation that could
plausibly make a difference. Then she has to train the assistant and parents
in the room-management system and schedule the sessions. She also has
to arrange for observations to take place to test the effect of the system
on the dependent variable – children’s engagement or ‘on-taskness’. She
decides to organise this by asking another parent who regularly assists in
the classroom to video the sessions and subsequently goes through the
video with a checklist, looking at each child in turn for a set period to work
out an on-task percentage figure for each of them.
Each child’s overall engagement (or ‘on-taskness’) is then plotted for the
two occasions. Using some simple statistics, Alyssa is able to show that the
differences between the two occasions are statistically significant. Simply
‘eyeballing’ the chart (that is, just looking at it intelligently) also shows a
number of interesting things. Because Alyssa ranked the children’s engage-
ment from lowest to highest in the first condition, she can see how those
who are at the lower end of the scale – that is to say, those children who find
most difficulty attending – are affected by the introduction of room man-
agement. The chart clearly shows that, for most of them, their engagement
improves markedly over the two conditions.
Such an experimental approach would not be sufficient, in my opinion, for a

thorough case study. A thorough case study is one in which we are aiming to gain
that polyhedron of understanding mentioned before – looking at a thing from all
sides, from the top and bottom. While most of the examples I have used so far do
indeed, in some shape or form, take such a multifaceted view, the experimental
method on its own does not, because it seeks to enumerate and control variables.
YOUR APPROACH 171
Added to other sources of information, however, a case study of this kind may well
add an invaluable dimension to that polyhedron.

take this …
Beyond a subject and an analytical frame for your case study, you will need to
decide on the kind of approach to take when collecting data and analysing them.
The approach you take will be shaped by the firmness of the ideas you have formu-
lated at the outset:
• If, on the one hand, you start with a firm idea or set of premises, you could be said
to be testing a theory.
• If, on the other hand, you prefer to see what ideas emerge as you immerse yourself
in the situation you are studying, you could be said to be building a theory.
• Alternatively, you may simply be aiming to illustrate something.
Beyond these broad divisions of approaches to case study research, you can pick
and mix to create the approach you want to take. So, you may choose to take:
• an interpretative approach as a participant observer, or

• a more formally experimental approach to some aspect of your study, or
• you may decide to do both.
Remember that the case study is like a wrapper: within it, the shape of the study is
determined by your question and your purposes.
REFLECTIVE ACTIVITY
In Figure 8.4 I indicate how a theory is developed in the case study about infection
in a hospital. Here, the storyboard points to the high infection in this particular
hospital having its roots in staff resentment about top-down management, and
a plethora of new initiatives about which there has been a lack of consultation.
Given what we all know about the spread of infection following the COVID-19 pan-
demic, think of an alternative theory which could be tested by a case study and draw
an alternative storyboard which might offer different routes for testing that theory.
Further Reading
Eisenhardt, K. and Graebner, K.M. (2007) ‘Theory building from cases: opportuni-
ties and challenges’, Academy of Management Journal, 50 (1): 25–32.
This is a highly traditional, but very solid and well-explicated account of theory
and the case study’s contribution to it, written from the point of view of manage-
ment science.
Field, A. (2017) Discovering Statistics Using IBM SPSS (5th edn). London: Sage.
This really tackles research from the point of view of statistics and, if you do an
experiment as part of your case study, it will be invaluable – not least because it is
presented via the main software package used in the social sciences, SPSS. If you
are doing a study of the kind I have suggested in this chapter, look especially for
non-parametric tests, such as the Mann–Whitney test and the Wilcoxon rank-sum
test.
Geertz, C. (1975) The Interpretation of Cultures. London: Hutchinson.

Read this for the discussion of ‘thick description’ (see also Chapter 11 below). Also,
try Googling ‘Geertz deep play notes on the Balinese cockfight’ for an example of
Geertz actually doing anthropology and thick description.
Goffman, E. (1956) The Presentation of Self in Everyday Life. Edinburgh: University of

Edinburgh Press.
The clue is in the title: it is about how we present ourselves … er … in everyday
life. We do so as actors, performing a series of roles on different stages. It is worth
reading even if you do not do a case study.
Luker, K. (2009) Salsa Dancing into the Social Sciences. Cambridge, MA: Harvard
University Press.
Luker discusses well the framing and construction of an interpretative research
project, starting from a hunch to the development of an investigation.
Malinowski, B. (1922) Argonauts of the Western Pacific. London: Routledge & Kegan
Paul.
Malinowski, B. (1982) ‘The diary of an anthropologist’, in R.G. Burgess (ed.), Field

Research: A Sourcebook and Field Manual. London: George Allen & Unwin.
Between 1915 and 1918, Malinowski studied the Trobriand Islanders of the South
Pacific. He broke new ground in the way he went about understanding people as
YOUR APPROACH 173
fellow human beings rather than as objects of study and, in doing this, he helped
to establish the ethnographic tradition. He was a participant observer, keeping a
diary of his observations, and his analyses were pioneering in their use of an inter-
pretative approach.
Spradley, J.P. (1980) Participant Observation. New York: Holt.

This is a beautifully organised account of the process of ethnography from a master
of the genre.
Thomas, G. (2007) Education and Theory: Strangers in Paradigms. Maidenhead: Open

University Press.
Here, I go into some detail (possibly too much) on the meaning of theory and what
it can offer to researchers.
Here Kevin Myers and I give more examples of the kinds of case studies outlined in
this chapter, analysing each of them in terms of the classificatory system used in
this book.
Whyte, W.F. (1985) Learning from the Field: A Guide from Experience. Beverly Hills,
CA: Sage.
This covers just about everything that an interpretative researcher could wish to
know.
9
YOUR PROCESS: THE SHAPE,
STYLE AND MANNER OF
YOUR CASE STUDY
How will you go about structuring your case study? What I am talking about here
is the style and manner of your case study – whether you are doing a case study of
one individual or several and, if the latter, whether they are all done at once or one
after the other, whether you will separate out nested elements of the single case for
special examination or look back at events that have happened in the past or collect
data as time proceeds (or both). We will focus in this chapter on the nuts and bolts
of the study, as summarised in Figure 9.1.
Subject Purpose Approach Proces s

Retrospective
Snapshot
Outlier Intrinsic Testing a theory
Diachronic
Key Instrumental Building a theory Singl e
Local Evaluative Drawing a picture Multiple Nested
Explanatory Experimental Parallel
Exploratory Interpretative
Sequential
Figure 9.1 The process as part of the design route

YOUR PROCESS 175
The single case

The single case could be called the classic form of the case study. When people
think about case studies, they think of one person – a doctor’s patient, for example,
or a child in a school or, perhaps, a classic legal case. The single thing is studied for
the lineaments of its structure, its character, with the emphasis on understanding
what is going on, taking one of the approaches I have already described.
The focus, when choosing the single case study, has to be on the characteristics
that give it some interest. Let’s imagine an example:
Security studies: A single case study

Let’s imagine that you are a political science student doing a master’s degree
in security studies. One of your taught courses has been on terrorism and –
this having interested you – you wish to follow it up by concentrating on ter-
rorism for your dissertation.
You do some reading and thinking about terrorism to try and shape your
ideas, and realise that nearly all terrorism is related to the work of well-
organised groups of varying sizes and forms, each with causes that inspire
loyalty from a significant proportion of some defined population with a per-
ceived grievance of some kind. The groups that manage to commit terrorist
acts successfully do so because they evade detection by the authorities. You
decide to make this the key angle of your inquiry – the lever that you will use
to force open the issue – and your research question is geared towards it.
Set up ‘cells’ that Careful planning with long

operate autonomously lead-in to action
Brutal treatment of those

Loyal following
who oppose them
How do terrorists
escape detection?
Figure 9.2 Terrorism storyboard

(Continued)
You draw your storyboard, which expresses your basic question, ‘How do
terrorists escape detection?’, as shown in Figure 9.2. In it, you sketch out
some ideas to follow up.
So, you have a question, but there are still a number of options open to
you concerning the route that would offer the most illuminating avenues for
tackling it. You could, for example, first try to group terrorist acts in some
way to see if they shared any characteristics. Then, using information in the
public domain, you might pick a couple of these groups to focus on, looking
for clues as to how they evaded detection by the security services. While this
would be instructive, it would not be likely to offer much data that would
provide an enlightening analysis. Really, on the question of evading the
authorities, something richer that taps a variety of sources and aims to knit
them together is likely to be more helpful. In other words, a case study.
Once you have come to this conclusion, you are at the stage of explain-
ing why a case study would be valuable in the circumstances of your current
knowledge and the question that you have sketched out. Your expectation is
that a case study of a particular terrorist incident will offer more insight into
this question than a range of other approaches. By enabling you to focus in
depth, a case study will mean that you can draw out the lines of reasoning,
to see the possible strands of causality at play.
You decide to focus on a terrorist incident that will provide a good deal
of detail about the people involved in it. The availability of data on the back-
grounds of the people involved is the key issue here. Once you have chosen
an attack, what might you do to look at how it evaded the gaze of author-
ity? Remembering that the case study is singled out by its emphasis on the
richness of the individual account or narrative, you might choose to focus
on one of the terrorists in a particular terrorist cell. You could focus on all
of them and treat the attack itself as ‘the case’, but this would necessarily
reduce the richness of each terrorist’s history and situation. With the depth
provided by one account, you can look at a great deal of detail: you can look
at the childhood of the terrorist, his education, his early adulthood and his
friends – was he a loner or did he have lots of friends? You can look at the
reasons for his attraction to fringe religious sects and, ultimately, groups of
people who were connected with Al Qaeda.
This initial inquiry might tell you a number of things. It might point, for
example, to his childhood in a wealthy family and his upbringing in a British
private school, where he developed an especially aggravated feeling of dif-
ference and alienation from his host culture. This might give you a clue to his
resentment of authority and Western culture and his willingness to engage
in acts of this kind. More relevantly for the dimension of escaping capture by
the authorities, it might point to a familiarity with isolation and a predilec-
tion for ‘keeping his head down’, avoiding notice by keeping quiet. There
YOUR PROCESS 177
is a range of strategies that could be involved in such a way of dealing with

life – not getting yourself noticed by being naughty or, indeed, by being
particularly good, not volunteering, always fitting in and going along with
the majority. In school life, such strategies are easily developed. You could
point out how easily such strategies might transfer to life in the adult world.
Then there is the issue of identity – the need for belonging. Having
been brought up feeling an outsider, such a need would be particularly
exaggerated – he would want to feel what the political scientist Robert
Putnam (2000) calls social capital. He would want to feel part of a club – not
literally a club, like a bridge club or a tennis club, but a group of people
who are brought together by a common aim and feeling a sense of togeth-
erness. Not getting this from his host culture – indeed, feeling alienated
from it – he would be vulnerable to all kinds of approaches from those who
could offer such an identity.
Always remember to keep your question at the front of your mind when
building an interesting pen picture such as this, as it is all too easy to get
distracted and go off on a tangent that is not related to your question. Here,
your question is about evading detection. How could you use the pen pic-
ture you are developing to help you answer this question? Well, the need
to feel part of a club might lead a person to particularly strongly bonded
groups – groups with a powerful sense of identity. Going back to Robert
Putnam for a theoretical framework, he uses the distinction between what he
calls ‘bridging’ and ‘bonding’ activities to strengthen identity. Bridging activ-
ity is the more inclusive kind of socialising, looking for new members and
looking outwards. With bonding, by contrast, members seek to strengthen
the group by excluding others and making the group exclusive rather than
inclusive. Such a group will seek ways to make itself invisible, not draw atten-
tion to itself, shun the approaches of outsiders or suspicious newcomers. It
will develop rituals of secrecy.
Already, we can see a potentially interesting personality case history
developing here, using Putnam’s distinction between bonding and bridg-
ing as the theoretical ‘tool’ (which I describe in Chapter 11). We can see a
boy who feels different from the crowd and has this feeling exaggerated on
going out into the big wide world. We can see the beginnings of an expla-
nation about how he and a group of similar young men might develop a
tight-knit, highly ‘bonded’ group that learned how to avoid being noticed,
especially by the authorities. They would be likely to assume low-profile
personae wherever they went, assuming these almost automatically, in situ-
ations where they were likely to meet with authority figures.
These kinds of personae could be seen as the backdrop, providing the
fabric within which their everyday encounters were framed. As an inquirer
(Continued)
into a project such as this, you could also seek to understand (within the
constraints of your own position, probably as a student inquirer with little
power to find out detailed personal information) what the incipient terrorists
did specifically to avoid observation. Were they known to have used coded
messages? Did they assume disguises? Did they use pseudonyms? These,
then, would be evidence of the more deliberate avoidance strategies.
A conclusion to the study might show how these strategies had broken
down. Perhaps the group had been infiltrated by the security services? Did
those services understand the modus operandi and psychology of groups
such as this, looking for psychological profiles and patterns of behaviour,
seeking out such groups to penetrate them?
All the time, you as an inquirer would be focusing first and foremost on
the individuals involved, building a narrative. Because this is a case study, it
is the narrative structure that is important – the story, in other words, build-
ing a picture of the personalities involved. Remembering, too, that this is a
single case study, the one individual would form the core that would bind
together all of the other elements, around whom the tapestry of your story
would be woven.
Time as a dimension of the case study

The great historian R.G. Collingwood (1946/1994: 72) pointed to the significance
of time and past events when he said that ‘truth … is the daughter of time; the best
knowledge is a fruit of the ripest and richest experience’. Time is a feature in most
case study research. You are looking back on a phenomenon, situation or event; or
you are studying that phenomenon, situation or event as it happens. For my clas-
sification of case studies in this chapter I have divided case studies into retrospective,
snapshot and diachronic studies, to indicate the ways in which time is treated and
employed differently in each type. I have further divided multiple case studies into
parallel and sequential studies.
The retrospective study

The retrospective study is the simplest of these, involving the collection of data
relating to the past phenomenon, situation or event. It happened in the past, so
you will be looking at documents and archival records such as registers, newspaper
stories, diaries, logs and photographs. You may be able to conduct interviews with
those who experienced the event or participated in it.
YOUR PROCESS 179
In several of my examples so far in this book I have used a retrospective study –

see the Berlin Wall example (pages 32–7) and that of the terrorists (page 175) – so I
shall refrain from giving another example here. Do remember, though, that there
are several ways of tackling the issue of time in a case study: you do not have to
do a simple retrospective. Consider the other frames for your design process that I
outline in this chapter before proceeding.
The snapshot
Here, the case is being looked at in one period of time. It could be a month, week, day
or even a period as short as an hour. That is the frame – the borders within which the
study is undertaken. Remember that a case study is already a bounded study, and the
boundaries are defined by something – possibly an institution, or the life of the person
in question – but here, in the snapshot, it is defined by a period of time.
If this is so, then time should in some way be important. If the snapshot defines
the study, then you are not getting contextual information arising from a whole
range of ongoing sources of data that you are able to mix and match. Your focus is,
if you like, ‘a day in the life of’ and the skill will lie in being able to draw data into
that narrow frame.
In doing this, the snapshot case study is illustrative, demonstrative and analyti-
cal, like other forms of case study, but the analysis here is aided by the temporal
juxtaposition of events – and that temporal juxtaposition (in other words, things
being next to each other in time) will help you and the reader to understand the
connectedness of events.
By presenting a narrative from one day, whether it is like a continuous video or
more like a scrapbook, you are offering the reader your analysis of the situation.
The object emerges out of this analysis. As Walton (1992: 129) puts it, ‘Cases come
wrapped in theories.’ In other words, the view being presented is a three-dimensional
picture with all of the colours painted in – the contour lines and the colour provided
by your own ‘theoretical’ stances, which are all of the things that make you who
you are and, in turn, your analysis what it is. The case inquirer is the central person
providing the theoretical wrappings, and this is done here in the snapshot especially
with a view to the picture being presented as a Gestalt over a tight timeframe.
What does this mean in practical terms? I’ve mentioned that your focus is like ‘a
day in the life of’, and it’s useful to think of one of the most famous examples of the
genre – Aleksandr Solzhenitsyn’s (1963/2000) One Day in the Life of Ivan Denisovich.
Although it is a work of literature rather than a research report, it is useful as an
example since Solzhenitsyn based the book on his own experience and it illumi-
nates his subject and enlightens his readers. Apart from any literary merit, this was
an experience worth knowing about since it gave a picture of the author’s reflec-
tions after his time in a Soviet labour camp – an experience he was forced to suffer
because of a derogatory comment he made about Joseph Stalin.
As with all other forms of case study, there should be a purpose for doing it. The
case study cannot be merely illustrative. If I were to provide a snapshot of my day
as One Day in the Life of Gary Thomas, it sadly would be nowhere near as revela-
tory as that of Ivan Denisovich. There would have to be a reason for my writing
it and, while Solzhenitsyn’s experiences in a Soviet labour camp are an immediate
and obvious source of interest, my own in my everyday life are not. Solzhenitsyn’s
experiences, however, can tell us about a whole range of things. We learn of the
desperation brought about by the intense cold and what this does to a human
being, the nature of the discourse permitted about Stalin, camaraderie, its nature
and its purposes. Then we can put all of this into the context that is known about
Stalin’s USSR and all that came after it. Not only does telling the story over one
day serve the narrative purpose of making it more interesting, real and compelling,
but also the timeframe connects the bits of the narrative. By describing the cold
and its effects on the guards (as well as the prisoners), Solzhenitsyn humanises the
guards and helps us to understand their reactions to the prisoners – both sympa-
thetic and unkind. Take this passage from the beginning of the book (Solzhenitsyn,
1963/2000: 7):
As usual, at five o’clock that morning reveille was sounded by the blows
of a hammer on a length of rail hanging up near the staff quarters. The
intermittent sound barely penetrated the window-panes on which the
frost lay two fingers thick, and they ended almost as soon as they’d
begun. It was cold outside and the camp-guard was reluctant to go on
beating out the reveille for long.
The camp guard becomes human – he is dispirited by the cold and we empathise
with him. Not only this, by empathising with him we understand better the other
actions he takes. The timeline, whether a week, a day or an hour, focuses the gaze
of the observer and encourages an analysis of the interconnections between parts
of the narrative. It helps an understanding of the ways in which one element of
the story is bound to another. It helps us to understand the fabric of the day, the
ways in which the inmates avoid or gain the guards’ attention and their facility
in harnessing from them small treats that could be traded and the ways in which
these became converted to a kind of currency. We learn about this in the context of
the intense and all-pervading cold and the prisoners’ single-minded preoccupation
with warmth and survival.
YOUR PROCESS 181
As I said above, A Day in the Life of Gary Thomas would not be as compelling as
that of Ivan Denisovich. Ivan Denisovich has the advantage, as a subject, of all
that we have just talked about. Ultimately, it helped to change the world. My being
woken – me and 6 million others – by my radio alarm competes poorly with the
camp guard’s reveille. My day could be interesting, however, as could anyone’s, if an
angle on it could be found. Here are two possible such angles, which would form
the object of my study.
• I am a senior member of staff at a university, negotiating a way between junior staff

members’ needs and wishes and the direction of the university’s management. I
could thus focus on the events in a working day that highlighted the tensions and
dilemmas and how they represent the effects of a culture of performativity and
audit and what this means for the erosion of trust in the context of the traditional
working environment of the professional. Choosing my day with care, the day’s
events could exemplify and explicate any or all of these issues.
• As a parent of daughters now in their 20s, I could choose a day that in some way
focused for me the tensions and life changes that are brought about by a young
person ‘leaving the nest’. It might be a special day, such as a graduation or moving
into the first flat. Moments during the day would provoke reverie, reflection, sad-
ness, hope, feelings of rejection (real or unreal) or joy, and these could be used as
a backdrop to an analysis of changing life circumstances and their effects.
As I have noted regarding Ivan Denisovich, this could not simply be a record. I
could not simply say, ‘this happened, then this happened, then that happened’.
That would simply be a diary, and a not very interesting one. The key in employ-
ing a snapshot is to find those interconnections over the day, week or whatever. It
is the finding of these interconnections that would constitute the analytic frame.
How did this link to that and what were my reflections on each? This is the stuff of
which analysis is made.
Here is another example of a snapshot, this one occurring over three days:
Understanding anorexia nervosa: A snapshot

case study
Katrina is on a postgraduate course in educational psychology, for which she
has to do a piece of reflective research to be written up into a 20,000-word
(Continued)
dissertation. Katrina is on a placement at a young persons’ unit of a psychi-

atric hospital and decides to do her dissertation based on this experience.
Among the patients at the unit is Briony, a 16-year-old young woman admit-
ted because of anorexia nervosa. It is difficult to understand this condition and
hugely frustrating for professionals to work with young people who display its
symptoms. It is a form of self-destructiveness that seems almost to have its own
power source behind it. The difficulty is understanding what this power source is.
Katrina wishes to do a snapshot case study since, she hopes, it will help her
to understand Briony’s condition by bringing together the interconnected-
ness of several elements of this young woman’s life – the food consumption
itself, its antecedents (that is, what went before) and its immediate conse-
quences and how all of these are connected to Briony’s social relationships
with her family, friends, patients and hospital staff.
Having established clearly the informed consent of Briony and her family,
Katrina decides to spend three whole days with Briony, staying at a guest house
close to the hospital, so that she can devote the whole of the three days to
being with Briony, talking with her, observing what she does and to whom she
talks, noting her eating habits, their context, antecedents and consequences.
As part of all of this, Katrina decides to conduct two long interviews with
Briony in which she intends to ask this bright, likeable young woman for her
own reflections on the condition (see Figure 9.3). She will focus on a recur-
rent theme in the literature on anorexia – that is, in dangerously restricting
food intake, people with the condition are taking control of their lives in a
way that they had hitherto found impossible.
Consequences
Behaviour
What happens after
Antecedents The eating itself.
the eating? Are
What goes before What is it
these consequences
any eating? characterised by?
in any way contingent
How does it happen?
on the eating?
Figure 9.3 An ABC of behaviour, linking events over time
Depending on Briony’s responses, Katrina will look to link this theoreti-

cal position with the young woman’s actual behaviour in situ at the unit.
Is there evidence of any kind of implicit or explicit control going on as the
day proceeds and as mealtimes approach, are encountered and pass? How
does this control take shape? Is it of her physical environment or the people
around her or her own moment-to-moment existence? How can any or all of
these be understood?
YOUR PROCESS 183
The diachronic study

The diachronic study shows change over time. In this way it contrasts with the
snapshot study, which is a here-and-now study. Again, by contrast, it is different
from a sequential study in that it is not two or more studies in sequence one after
another, but one study that reveals differences as it proceeds. Here is an example:
Investigating the Hawthorne effect: A

diachronic case study
You may have heard of the Hawthorne effect – a well-known term in organi-
sational psychology. It is about the phenomenon of productivity increasing
when an interest is shown in staff (see Figure 9.4). It does not matter what the
interest is – any interest seems to have this effect most of the time.
The effect was named after the
Hawthorne Works, a factory near
Chicago making electrical parts for
telephones and radios. In 1924, the
US National Research Council sent
two engineers to run experiments
there in the hope that they would
learn how improving lighting on
the shop floor influenced workers’
productivity.
To cut a long story short, they
did discover that improving light
levels improved productivity. Not
only this, but subsequent analysis
showed that any change – the main-
tenance of tidy workstations, moving
benches around, clearing floors and
even reducing light levels – all had
the same effect. The consequence
of changes such as this came to
be known as the Hawthorne effect,
sometimes also called the experi- Figure 9.4 Is everybody happy? The
menter effect, and was written up in Hawthorne effect. Photo by Austrian
the literature by Roethlisberger and National Library on Unsplash
Dickson (1939).
(Continued)
Why should the Hawthorne effect happen?

Working in a factory that makes car parts, Bev is studying part-time for
an MSc in human resource management and decides to study the phenom-
enon as the topic for her 20,000-word thesis. She decides to use her own
place of work to take this question further, clearing with her bosses, and
three employees on part of the assembly line, the making of a series of
small changes to work routine, during which time their productivity will be
assessed. Clearly, this involves gaining the trust of her colleagues, clearance
from the staff’s trade union and ethical clearance from the university, but,
given the good industrial relations at the plant, none of these is too difficult
to achieve.
Bev’s starting point is that the Hawthorne effect is a real phenomenon,
even though there has been some evidence (Levitt and List, 2009) to suggest
that it might have been a result of the way the data were originally collected
in Chicago all those years ago. Bev, however, after having discussed this with
her tutor, took the effect to be so well established in subsequent replications
that it was worthy of additional study.
She was interested in the why and the how – why and how does produc-
tivity improve when a change is made to working conditions, and why does it
not tend to maintain itself over time? A case study, Bev feels, may help to get
to the means by which the effect works. It may give clues to its mechanism
of action. A time element will clearly be at the core of this study, looking at
changes that occur over time as the change is introduced and as it becomes
part of the routine. Time, in other words, provides the basis of the analytical
frame.
The change that Bev decides to introduce involves raising the height of
these three employees’ seats by 2 centimetres – not very much, but enough
to make a noticeable difference.
Now, remembering that a case study is a bounded case, Bev has the situ-
ation she wishes to study bounded by:
• the physical context – the workbench, seat, engine parts zinging down
the assembly line
• the social context – the other staff and the group’s supervisor
• time – the key one, with changes happening as time proceeds.
Bev has to remember in doing this case study that it is not governed by the
same principles as an experiment (unless she were to deliberately set up
an experiment as part of the case study, but she is not going to do that).
So, she is not going to be changing the value of a variable (the height of
the seat) while holding the other variables constant in order to assess the
YOUR PROCESS 185
effect of the change. Nor is she going to be formally comparing this group
with another group of workers further down the assembly line who have
not had their seats raised. Rather, she is going to be looking at the case as a
whole from a number of directions and interpreting what she finds. Always,
in analysing her findings, she will have to have at the front of her mind her
purpose for doing the study, which is to understand the Hawthorne effect
in her particular circumstance. Her understanding is informed by all of the
work on the Hawthorne effect that she has discovered during her literature
review. On the basis of the latter and her thinking, she draws up a story-
board about potential explanations, as shown in Figure 9.5.
Psychology of well-being
Doesn’t a certain
and happiness
amount of stress improve
Reduction in performance?
Why should stress?
feeling special
make people work
harder? Why should cohesion
make people work
harder?
Feeling special and

feeling cared for Sense of group
cohesion?
What causes the Hawthorne effect?
Figure 9.5 Storyboard for the causes of the Hawthorne effect
So, Bev starts the study and raises the seats of the three staff. She
assesses their output and finds that it does indeed rise in all three cases.
(Suppose it had not risen. Would this have been the end of the inquiry? No.
Bev could have proceeded with the study, either modifying it in such a way
that another change or changes were introduced until an effect was found,
or else she could simply have continued with the study as an exceptional
example, offering reasons for there having been no Hawthorne effect in this
case. It could, for example, lend weight to the findings of Levitt and List
(2009), who suggested that Hawthorne effects are not real, which would then
point to a new direction her inquiry could take.)
On finding a Hawthorne effect – the rise in output of these three mem-
bers of staff – Bev could proceed to the crux of the study, which is an
(Continued)
assessment of the social and psychological processes that might contribute

to the improvement in performance.
To guide her data-gathering and its subsequent assessment, she uses the
storyboard she drew (Figure 9.5). She has long semi-structured interviews
with each of the staff separately and conducts a focus group with them all
together. In this she inquires about their feelings of security, ‘place’, stress
and happiness using focus materials – photographs and video clips – that
stimulate discussion about safety, confidence and well-being. She asks
whether they are getting on better or worse with others participating in the
study and with those outside it. She uses the ‘5 Whys’ technique employed
by the Toyota Motor Corporation (see Ohno, 2006) for trying to get at the
‘root cause’. She also observes what appears to be going on down the
assembly line. Is there, for example, any teasing or ignoring or even bullying
going on?
Bev focuses especially on the ideas raised in her storyboard and proposes
from her findings that the effect is not produced as a result of any improved
sense of well-being. Rather, the interviews, focus groups and observation
seem to point to a sense of involvement in the work of the company. Rather
than being ‘us and them’, there is more a sense of just ‘us’, with communica-
tion appearing to take place and a sense that the staff involved are being
listened to.
An interesting feature of Bev’s study is that it involved an alteration in the envi-

ronment of her staff, in the way that is sometimes thought to be appropriate only
for an experimental study. Yet this was not an experiment. It invoked the change
in order to provide the basis for an interpretative study – one that is idiographic in
nature.
The multiple or collective or comparative

case (or cross-case analysis)
when there is even less interest in one particular case, a number of cases
may be studied jointly in order to investigate a phenomenon, popula-
tion, or general condition. I call this a multiple case study or collective
case study. (Stake, 2005: 445)
In a multiple case study, because there are several subjects, each individual sub-
ject is less important in itself than the comparison each offers with the others.
YOUR PROCESS 187
For example, a study of two schools’ different responses to a peripatetic education

support service might be done to try and throw light on the differences in these
schools’ biographies and characteristics. The key focus would not be on the shape
of relationships per se in one school, but, rather, on the nature of the differences
between one and the other. Selection would have to be on the basis of prior knowl-
edge by the researcher of these levels of difference, and the interpretations would
also be made in the context of this knowledge. The key to this research would be
to drill deeper, to find out more and undertake
more searching analysis of the cultures of the
Another term for multiple case
two environments. This comparative element study is ‘cross-case analysis’, since
is why Schwandt (2001) calls this kind of case the emphasis is on the comparison
study ‘cross-case analysis’, putting the empha- between the cases.
sis on the comparison that is done.

Because you are using multiple studies and making comparisons does not mean
that you are in some way reinvoking the need for representative samples of each
phenomenon. Just as much as in any other case study, a cross-case analysis is about
the ‘guts’ of the case, seen in its wholeness. There is a platform, though, on which
sets of wholeness are compared. Remember (once again) that case studies can never
form a sample from which you can generalise. In the somewhat pompous language
of research methodology, the case you choose to look at will be a non-probability
sample. (As I mentioned, though, on pages 66–8, there are good grounds for saying
that your case is in no sense at all a ‘sample’.) So, if you are comparing cases, you
are comparing these cases for what they show.
Interestingly, one of the very first case studies ever to have been systematically
conducted was a comparative case study. It was undertaken by a French mining
engineer, Frédéric Le Play, in nineteenth-century France. Le Play had travelled
as part of his work as an engineer, but, perhaps to the chagrin of his employers,
became more interested in the people he met than the mines he was visiting. On
his travels, he would stay with a family and remain with them long enough to get
a full picture of their way of life (Mogey, 1955). Drawing from such intense obser-
vations of individual families, he was able systematically to compare the lives of
families in rural German mining villages with those of French urban workers. Partly
because of the vividness of these first-hand accounts, his comparative case studies
achieved literary as well as popular acclaim and led to significant reforms – initiated
by employers – in the treatment of workers in urban France.
Returning to the twenty-first century, at several points in his lovely book The
Craftsman, Richard Sennett (2009: 182–93) uses case comparisons, or cross-case
analysis, to set off strands of analytical thought in his readers’ minds. I shall draw
from one of these here:
Sennett’s comparison of chefs: Cross-case

analysis
The Craftsman is a highly readable work of scholarship about craft knowledge –
how it is acquired and how it is passed on to others. Sennett legitimately uses
a panoply of different sources to compile his thesis – historical records, case
studies, conversations, reminiscences, personal experiences, scientific literature
and more.
Sennett uses a comparison to make a point about instructions. His point
is ‘show, don’t tell’. In other words, in helping to share a craft with another
person – whether it is woodworking, cooking or whatever – the mentor or
craftsperson should give examples rather than try to explain. It does not
matter how strange the examples are, they are better than instructions
based on task analysis. The mind does not work like a computer, with
watertight sequences of ‘do this; don’t do that; now do this’. We need real
examples from life that we can get inside, like a comfy pair of shoes – we
need metaphors from practice. We learn by empathising, imagining and
doing, feeling our way with the guidance of a tutor. Whether this tutor is
a real flesh-and-blood person or a book (as in Sennett’s examples about
recipes) is immaterial. What is needed is the feeling that you are sharing
the task, understanding its nooks and crannies, being guided rather than
told. For those familiar with the psychology of learning, the congruence with
Vygotsky’s thinking is clear.
Perhaps I should rephrase the comment I made just now about learning by
empathising and doing. It should really be, ‘We learn best from empathising
and doing.’ Of course, we could simply learn – no more and no less – from a
set of instructions, but our learning would have no depth, no feeling. To bring
the craft, the magic, into the learning, to make a wooden box with enjoyment
and care – a box with soul rather than just a box – needs skills learned by the
side of a craftsperson or those honed over a lifetime’s experience.
Anyway, this is the essence of what Sennett is trying to illustrate with
his comparison of four chefs and how they cook various chicken recipes.
Sennett’s case study is:
• local, since one of the cases draws from his own experience, and key
because they all are exemplars
• instrumental, since it serves a purpose
• illustrative, since it is showing something
• multiple or comparative, since the core of the message comes out of the
comparison of cases.
YOUR PROCESS 189
So, remembering back to the taxonomy of purposes, approaches and pro-

cesses, the path that it takes is as set out in Figure 9.6.

Single
Multiple
Explanatory Experimental • Retrospective
• Diachronic
Figure 9.6 Sennett’s case study
The first chef Sennett describes is Richard Olney. Sennett picks on his
description of how to bone a chicken. The instructions provided by Olney,
says Sennett, tell instead of show: ‘If the reader already knows how to bone,
this description might be a useful review; for the neophyte it is no guide.
Many unfortunate chickens will be hacked to bits if a beginner follows it.’
Here is a sample of the instructions that Sennett is talking about: ‘Sever
the attachment of each shoulder blade at the wing joint and, holding it firmly
between the thumb and forefinger of the left hand, pull it out of the flesh
with the other hand.’ He likens this kind of writing to engineering: ‘Not only
do engineer-writers leave out “dumb things” that “everyone knows”: they
repress simile, metaphor, and adverbial color. The act of unpacking what’s
buried in the vault of tacit knowledge can make use of these imaginative
tools.’
The second part of the multiple case study consists of a word picture
that Sennett draws of chef Julia Child. He gives an account of how Child’s
recipe is like a story that gets inside the chef’s head, expressing forebod-
ings and sympathy. She uses analogy, focusing on the cook rather than
the chicken. She gives clues rather than direct instructions and guides
the reader like an expert with an apprentice, anticipating difficulties and
suggesting ways around them: ‘for a moment Child will imagine holding
the knife awkwardly; the cello master will return to playing wrong notes.
This return to vulnerability is the sign of the sympathy the instructor gives’
(Sennett, 2009: 186).
Part three is a pen portrait of the famous Elizabeth David, who shared
her skills with her readers by giving the cultural context of the food. She
(Continued)
talks about local cooks in France touching and prodding the bird. She
tells stories about when, how and with whom she ate and gives interest-
ing titbits about how tarragon might be used by Bordeaux cooks while,
for the same recipe, sage was used by the cooks in Perpignan. So, as a
reader you think about the food you are cooking, being encouraged to
consider the process as creative, variable and intuitive rather than a rigid
set of procedures.
The fourth case study offered by Sennett is perhaps the most interest-
ing. It is of Madame Benshaw and her recipe for poulet à la d’Albufera.
Madame Benshaw is different from the others in that she is not famous –
she emerges from Sennett’s own personal knowledge – and because
she talks almost entirely in riddles, devoid of instructions in the formal
sense. Sennett was a student of hers at an evening class and he says that,
because her English was poor, she would teach almost entirely by exam-
ple, ‘coupled with slight smiles and emphatic, frowning contradictions of
her thick eyebrows’. With Sennett’s help, Madame Benshaw wrote down
the recipe:
Your dead child. Prepare him for new life. Fill him with the earth. Be
careful! He should not over-eat. Put on his golden coat. You bathe him.
Warm him but be careful! A child dies from too much sun. Put on his
jewels. This is my recipe. (2009: 190–1)
By the ‘dead child’, she meant the chicken. The ‘preparation for new life’
meant boning. ‘Filling with the earth’ was about stuffing, and ‘not overeat-
ing’ about not overstuffing. The ‘golden coat’ was about browning before
baking. The ‘bathing’ referred to the preparation of a poaching liquor and
the ‘jewels’ the pouring on of this liquor, basting.
This recipe, then, is told entirely via metaphors, and these allow the nov-
ice cook to feel and understand. Calling the chicken a ‘child’ immediately
evokes a sense of intimacy, protection and tenderness, so we understand
that we should place the bird only in a cool oven. As Sennett puts it, meta-
phors ‘roll forward and sideways’, allowing us to gather different meanings
as they do so.
This comparative case study illustrates exactly the point that Sennett
wants to, which is that learning a craft is not about simply following instruc-
tions but a human process. He builds the theory that in doing this, we do
well to follow our noses, to use, if you like, our ‘instinct’ for learning. This is
something that is done by modelling and copying, getting inside another
person’s head. To build the theory, he draws out the key points and synthe-
sises them, as shown in Table 9.1.
YOUR PROCESS 191
Table 9.1 Chefs cooking chickens – Sennett’s multiple case study
Chefs Style (for comparison) Analysis

Richard Straight instructions that give almost clinical The general point is about
Olney directions about what to do learning coming from
imagining and doing and
Julia Child Use of analogy, focus on the cook rather
using the knowledge you
than the chicken and use of the learner’s
already have to help you
vulnerability
learn something new. It does
Elizabeth Talking about the culture, taking the reader on not come through formal
David a journey, discussing variations – making it, instruction
therefore, less rigid, less fixed
Madame Talking entirely in metaphors that let the
Benshaw reader ‘get inside’ the mind of the chef and
understand what she is saying
This is where many student researchers go wrong in doing a case study of this
kind. There is too much description – and it is just description. Too little actually
comes from looking at the differences between the cases. It is this that is, in fact,
more important than the presentation of the cases. Fascinating though Madame
Benshaw’s case is, it is far less interesting on its own than it is when compared with
the others. The analytical frame is provided by the comparison.
An interesting comparative case in the literature can be found in the work of
Jiménez and Gersten (1999). They made comparisons between the teaching styles
of two Latina/o teachers and the different ways in which they went about teaching
literacy.
Nested case studies

Yin (2018) draws a useful distinction between multiple case studies and what he calls
‘embedded’ case studies. I prefer to call the latter ‘nested’ studies, partly because it is
one less syllable than ‘embedded’, but mainly because ‘nested’ gives more of a sense
of a subunit fitting in with a larger unit, rather than its being implanted there. It is
that fitting in that is of interest – how does the subunit connect with other subunits
and the whole?
With multiple studies, the emphasis is on comparing clearly different examples,
and the contrasts found between and among the cases then throw the spotlight on
an important theoretical feature. By contrast, with nested studies the breakdown
is within the principal unit of analysis, such as classrooms within a school. This is
summarised in Figure 9.7.
Nested
Nested
Case 1 Case 2 unit 2
unit 1
Wider
case
Nested Nested
Case 3 unit 3 unit 4
Case 4
Multiple case studies, offering

Nested units within one case
comparisons
Figure 9.7 Multiple and nested case studies compared
A nested study is distinct from a multiple study in that it gains its integrity, its
wholeness, from the wider case.
As with all decisions regarding case studies, there has to be a reason for conduct-
ing the study in a particular way and, with the nested study, that reason comes
out of contrasting the units as part of the wider case. If you were looking at three
classrooms within one school, but the school had no significance other than that
it physically housed these three classrooms, then the cases would not be seen
as nested. They are nested only in the sense that they form an integral part of a
broader picture – integral to something that might be happening within the school.
Let’s flesh out this example a little.
Comparing six classes: A nested multiple

case study
Franklin Drive Primary School has just decided to implement an inclusion
policy. Deputy headteacher Lara, who is studying for a master’s degree in
education, decides to examine the ways in which the policy can be under-
stood and implemented throughout the school. This analysis will be done
YOUR PROCESS 193
for the dissertation she has to complete for her degree, but she expects that
it will also form the basis for developmental work in the school afterwards.
Lara’s starting point is that the kinds of changes expected to result from
an inclusion policy of this kind are difficult to implement at a practical level
because they depend on the cooperation of individual people, and different
people notoriously behave in different ways.
Lara’s tutor has referred her to a classic paper, ‘The myth of the hero-innovator’,
by Georgiades and Phillimore (1975). The point that Georgiades and Phillimore
make is that implementing change in an organisation is far more difficult than
it might at first sight seem. Because of this, the organisational landscape is lit-
tered with the carcasses of brave people who have tried to change things but
failed. It is almost as if people working within organisations, whoever they are –
teachers, doctors, firefighters, zookeepers – conspire to resist change. Allied
to this is Lara’s knowledge that there are few recipes for success with an inno-
vation such as the implementation of an inclusion policy. If trying to introduce
something of this kind, you have to adapt and modify to suit the organisational
conditions you find in different places.
This is the starting point for Lara’s study – that is, the six classes of her
primary school are all very different since all the teachers in the school have
their own styles which will have to be accommodated in the implementation
of the policy.
She chooses three teachers and their classes as nested case studies within
the school. The core of Lara’s study will focus on the inclusive principles
espoused by the school and the ways in which it is possible to implement
these aspects of the inclusion policy, given the different styles of the teach-
ers working at Franklin Drive. The principles of the inclusion policy include:
• admissions policy and procedure

• the way that the school allocates resources
• assessment procedures
• staff development
• specific inclusion strategies
• the curriculum
• parental involvement.
The first four of these are whole-school issues that will not be affected by the
idiosyncrasies of the teachers’ styles. Lara is concerned about the last three
(Continued)
points, however – namely, specific inclusion strategies, the curriculum and

parental involvement. She knows that teaching styles differ greatly in the
three classes on which she is focusing. The teachers’ attitudes to parental
involvement also differ. The teachers are:
• Jon, a newly qualified teacher, who does things by the book and enjoys
involving parents, but is rather in awe of them
• Meryl, who retires next year and is a formal teacher who firmly declines
offers of help from parents or volunteers
• Jane, a mid-career teacher who is spontaneous, dismissive of govern-
ment directives, popular with parents and works easily with them.
Lara decides to focus in these studies on the ways in which these three teachers
can implement the school’s new policy, given their different styles. She consid-
ers forming a focus group with Jon, Meryl and Jane in which she will present
the policy and discuss views on its implementation. On reflection, however,
she rejects this idea, realising that it may accentuate any differences in style
and make Meryl defensive and alienated from the others. What she needs to
do is study each of these classrooms as environments for the development of
the policy, each working in its own way. In the sense that the study will help to
change practice, this is a case study that is also a piece of action research.
Lara ultimately decides to conduct semi-structured interviews with each of
the teachers, in which she will ask about inclusion strategies, the curriculum
and parental involvement. These will be analysed using the constant com-
parative method (see Chapter 11) for both unique and shared ideas about
implementation in class. The analytical framework that she will develop will
serve as the basis for a summative plenary discussion with the teachers in
which their ideas are shared.
As Lara expected, the interviews reveal quite different attitudes to the
curriculum, learning and the community, as well as education’s place in it.
She uses construct maps (Thomas, 2017: 246–8) to show diagrammatically
the differences in thinking and try to draw strands of commonality from
them. She fails in this last aim, however. While Jon and Jane show some
commonality in their thinking, there is no congruence at all with Meryl.
Lara’s conclusion is that the units of the school operate as quite discrete
elements in the larger structure of the school, and she uses Bronfenbrenner’s
systems theory (see Chapter 3) to frame her analysis here. She speaks in her
write-up of the insularity of the teacher role within the school and of the
resilience of this insularity. She also questions if it would, in fact, be a good
thing to attempt to break this down, coming back to her starting point
about styles – styles that are built on attitude, belief, personality, habit
YOUR PROCESS 195
and disposition. These facets of individual teachers’ make-up would be dif-

ficult if not impossible to break down, and in any case should not be broken
down, Lara decides. Lara’s conclusion also includes reflections on the role of
policy and the extent to which attempts at the implementation of policy may
undermine the teachers’ identities and effectiveness. The latter, she suggests,
is framed around the characteristics of each teacher and the circumstances
of his or her class, and attempting to change this may be counterproductive.
I hope that you can see how the nested study is different from the simple multi-
ple study, comparing particularly the multiple study of chefs with the study of the
three classrooms. Comparisons are at the heart of each kind of study, but in the
nested study these occur in a wider, connected context.
Parallel and sequential studies

Parallel and sequential studies are both forms of multiple study. In the parallel
study, the cases are all happening and being studied at the same time, as in Lara’s
case above, while in the sequential study the cases happen one after the other and
there is an assumption that what has happened in one or in an intervening period
will in some way affect the next.
Given that Lara’s study suffices as an example of a parallel study, let me give an
example of a sequential case study to complete the picture:
Accessibility of public spaces: A sequential

multiple case study
Sarah is studying the accessibility of public spaces – particularly public build-
ings, roads and parks – for people with disabilities, including autistic people.
She works for a large charity that has agreed with a local authority to share
funding of the replanning and landscaping of a small city-centre park. This
project will be used as a pilot and, on the basis of its findings, the charity will
aim to work with other local authorities to make changes elsewhere, inform-
ing the conduct and design of these further projects.
(Continued)
Under Sarah’s guidance, the park is replanned by a group comprising

a landscape architect, disabled people, autistic people and local authority
planners. They work with a computer design expert from the local univer-
sity to produce virtual reality environments and computer visualisations that
enable disabled people and autistic people to offer their advice on possible
changes. They expect ultimately to change paths, signage, access to build-
ings and the space devoted to various activities, including play facilities for
autistic children and teenagers (Kossyvaki and Papoudi, 2016; Wood, 2019).
Sarah’s plan is to design a case study in two parts, one before and one
after the change, as a component of a larger inquiry in which the use of the
park by various groups of people is monitored. This therefore involves a
before-and-after comparison, so it draws its rationale directly from evalua-
tive research.
Sarah’s choice of participants is clearly important given the topic. A selec-
tion of people with different kinds of disability is necessary. She chooses to
talk in depth to six people before the changes and the same six people after
the changes. Three participants will be wheelchair users and three will be
identified by a local support group run by autistic people.
The first part of the study (A1 and B1 in Figure 9.8) involves, first, an
examination of statistics on usage of the park by disabled people before the
changes (A1), and next, initial interviews and ‘walk-throughs’ of the park with
the six participants (B1).
The second part of the study (A2 and B2) involves, first an examination of
statistics on usage of the park by disabled people after the changes (A2), and
next the follow-up interviews and walk-throughs with the six participants (B2).
Sarah conducts her work with her participants by engaging them in
unstructured interviews and the ‘walk-throughs’ of the park, with a non-
directive invitation to offer ideas and comments about anything that occurs
to them about the park’s design.
Change implemented
Planning A1. Statistics B1. First A2. Statistics B2. Second

in park
using virtual on usage, interviews on usage, interviews

reality collected and ‘walk- collected and ‘walk-
mock-ups pre-change throughs’ post-change throughs’
Time
Figure 9.8 Sarah’s research, showing the sequence of elements

YOUR PROCESS 197
The first part of the study (A1 and B1) provided confirmation of an increase
in usage and an endorsement of the improvements in access to the park.
The in-depth interview and ‘walk-through’ element (B1 and B2) gave mixed
information about the nature of the changes, with a difference between the
wheelchair users and the autistic adults in the way they responded in their
interviews and walk-throughs. Those using wheelchairs were indeed content
with the primarily access-based changes. The autistic adults, however, were
unhappy with what they felt was an unthinking association of autism with
learning disabilities, and this raised ethical and practical issues that Sarah
and the charity had to pursue. On the substantive subject of the project –
park design – the autistic participants made comments about how green
spaces can provide a sense of security and enable autistic people to self-
regulate from sensory and emotional overload. They talked of the need for
quiet spaces, not too crowded, and a need for planning for noise reduction.
Predictability, they said, needed to be thought about, with uncluttered sig-
nage which used clear and unambiguous language with visual supports.
By doing her case study sequentially in this way, Sarah was able to offer
useful information to aid the rolling out of the programme to other parks.
For example, she had been able to understand how the design that had
been produced with the aid of computer visualisation had not met all expec-
tations. She was then able to make suggestions about future projects. Also,
although the case studies did not draw from a representative sample of disa-
bled people, they enabled a clearer understanding of the needs of these
users, pulling in ideas that would not have been garnered from a study with
a more one-dimensional focus.

take this …
Ask yourself if your research question can be addressed by a single focus on one person
or situation or whether a comparison between different cases would be better. Is there a
time element to your question that will be addressed by looking at a sequence of events
or is it better to examine one tightly defined period in time, such as a day? Would it
be helpful to extract a number of nested elements from your main focus and examine
these in detail? Thinking about these questions and how we might categorise different
forms of case study, we can first of all say that the subject of a case study may be:
• single
• multiple.
Beyond this, time can be thought of as a key component in structuring the case
study. We have thought about:
• retrospective case studies

• snapshot case studies
• diachronic (or longitudinal) case studies.
Each of these employs time in a different way. And if the study is a multiple one,
the multiple subjects may be studied:
• as nested units within a larger whole

• as parallel units, being studied separately
• as sequential elements.
As with all questions about design, thinking carefully about what you need will be
repaid in your findings and the analysis you are able to draw. Keep asking yourself,
‘What kind of analysis, what kinds of insights am I seeking?’, and then structure the
case study – thinking about both the subject and the object (pages 16–18) – around
these questions.
REFLECTIVE ACTIVITY
From your own area of interest, outline two case studies (identifying subject and
object in each), one of which would usefully be constructed around …
• a single case
… and the other which would be constructed around
• multiple cases
… and in both case studies say whether the study would be using time by
looking back (that is, retrospectively), or at a particular snapshot in time, or
diachronically. You may also like to draw each case study’s path through the
typology.
YOUR PROCESS 199
Further Reading
Mills, A.J., Durepos, G. and Wiebe, E. (2010) Encyclopedia of Case Study Research
(Vols 1–2). Thousand Oaks, CA: Sage. doi: 10.4135/9781412957397
These volumes give some good examples of different kinds of case study research
and offer a particularly interesting discussion of multiple case study.
Sennett, R. (2009) The Craftsman. London: Penguin.

I offer The Craftsman as further reading for this chapter partly because I have drawn
on Sennett’s examples, but mainly because his thesis is so apposite for the case
study. It is about how we make sense when we become skilled in a craft, which is
as relevant for the case researcher as it is for the goldsmith (one of Sennett’s exam-
ples), carpenter or chef.
Thomas, G. (ed.) (2013) Case Study Methods in Education (4 vols). London: Sage.
Here, I offer examples of different kinds of case study in many kinds of social
research. Volume 3 gives examples of case studies on a range of topics, and Volume 4
deals specifically with case studies in education.
Yin, R.K. (2018) Case Study Research: Design and Methods (6th edn). Thousand Oaks,
CA: Sage.
Yin draws some interesting distinctions between different kinds of case studies.
STAGE 3
COLLECTING EVIDENCE,

Chapters 1 to 5


Chapters 6 to 9


Chapters 10 to 12

In this third part of the book, I look at:
• ways to collect data for a case study
• ways to analyse those data
• writing up a case study.

10
OUT IN THE FIELD:
SOME WAYS TO COLLECT
DATA AND EVIDENCE
‘In the field’ is a term used by researchers to refer to the process of actually collect-
ing data, as distinct from, say, doing a literature review. So, when you are ‘in the
field’, you are doing ‘fieldwork’. Fieldwork is a distinct part of a research project.
When you are doing fieldwork, you are collecting data and/or evidence.
It is important to remember that there is a distinction between data and evi-
dence. Data is another word for ‘information’ – bits of information of whatever
kind, whether they be observation records, numbers (such as test scores), inter-
view transcripts, photographs or documents. Evidence, however, is data in support
of some proposition. The propositions you come up with in your inquiry will
develop as your research progresses. As I hope I made clear in Chapter 2, the
initial, prima facie questions that you lay out in your introduction are an expres-
sion of your first thoughts about a subject, and these are refined as your work
progresses – as you read more for your literature review and your ideas become
clearer.
It is as you go through this process that the
Data are bits of information.
propositions become clearer and you, in turn, Evidence comes from data that
can be more focused about the evidence you support or refute a proposition.
need to support or refute those propositions. If
we take the example of the case study on the Hostos-Lincoln Academy (see page 42),
Rageh Omaar started with a fairly undefined question, but, as his immersion in
the relevant research and literature progressed, he developed ideas, propositions
and theories about his questions. The case study that he ultimately conducted was
aimed at gathering data that would support or refute the propositions emerging
for him.
You may be seeking data, pure and simple, to see where this takes you in develop-
ing your ideas about a topic. Alternatively, you may be seeking evidence in support
of a clearly defined proposition. Which you do depends on the kind of research
you are conducting. If you go into your case study without a tightly constructed
theory or set of propositions to guide your research, you will be seeking data that
will gather around ideas which emerge as the study progresses. If, however, you
begin with a clear hypothesis or a well-defined theory at the outset, then you have
propositions that will be supported (or not) by the data that you collect – you are
looking for evidence.
With this distinction in mind, how can you go about collecting information that
will help you with your case study? I have mentioned many of the methods that
can be used in the examples I have given in Chapters 6–9. Let’s do this more sys-
tematically now, though, and run through some of the most commonly used ways
of collecting evidence as a guide for you when you actually get down to doing your
study. This is only a guide and an aide-memoire, though, and I give more details
about methods for gathering data in How to Do Your Research Project (Thomas, 2017).
I summarise these commonly used methods of collecting evidence in Table 10.1.
Let’s look at each of them in a little more detail.
Table 10.1 Commonly used methods of collecting evidence
• Interviews mainly use words

{ structured
{ unstructured
{ semi-structured
• accounts
• diaries
• reflective journals
• group interviews
• focus groups
• interrogating documents
• questionnaires use words, images and/or numbers
• observation
{ structured observation
{ unstructured observation
{ participant observation
• image-based methods
• measurements and tests mainly use numbers
• official statistics
• other numerical data
SOME WAYS TO COLLECT DATA AND EVIDENCE 205
Interviews
Structured interviews
In a structured interview, you meet with another person and ask a predetermined list
of questions. The structured interview has a limited range of strengths:
• it can be administered relatively easily and quickly

• interviewees’ responses can be easily coded.
But, beyond these, it does not have very much in its favour. Because there is no
great advantage in giving it in a face-to-face manner, it may as well be given in
written form – in other words, as a questionnaire (see page 212).
Unstructured interviews
Likely to be used in an interpretative case study (see Chapter 8), an unstructured
interview is like a conversation. There is no fixed way to conduct such an interview.
You do not present your interviewee with a list of questions. In fact, the idea really
is for interviewees to set the agenda. They should be the ones who determine the
direction of the interview and the topics that emerge. As the researcher, you go in
with an open mind and just try to listen and facilitate. This is of course what is
wanted in interpretative research: in this kind of research you are looking for your
respondents to set the scene and let them tell you what the issues are.
Beyond this desire to let the interviewee lead the way, there are variations on
the theme of just how unstructured the interview should be. If your respondents
go completely off the topic, then you would wish to bring them back to it in some
way. However, this would need to be done sensitively, since it may be the case that
this is the way that the respondents talk – going ‘off message’ for a while before
they return to the topic. Or they might like to tell a story when they are talking,
and the unstructured interview gives them a chance to do this. If it seems the case
in this sort of circumstance that something of a delicate or sensitive nature is being
disclosed, you should clear with your interviewees after the interview that they are
happy for the specific information to be used and in what guise. You could reveal
it only as a general comment, perhaps, or it may be acceptable to your interviewees
to quote from them in an anonymous way on which you can agree. (See also the
discussion on ethics and confidentiality in Chapter 5.)
The unstructured interview will be more like a conversation than a formal inter-
view. Sometimes, though, you may wish to move the conversation in a particular
direction and prompt the interviewee. How should you prompt without setting the
agenda yourself? There are degrees of strength of prompt: you may simply say, ‘Can
you tell me more about that?’, or offer something stronger such as ‘How did that
make you feel?’ or ‘What happened next?’ However, if you really are interested in
what the interviewee has to say you would avoid a question such as ‘Does that make
you feel angry?’ This really would be putting words into the interviewee’s mouth.
Semi-structured interviews
In a semi-structured interview you provide the structure with a list of issues (rather
than specific questions) to be covered and you have the freedom to follow up points
as necessary. Because of these advantages, it is the most commonly used kind of
interview arrangement in small-scale social research.
However, for case study, the fact that it is the most commonly used does not
mean that it is necessarily the best. If you really are interested in interpreting your
interviewees’ comments and if you are a participant observer in the situation that
you are researching, an unstructured interview may well be a better choice. I make
this point because too many students (in my opinion) opt for the semi-structured
interview as the most straightforward and seemingly obvious choice, where in fact
it can lead them to do a different kind of research from that which they set out
to do. If you really do intend to do an ethnographic, interpretative case study,
entering your research environment as a participant observer, you are trying to
understand how the ‘players’ in that environment are playing their roles – depend-
ing on the meanings that they construct. If this is the case, if this is what you want
to do, then the semi-structured interview may be too rigid.
If you decide after reflection that you do need semi-structured interviews, though,
you will start with an interview schedule, which is a list of issues that you intend to
cover. You are not obliged to go through these points in order – or in any way keep
to a formal set format for the interview. Rather, it gives you a reminder of what you
want to cover. It reminds you not just of the issues but also of potential questions,
possible follow-up questions and ‘probes’, which may encourage the interviewee to
say more on these follow-ups. Probes may be verbal – for example, ‘Go on …’ – or
non-verbal, such as a nod, smile or tilt of the head.
Accounts
An account comes from a respondent and is the result of a format that allows him
or her to communicate experiences and feelings freely. Accounts are really like the
products of unstructured interviews, but without having been collected in an inter-

view. The account could be provided in a long, written piece of prose, like an essay,
or recorded in audio form, to be transcribed later. An account is handled in the
same way as the data from an unstructured interview.
Diaries
A diary may involve you, or a participant in your case study, making a record of
ideas, reflections, thoughts, emotions, actions, reactions, conversations and so on.
Alternatively, it may involve making a structured record of activities. Diaries may
be divided into three types:
• interval-contingent, where participants report on their experiences at regular

intervals; this is the most common kind of research diary
• signal-contingent, where some signalling device (a phone call or text message, for
example) prompts participants to give diary reports at particular intervals (fixed
or random)
• event-contingent, where participants provide a diary entry each time the event
occurs. This enables the capture of rare events that would not necessarily be
caught by fixed or random interval assessments.
If you are completing a diary yourself, it is best done immediately after your session
in the field, recording an assortment of opinions, views, interpretations, remem-
bered conversations and so on. It can take written form or be an audio or even a
video recording.
Some people will be more fulsome in accounts in a private disclosure of the kind
given in a diary than they would be in a face-to-face interview. This is often the case
with children and young people, as long as they have the writing skills. If they do
not have such skills, you could consider the use of an audio diary. You may even
wish to keep a video diary, where expressions, gestures, sighs, grimaces, frowns,
tears even, may be recorded – features of the situation that would be missed in a
written or audio diary.
Keeping a reflective journal

Similar to a personal diary is the reflective journal. It is similar, in that you regularly
keep a record. The emphasis in the reflective journal, though, is that you go a step
beyond simply recording events, putting the emphasis instead on your thoughts,
opinions, appraisals, judgements, feelings and insights. These reflections can then
be used to help with …
1. your own process of learning how to research, helping you to reflect on and
improve your own research practices and skills, and
2. actual data for inclusion in your findings.
Table 10.2 shows an example, recording reflections on a classroom observation.
Table 10.2 Reflections on a classroom observation
Occasion:
Observed in Ms X’s
classroom: 06/06/21
What did I From unstructured observation I discovered that Z [focus child] is highly
discover? active, moving about the classroom a great deal. I discovered that if I am
a ‘participant observer’ I need to participate in all areas of the room, not
getting too involved with one child or group as I am likely to miss things that
Z is doing as he moves around
What went well and I was able to talk to Z on a few occasions and engage him in conversation
why? in a reasonably relaxed way. I think I got more from him in these ‘natural’
meetings than I would have if I had withdrawn him for an interview. In the
unstructured observation I could see Z’s behaviour in context and I could
begin to understand it in the light of what the other children were doing and
how they were responding to him
What could I have I could have asked Ms X if she could have introduced me to the class
done better? with a bit more detail for the children as they kept asking me who I was.
I could have moved around the classroom more, not getting too involved
in the work of one or two groups. I left it too long before writing up notes
and I should have asked Ms X for her views on my observations and
interpretations
Long-term Read up more about participant observation

implications Always write up notes immediately
Always check notes and observations with key participants
The journal does not have to take this form; it could instead be more free-flowing,
recording your reflections day by day on your research work.
You can find ideas for keeping a journal and other ideas for reflective writing
in your university library website. The advice from my own university library can
be found at https://tinyurl.com/ycq78w56 and an adapted version of my univer-
sity’s tips for keeping a reflective journal, with more information, at https://canvas.
bham.ac.uk/courses/11841/pages/keeping-a-reflective-journal
Tips for keeping a reflective journal

• Write in your journal regularly, even if individual entries are sometimes short.
• Describe what has happened, certainly, but try also to take an analytical approach,
considering why things happened a certain way, rather than just what happened.
• Try techniques such as mind mapping, diagrams or sketches.
• Review the entries you have written to see if you can find themes and recognise the lon-
ger term action you might need to take (for example, to improve a particular study skill).
• Whatever you choose to write, remember that there are often no right or wrong
answers, only a learning process.
As I have noted, the reflective journal is not simply for your own benefit in the
development of your own research skills, but also may be included in your research
findings, either in the main body of the text or in an appendix. Indeed, in a case
study, a reflective journal may form a central element of your data collection
armoury, allowing you to incorporate a form of ‘thick description’, recorded natu-
rally while you do your research.
Quoting from your journal as an interpretative researcher also helps you explain
your own positionality. Here is how one researcher describes how she used a
research journal:
I drew on my reflective journals as a way of making my history, values,

and assumptions open to scrutiny, not as an attempt to control bias,
but to make it visible to the reader. For example, in the introductory
chapter of my thesis I used excerpts from my pre-research journal to
make it clear what my experiences, values, and assumptions were prior
to beginning the research. (Ortlipp, 2008: 698)
Ortlipp also describes how the reflective journal helped her to reflect on her research
methodology and change from her original plans:
In some instances critical self-reflection prompted me to change my

approach during the research process, to use methods that I had not
initially planned to use, and to discard pre-planned ways of going about
the research that I had included in my research proposal. (2008: 699)
Being open about your reflections on a topic, on your own process of working and
on your own positionality are important facets of case study research, and keeping
a reflective journal will help you in the process of making these explicit.
Group interviews and focus groups

Group interviews
Before you set up a group interview, you need to establish your reason for holding
a group rather than individual interviews. If you are interviewing a group, it should
be because the group psychology itself has some impact on the situation that is
of interest to you. You may, for example, want to find how the group (as a group)
is behaving, how it might behave in response to an imaginary event, or compare a
group attitude with individual attitudes within the same group, perhaps to judge
the power of one or two group members who may not be at all representative of
the general opinion.
People behave differently in groups: particular individuals may become more
talkative or less talkative; some people take the lead, while others follow; some will
tend to be ‘stroppy’, others helpful. And there are particular ways a whole group
will behave, differently from individuals. So, for example, there is an inclination
for a group to display what is called a ‘risky shift phenomenon’, a tendency well
established in social psychology. This is about the likelihood that a group will make
a riskier decision than an individual. If you asked a set of groups a question such as
‘Would you have a person with a criminal record to stay in your house?’ and then
asked an equivalent number of people but as individuals, you would probably find
a riskier decision (that is, more likelihood of saying ‘yes’) in the groups. There’s
safety in numbers, and this maxim applies even if decisions are being made about
wholly imaginary happenings.
So there are good, legitimate reasons for using a group interview, beyond the
mere fact that it may save you time.
Focus groups
The term ‘focus group’ has come to be used interchangeably with ‘group inter-
view’, but the two kinds of group are different in important respects. In group
interviews the emphasis is on the researcher taking a lead role, asking questions
and being in control of the discussion – rather in the way that an interviewer
is leading the discussion in a structured or semi-structured interview. So the
researcher asks questions and respondents answer. But in focus groups the
researcher plays the role of facilitator or moderator. If you are running a focus
group your aim is to facilitate or moderate discussion among participants, not
between yourself and the participants. The idea is that you take a marginal rather
than a pivotal role.
In the focus group, participants, usually eight or so, who have a relevant charac-
teristic or feature of their lives in common will be brought together in an informal
setting to discuss the topic in question. As in an unstructured interview with an
individual, the idea is to get at the understandings, beliefs and values of the partici-
pants. And in the same way that in an individual unstructured interview the aim
is to let the individual take the lead in setting the direction of the discussion, so in
the focus group the aim is to let the group take the lead. As facilitator, your role is to
stimulate discussion, and you may do this with your comments or you may prepare
a range of focus materials for the group to discuss. These are materials – for example,
objects, artefacts, photographs, drawings, newspaper clippings, short videos, audio
recordings – that help the group focus on the topic of interest.
Given that you need your wits about you to facilitate effectively in a group such as
this, it is common practice in professionally conducted research to use an observer
to record information about context, environment and participants’ behaviour. In
small-scale research it will not usually be possible to get help in this way, so it may
be helpful to make an audio and/or video record of proceedings.
Interrogating documents
Case studies take many and varied forms, and a retrospective case study may depend
entirely upon documentary evidence. It may be the case that there are no living sur-
vivors of the event, or no one who is accessible to offer their thoughts. Even in the
case of an event about which there are potential informants, these people may not
be able or willing to be interviewed.
Gathering data from documents represents an entirely different proposition from
gathering data from people. Essentially, the knack is to find the right documents,
read them and think about them.
There are methods for examining documents with your word-processing pro-
gram for, for example, readability or the existence of certain kinds of words, but
there is no substitute for your own careful reading for meaning and substance in
a document. Fortunately, reading documents has been made easier since most are
nowadays available in Word, as PDFs or online and can be searched and cut-and-
pasted from.
Given the ease with which documents can now be downloaded (particularly gov-
ernment and other policy documents), a key shortcut is in some basic document
interrogation using computer software. This is easy if the document is in Word,
but a bit trickier if it is in PDF. If the latter, you can, after checking copyright, copy
the whole thing and then paste it into Word. (Tip: To copy the whole lot in a PDF
file, press Ctrl-A, which selects the whole document; then Ctrl-C, which copies it,
then switch to a blank Word document and paste it into that with Ctrl-V.) If you
are not able to download the relevant document in word-processing or PDF format,
the one technical trick that is widely available now is through electronic scanning –
through a smartphone, a printer or a dedicated scanner. Most of these devices now
are provided with optical character recognition (OCR) software which will do the
‘reading’ (but sadly not the thinking or understanding) for you. When you scan
your text your software should give you the option of enabling the text to be ‘read’
as text (as distinct from simply taking a picture of the text). It will then save this into
a word-processing file, and once this is done there are a number of ways in which
your computer can help you to analyse the text; these forms of analysis are covered
in Chapter 11.
Questionnaires
A questionnaire is a written form of questioning and the questions may be closed
(‘Do you think parents should be allowed to smack their children? Yes or No’) or
open (‘What are your feelings about parents smacking their children?’). You may be
collecting facts or, as here, assessing people’s attitudes on a topic.
A questionnaire can be tightly structured, but can also allow the chance for more
open responses if required. It may be read out by interviewers (either face-to-face
or over the phone) or sent to respondents for them to complete themselves. It may
be sent by post or email or presented online. For online questionnaires, there are
various Web-based services available (such as SurveyMonkey, available at: www.sur-
veymonkey.com, which lets you construct your own questionnaire free for a survey
of up to 100 respondents).
Ways of questioning
The first thing to consider is whether you are going to used open or closed questions.
Open questions leave the response open to the respondent, and I will come to these
in a moment at the end of this section. Closed questions limit the response and can
be organised in a number of ways.
Dichotomous questions
‘Dichotomous’ means ‘two-way’, and the dichotomy is usually ‘yes’ or ‘no’. For
example:
Have you ever applied for a managerial position in this company?
Yes 
No 
These can often be screening questions. In other words, you may use the question
to separate respondents into groups, who can then be questioned separately. If they
have not applied for a post of responsibility (in this example), why not? Your sub-
sequent questions to this group will pursue this theme. And if they have applied,
what happened? A separate route of questioning can be used for this subset.
Multiple-choice questions
These contain two or more answers where respondents can be told either to tick
one box or to tick as many boxes as needed.
Rank-order questions
Here, respondents have to rank items (that is, put them in order) on a list according
to some criterion – for example, best to worst, most able to least able, or degree of
difficulty. Within this ranking, you can either ask for a limited number of choices
(for example, first, second and third), or require respondents to rank the whole list.
Rating-scale questions
These require the respondent to rate some experience, attribute, attitude, etc., along
a continuum. You may for example wish to ask adults about their experience of
testing and assessment when they were at school:
Remembering back to when you were at school, would you say that your
experiences of formal assessment and testing were:
Very positive 
Positive 
Neutral 
Negative 
Very negative 
The respondent will tick only one of these boxes.

Matrix or grid questions

Matrices (grids) provide a series of questions, which all have the same answer scale.
For example, if you were interested in canvassing patients on what constitutes a
good nurse, you might ask them to suggest how important each of these criteria
were, all on the same scale of 1 to 5.
Importance 1 2 3 4 5
Caring     
Efficient     
Reliable     
Kind     
Knowledgeable     
Understandable     
Helpful     
Scales
Scales, that is to say, sets of items and responses, appear in some of the question
formats above. However, the Likert scale is worth mentioning as an easily used tool
for measuring attitudes: respondents indicate their levels of agreement to state-
ments provided by the researcher relating to that attitude, belief or characteristic.
The respondent, rather as in rating-scale questions, responds to each item on a
five-point or seven-point scale, for example corresponding to strongly agree, agree,
neither agree nor disagree, disagree, strongly disagree. To remove the tendency for
some people to over-choose the middle option, this middle option is sometimes
removed, making a four-point scale. The latter is used in the Rosenberg Self-Esteem
Scale, where the first three items are as shown in Table 10.3.
Table 10.3 The Rosenberg Self-Esteem Scale
Strongly
Strongly agree Agree Disagree disagree
1. On the whole I am
satisfied with myself
2. At times I think that I
am no good at all
3. I feel that I have a
number of good
qualities
Open questions
You can think of open questions in a questionnaire in the same way that you
think about unstructured interviews, diaries and accounts. They are bounded by
the same considerations, in that you are aiming to get at the central concerns of
an issue as your respondents see them. In a written questionnaire, though, where
there is little in the way of stimulus (as there would be in an interview), if you
want to use open questions you will have to give more of a prod to jog the mind
of your respondent.
Open questions may be simply a question such as ‘How would you describe your
manager’s management style, in two or three sentences?’ Or you may structure
them differently, for example:
If I had to sum up the culture at my place of work in a word or phrase,

it would be …
I am/am not enjoying my work because …
Observation
Observation is a key way to collect data, though there are quite different ways in
which to go about it. The core difference is between:
• the kind of observation in which you systematically look for particular kinds of
behaviour, and
• the kind of observation in which you watch informally (but methodically), recording
important facets of what is happening.
The first kind of observation, where you watch for particular kinds of behaviour, is
called structured observation. The second kind, where you are in the situation, take
part, record and watch from within, is called unstructured observation.
If you undertake a structured observation, you make assumptions that the
social world is viewable through a prism that enables the breakdown of social
activity into quantifiable elements – bits that you can count. You, as the observer,
have to define what these bits are. They may be individual pieces of action or
language, such as a child physically touching another child or a teacher using
a particular kind of question. The next thing that a structured observer has to
do is devise some way to count these elements. This can be done in a number
of ways.
Structured observation
Duration recording
Here, the observer measures the overall time that a target behaviour (such as ‘child
out of seat’) occurs. You will end up with an overall time, for example 27 minutes
in a session of 60 minutes.
Frequency-count recording
The observer records each time a target behaviour occurs. You will end up with
an overall number, for example four times in a session. (This is also called ‘event
sampling’.)
Interval recording
You decide on:
• an interval (3 seconds, 10 seconds, 20 seconds, or whatever – depending on the

complexity of what you are looking for)
• target individual(s)
• categories of behaviour (for example, on-task, off-task).
You will end up with data that can be processed in a number of ways. The most
usual way of processing data is to count up the number of times that the target
individual has scored in, say, the behaviour category of interest and then to express
this as a percentage of the total number of possible observations.
Unstructured observation
By contrast with structured observation, an unstructured observation is undertaken
when you immerse yourself in a social situation, usually as some kind of participant,
in order to understand what is going on there. This is the more likely scenario in a
case study. Often this kind of observation is called participant observation, because it
is associated with researchers becoming participants in the situations that they are
researching. Ethnographers, with whom this kind of observation is usually associ-
ated, often discuss social life as if it were a stage, or a set of stages, on which we act
out roles (see the discussion of Goffman’s work in Chapter 3). Unstructured obser-
vation will be geared to understanding how these roles play themselves out on the
stage of life.
As Burgess (1982) points out, the term ‘participant observation’ is a little con-
fusing since it connotes much more than simply observation. It entails talking to
people, watching, reading documents, keeping notes and anything else that ena-
bles you to understand a situation.
Image-based methods
Here, photographs, videos, films, graffiti, drawings, cartoons and so on are used as
data sources. Especially now that digital photographs and videos are more or less
free and disposable, image-based methods provide a powerful extension of obser-
vation and open up a range of possibilities for case study research. Image-based
methods not only slot effortlessly into the illustrative case study but also are a
means of recording observations and eliciting responses from your research par-
ticipants in more or less any kind of case study. There are a number of advantages:
• You can more easily include the person or persons on whom the research focuses.
While observation pure and simple is always from researcher to researched, image-
based methods offer the reciprocal relationship as well – that is, the researched
can set the agenda and the research can be more inclusive. In research with young
people in a secondary school, Cremin et al. (2011) explored the students’ ‘voice’ by
giving them a camera.
• You can capture a social scene far more quickly than you can with notes. The cap-
tured scene will enable you to freeze it in time for your subsequent analysis at
leisure.
• The scene can be captured discreetly, with little input from you, ready for your
subsequent interpretation. Not only is it more subtle than, say, interviewing, but
it can also break the ice and help conversation to start. As Schwartz (1992: 1) put
it, the camera was ‘an excuse to start up a conversation, and the longer I made
photographs, the more people I met’.
• You can adapt your method to your case study ‘process’. For example, in sequential
or diachronic studies (see Chapter 9) you may wish to repeat photographs of the
same situation over a range of times and dates.
• These methods can be blended extraordinarily easily with other methods. Prosser
and Loxley (2008) describe a study about city gentrification that combined pho-
tography, ethnographic fieldwork, grounded theory and analysis of detailed field
notes over a period of 16 years.
• Images can enable the use of the photo-elicitation method – that is, photos
(or other images or videos) are shown in interviews to evoke a response. This
can be particularly useful with

children, with whom it is often
difficult to engage if you limit
yourself to words. Prosser
(1992), for example, showed a
photo of ‘smokers’ corner’ (see
Figure 10.1) to pupils and staff
as a way to compare the pri-
orities of teachers and pupils.
Other examples are online – for
example, PhotoVoice is a multi-
Figure 10.1 Photo-elicitation: smokers’ corner. sensory project with the charity
Reproduced by kind permission of Jon Prosser Sense UK, which uses, among
other things, visual imagery to
help young people with sensory difficulties to express themselves – see https://
photovoice.org/sense/
• Prosser and Loxley also point to the ambiguity of an image and its unpredictable,
almost haphazard capacity to kindle a response that may be quite unexpected.
Prosser (1998) gives an excellent overview of this emerging method.
Measurements and tests

Tests are used to check the extent of something, whether this is reading ability or
blood pressure. Test results will nearly always be given in the form of numbers. Tests
can exist in simple, informal measures devised for a particular and local purpose
as well as in complex standardised forms. An informal test assesses something that
has been taught or a level of pre-existing knowledge. A teacher may devise, say, a
spelling test and give marks out of 20. In the same way, for an evaluative case study,
you may devise a test to assess the extent of learning.
Tests can be either norm-referenced or criterion-referenced:
• a norm-referenced test compares the person being tested with a sample of similar
people
• a criterion-referenced test assesses whether or not someone meets some crite-
rion, irrespective of how well other people perform on the test.
Tests of ability and attainment are usually norm-referenced since they seek to tell
testers how well those tested compare with other people. The driving test is the
usual example given of a criterion-referenced test, since there is no comparison

going on with other people doing the test. It does not matter how many people can
or cannot do a three-point turn – if the person being tested can do it, then they will
get that box ticked.
Official statistics and other numerical data

Official statistics are used surprisingly little in student projects. As I hope I have
shown in several of the examples I have used in this book, however, you can do the
groundwork that will enable you to undertake a case study really easily using these
statistics. For example, in the Hostos-Lincoln case study (page 42) I was able to find
out that this school is a fine example of a school that can ‘buck the trend’, simply
by Googling its intake and attainment figures.
There are many websites now from which you can download data of extraordinary
detail and richness. In the UK, there is the Department for Education’s invalu-
able document Education and Training Statistics for the United Kingdom (available if
you search for the title online). Another excellent resource – particularly for those
undertaking comparative research – is the wide range of statistics gathered by the
Organisation for Economic Co-operation and Development (available at: http://
stats.oecd.org/wbos/Index.aspx?usercontext=sourceoecd). The Office for National
Statistics Neighbourhood Statistics (available at: www.ons.gov.uk/help/localstatistics)
also gives a mass of information about, for example, census data, which includes
accommodation types, numbers of cars, country of birth, distances travelled to work
and much more, related to people and regions in the UK. It is especially useful if you
wish to relate your data to general statistics for a region or nationally.
There are some other useful sites.
• http://gapminder.org is excellent for comparative data between countries. Not

only will it plot, say, countries’ national income against their citizens’ life expec-
tancy, but it will do this dynamically over 150 years. I used this site for the Kerala
case study on page 118. Do try this website.
• Different countries’ census bureaux will offer information on a range of population
data. Search for ‘Population Census Bureau’ for US census data, ‘Census Office for
National Statistics’ for UK data, or similar search terms for the country you are
interested in.
• http://ess.nsd.uib.no provides general social data from different countries in
Europe from the European Social Survey.
Using social media

Many possibilities are presented by social media such as Twitter, Facebook,
WhatsApp and Instagram, not just for data collection, but also for collaboration
with other researchers or your research participants themselves. YouTube videos
may help (though they are of very variable quality), and you may find out new
information on what key academics in your field of study are doing and thinking
from sites such as Academia.edu. You can follow key names and discover what
they are currently publishing. On Twitter, follow hashtags such as #phdchat, which
helps to create a sense of community in the lonely process of doing research.
You can use any of the several forms of telecollaboration now available either to
communicate with other researchers or to gather data from participants far and wide –
nationally and internationally. There are two types of telecollaboration: synchronous
(that is, everyone doing it at exactly the same time), such as in a videoconference
on Zoom; or asynchronous (that is, people not necessarily talking at the same time),
such as in emails, discussion forums, Facebook, Twitter, blogs or wikis. Your university
online learning platform, such as Canvas, may also provide the opportunity for this.
As far as synchronous communication is concerned, consider using a Zoom or
Skype group meeting for a focus group. Your university may provide the opportu-
nity for you to use a platform such as Microsoft Teams.
Forums, which are discussion areas on Internet sites devoted to particular inter-
ests, may be a particularly good way of contacting people who would be willing to
participate in your research. Accountancy Age, for example, allows you to set up
a blog through which you will be able to contact interested readers. Or Mumsnet
hosts discussions – ‘talk topics’ – on a wide range of subjects, from books to educa-
tion and health. Students of mine have posted SurveyMonkey questionnaires in
some of these forums and have received a useful number of replies. Or you might
try doing the same (that is, pasting a link to a survey) in a tweet.
Do remember, though, that the people you contact via Twitter, Facebook, Mumsnet
or any form of social media are most unlikely to be representative of larger populations.

take this …
Remember that this chapter has only been about tools – data-gathering tools – for
enabling you to find stuff out. They do not constitute your case study. Good tools,
however, help you to do a good job, so, while they are not a substitute for your
analytical skill, they can certainly help by providing you with the right kind of
data. Do try to get a grip of the range and variety of data-gathering instruments and
understand them as well as you can. In this chapter we have reviewed the use of …
• interviews
 structured
 unstructured
 semi-structured
• accounts
• diaries
• group interviews
• focus groups
• document interrogation
• questionnaires
• observation
 structured observation
 unstructured observation
 participant observation
• image-based methods
• measurements and tests
• official statistics
• social media.
Having made use of these tools and techniques to gather together a body of data,
you will want to proceed to analysis. To analyse the data you gather you will need
to think about analytical techniques, and these are introduced in Chapter 11.
REFLECTIVE ACTIVITY
Using the example of Nav, the dentist, and his concerns about patient experi-
ence at his dental practice (see p. 155 and Figure 8.6), identify the forms of data
collection Nav planned to use. Think further about the storyboard in Figure 8.6
and see if it can be developed further and in different directions. Are there other
ways that Nav could collect data, either using methods outlined in this chapter,
or using methods you have seen outlined elsewhere?
Further Reading
Bryman, A. (2015) Social Research Methods (5th edn). Oxford: Oxford University Press.
An excellent compendium on research methods and data-gathering techniques.
Cohen, L., Manion, L. and Morrison, K. (2017) Research Methods in Education (8th
edn). London: Routledge.
Cohen et al. give an excellent overview and explication of a wide range of research
methods used in social research, and in particular in education.
McCulloch, G. (2004) Documentary Research in Education, History and the Social

Sciences. Abingdon: Routledge.
Good for understanding the use of documents – archives, records, newspapers and
so on.
Ortlipp, M. (2008) ‘Keeping and using reflective journals in the qualitative research
process’, Qualitative Report, 13 (4): 695–705. Retrieved from https://nsuworks.
nova.edu/tqr/vol13/iss4/8
This article gives some excellent advice on using a reflective journal.
Prosser, J. (1998) Image-Based Research. Abingdon: Routledge.

The bible of theory and practice on using image-based research.
Rose, G. (2016) Visual Methodologies (4th edn). London: Sage.

An authoritative text on interpreting visual culture, which spans an enormous
range of visual material from archival photography to documentary film, websites
and social media.
Salkind, N.J. (2019) Statistics for People Who (Think They) Hate Statistics (7th edn).
London: Sage.
Good on basic statistics, with some information also on not-so-basic statistics.
Smith, E. (2006) Using Secondary Data in Educational and Social Research. Maidenhead:
Open University Press.
An excellent discussion of the use of secondary data.
Thomas, G. (2017) How to Do Your Research Project (3rd edn) London: Sage.
Chapter 7 covers data gathering and goes into more detail than I do here about
interviews, accounts, diaries, questionnaires and observation.
11
ANALYSIS: A TOOLKIT
FOR ANALYSING
AND THINKING IN
CASE STUDY
As I have been noting throughout the book, there are two parts to a case study:
(i) the subject; and (ii) the analytical frame or object. The way in which you develop
the analytical frame takes shape in the context of all of the considerations that I
have already touched upon, and these considerations will require that you think
long and hard about what you wish your case study to be a study of. In Chapter 1 I
gave the example of the choice of World War II as a subject. On its own World War II
would not be a case; only with an analytical frame such as ‘a case study of a “just
war”’ would the subject derive meaning as a case study.
So, the analytical frame – the object – is essential. While the development of the
analytical frame is crucial – it is, after all, at the core of the study – the way in which
you go about actually undertaking the analysis is almost equally important. It is
about how you do the analysis.
There are myriad ways in which you can analyse data. Since we are concentrating
on analysis in a case study here, though, I am going to say a little about methods
that seek a holistic analysis rather than those encouraging a separate analysis of
particular parts. I do hope that you will not feel that these are compulsory or rec-
ommended, since case study, as I have been at pains to point out, is not defined
by an approach, nor by a method, nor even by a set of methods – even if these
conspicuously stress the holistic. Rather, a case study is a container, a wrapper, for
a situation or a set of circumstances and it may contain a range of phenomena to
be analysed. The analysis of those phenomena may, of course, take multiple shapes
or forms depending on the nature of the phenomena, so you may use statistics or
sociograms or network analysis (Thomas, 2017) or whatever, and these may all pro-
vide perfectly satisfactory means of analysis of these different elements.
You will wish to take a line in your study, however, that explicitly frames your
analysis in a holistic context. It is a case study, after all; it is holistic. So, the ways
of thinking about analysis that I offer in this chapter stress a view of the whole
situation. It is, perhaps, rather an unusual assortment and by no means covers
exhaustively the analytical methods that are available, but I have brought these
together as means of seeing patterns and developing connections. They are, of
course, no substitute for your thinking. In fact, on the contrary, they require your
thinking, but what they do is help your thinking progress.
Interpretative inquiry: eliciting themes

You may recall that I mentioned in Chapter 8 that interpretative inquiry seems made
for case study. It’s like love and marriage (they go together like a horse and carriage),
because the starting point of the interpretative inquirer, like that of the case inquirer,
is the indissolubility of the situation to be studied. The interpretative inquirer starts
with the view that situations cannot be fractured into variables. We have to study
the meanings that people are constructing of the situations in which they find them-
selves and proceed from these meanings in order to understand the social world.
So, interpretative inquirers collect data in the form of interview transcripts, infor-
mal observations and so on, as noted in Chapter 8. The tricky bit is in drawing
something meaningful from all of these qualitative data – interpreting them in
such a way that they contribute to a useful analysis. The problem with qualita-
tive data is that they can seem shapeless and without structure. The trick is to
find points of congruence and similarity – places of coherence in the seemingly
amorphous mass of data. Those must not just re-emerge, undigested as ad hoc
quotations; analysis requires more than this.
Interpretative inquirers believe Categorisation, sorting, finding coherence,
that you cannot fracture the social simplifying, synthesising: these are at the heart
world into variables like a geologist
of analysis and there are various means to help
fractures rock.
you achieve good analysis.
Constant comparative method

The constant comparative method is the basic method of interpretative inquiry,
rather like curry sauce is the basic substrate of Indian cookery. There may be many
ANALYSIS 225
bells and whistles that are added to the constant comparative method, but, how-
ever elaborate, it will always be defined by the simple principle of going through
data again and again (this is the constant bit), comparing each element – phrase,
sentence or paragraph – with all of the other elements (this is the comparative bit).
That’s all there is to it, though it has been pimped up by the addition of many
and varied ‘improvements’, which you can find detailed in textbooks on research
methods. You may find some of these useful, but most, in my opinion, make things
appear to be more difficult than they really are.
The basic principle governing the process of constant comparison is that you
emerge with themes that capture or summarise the essence (or essences) of your
data. Taylor and Bogdan (1984: 126) sum up the process thus:
in the constant comparative method the researcher simultaneously

codes and analyses data in order to develop concepts; by continually
comparing specific incidents in the data, the researcher refines these
concepts, identifies their properties, explores their relationships to one
another, and integrates them into a coherent explanatory model.
The general process is as follows:
1. READ. Examine all of your data – read the interview transcripts, diaries,
notes from unstructured observations and so on, look at videos and lis-
ten to audio recordings.
2. COPY. Make an electronic copy of all of your raw data and mark it ‘RAW’.
You now have two copies: your raw data (which you do not want to lose
or corrupt) and your working data files. Keep them separate. In fact,
employ belt and braces: keep a copy of your raw data on a memory stick
as well as on your hard drive. Now, rename your working files, adding
the suffix ‘WORKING’.
3. CODE. Read through your working files. As you are reading, underline
or highlight parts that seem to be important. This underlining or high-
lighting is called coding. If you have audio or video from which you have
transcribed, look at or listen to this as well. So much more comes from
that context. As you proceed, you will get an impression of important
ideas or subjects that are recurring. These can be called your temporary
constructs or prima facie constructs. Make a list of them.
(Continued)
4. READ AGAIN. Read through the data a second time, using the list of
temporary constructs from your first reading to check against. Draw up
a grid, with the temporary constructs on the left and page references to
where the constructs are evidenced on the right. Make notes and obser-
vations on the grid as you do this.
5. SORT AND SIFT. Get rid of any temporary constructs that do not seem to
have been reinforced in the rest of the data. Do not delete the actual data
themselves, though – they may form an important counter-example for
the general themes that are emerging. Highlight these counter-examples
in a different colour in your working data records and keep a separate list
of them.
6. RE-CODE. From the second reading, come up with second-order con-
structs that seem to be a good ‘fit’ with your data. These second-order
constructs should do a good job of summarising the important ideas in
your data.
7. EMERGE WITH THEMES. Look through once more, refining these sec-
ond-order constructs now as marker posts for the final organisation of
your data. Once you are satisfied that these capture the essence of your
data, label these as your final themes.
8. THINK ABOUT THE THEMES. How do they seem to be connecting
together? What matches with what? Are there any unanimous areas of
agreement? Are there any contradictions or paradoxes?
9. MAP. Find ways to map your themes (see below).
10. ILLUSTRATE. Select good quotations or sections from your work to illus-
trate the themes.
These themes, or categories, are the essential building blocks of your analysis.
Remember that the aim in using an interpretative approach is to emerge with the
meanings that are being constructed by the participants (including you) in the
situation.
There are various ways in which you can map your themes to show the intercon-
nections between them. This mapping is often the weakest part of students’ use of
interpretative data. While the identification of themes is important, students some-
times go little beyond this and do not manage to explain how ideas are related to
one another.
ANALYSIS 227
More on coding
As I have just noted, when you are using the constant comparative method you
mark up your data with codes – abbreviations, names, marks and/or colours – that
describe important facets of the data. As I outlined in the points above, you can call
these temporary constructs and second-order constructs – different commentators use
different names for this process of sifting and categorisation. Do not worry about
the different words – they are describing the same process. Your main aim in doing
this is to pick out the important bits, and eventually emerge with themes which
capture or summarise the contents of your data. For example:
Here is a passage from an interview with a teacher about bullying in her

class. The coding here picks out three topics, as follows:
Grey = noticing difference

Light blue = hostility, retaliation or aggression from other children
Dark blue = provocation/‘incitement’ of other children.
Teacher: Oh, we’ve got a little girl in here; she looks different and she
acts different from the others, and they’ll make up a name that they
call her. Um, it seems like every year there’s one child that gets picked
on more than somebody else, because they’re different, because
they might look different, they might act different. Say, for instance,
she’ll say something or she gets very excited about something, they’ll
tell her to ‘Sit down’, ‘Be quiet’, ‘Stop doing that’ – they’re like on her
case all the time. Then you’ve got another boy in here who, well he
loses control of himself, and so he blurts stuff out or yells out, and
the other children will turn around and yell back at him. And out on
the playground they do tend to sometimes get physical. Like with
her [pointing to a desk] I’ve watched them actually walk by this little
girl and purposely bump into her or something like that, but then
even though you’re watching them, the child will turn around and say,
‘Well, I didn’t do that’ when you confront him. And it’s not everybody,
it’s just a few, and it’s a few that have behaviour problems that seek
attention, and they don’t know how to get attention any other way
besides a negative way.
(Continued)
There are no right or wrong answers about how you code: you may choose
to code this passage entirely differently from the way I have done it here: you
will be guided in your coding by the focus of your study.
Ultimately, you need to read into what your codes (and eventually your
themes) seem to be identifying (and a fuller transcript is given in the online
resources). For example:
• The teacher’s comments on bullying put a lot of emphasis on the behav-

iour of the bullied child being different or provocative, ‘bugging’ the
others. Is she almost sympathetic to responses of most of the children
to the bullied children’s inappropriate ways of dealing with stress?
• Only once or twice does she comment on something positive she has
done to neutralise conflict by helping the target child to cope (for exam-
ple, by moving towards them and putting her hands on them).
• There is little sense of the ‘pack’ behaviour of the other children in picking
up deliberately on difference. Can pack behaviour be curbed in some way?
• If any recommendations were to accompany this, they might focus on
the structure of the groups, and ‘buddy’ systems for new children, given
the particular problems that are highlighted about new children.
Theme mapping
While the constant comparative method can identify the themes in a case study, on
its own it does nothing to show the relationships among those themes. You need
to make it clear from your commentary and discussion how the themes interrelate.
Theme mapping makes such interrelationships clearer by presenting an interview
and its themes in picture form.
Theme mapping begins, as does most qualitative analysis, with the constant
comparative method. Once you have established your themes, you go through your
working data files and look for good quotations that illustrate those themes. Then,
in the order that those quotations appeared in the interview, you can put them into
boxes on the page (see Figure 11.1). The page now becomes your ‘map’. You may
also find other quotations that in some way complement or contrast with these
quotations. Put these in boxes on the ‘map’ too. Now, label the boxes with the
names of the themes and draw dashed lines if ideas seem to be connected between
themes, and solid lines with arrows where one theme (the theme towards which
the arrow points) seems in some way to account for or explain the theme at the
other end of the arrow.
1 PERSONAL 2 THREAT/SUSPICION 3 TERRITORY
Response to me Without exception, they (the ones

Sometimes I’m asked to sit in a
varies a lot. It that are difficult to work with) are
particular
depends on the the ones with the discipline
chair – those are the ones
confidence of problems. They seize up and
we’re still fighting
the teacher don’t want you there
4 COMMUNICATION 5 PROFESSIONAL 6 7
We’ve tried to have Sometimes

With some teachers, If colleagues
regular meetings I don’t have the subject
depending on what they’re stick rigidly
but time is a problem. knowledge. It’s quite
doing – like watching a to a text book,
Trying to talk before frightening but
video – you can feel like a there is
or after a lesson is good to know what the kids
spare part little role for you
often impossible feel like
The main problems come

from not having agreement
on classroom management You don’t know We’re most use when
procedures. It sometimes whether to intervene there is a wide range of
isn’t appreciated if we get or not if there is poor teaching and learning
up and move around discipline styles in the class
8 ROLE CLARITY 9 10 ORGANISATION
Figure 11.1 Theme map of an interview

The example I have given in Figure 11.1 is a theme map following the analysis of
some interviews I conducted with support teachers for children with special educa-
tional needs (see Thomas, 1992). Rather than withdrawing children to work with
on their own, these teachers had changed their way of work to be in the classroom,
working alongside the class teacher. I was interested in their perceptions and feel-
ings about their role in the classroom. Did they feel useful? Were there tensions,
and, if so, of what kind? From my interviews and from my own diary, working as
a support teacher, I emerged with a number of themes relating to these questions.
These were: status and self-esteem, territoriality, threat/suspicion, interpersonal fac-
tors, ideology–professional, communication, organisation, school policy and role
clarity. You will see that seven of these nine themes are flagged in the interview
theme map in Figure 11.1. Aside from anything else, the theme map gives a kind of
mini-representation of the interview, because the illustrative quotations are given
in the order that they have appeared in the interview.
NVivo
Remember when you are doing all of this that it is the twenty-first century and
there are many software tools to help you analyse data in the form of language. You
may, for example, find that packages such as NVivo and ATLAS.ti help if you have
a large amount of data. Your university will probably have an institutional licence
that will give you free access to one or both of these packages.
NVivo enables you to code data – documents, PDFs, audio, video – and sort these
data into themes. The process for examining the data is essentially the same as I
noted above for the constant comparative method. In other words, you explore
the data, come up with tentative ideas and code interesting bits, search the data
for similarities, mark these, group them under ‘nodes’, and eventually emerge with
final themes.
Figure 11.2 gives an example of an NVivo page from a project in which I was
involved with my colleagues Liz Ellis and Liz Hodges (Ellis and Hodges, 2014). This
was a project funded by the deafblind charity Sense, exploring the life experiences of
people with Usher syndrome, which is the most important single cause of deafblind-
ness. Forty-two people with Usher syndrome were interviewed and their interviews
were transcribed into a Word document. This document was imported into NVivo,
the data were coded and nodes identified. Figure 11.2 shows those nodes.
You will see that 17 nodes were identified in total; the ‘Education’ node has
been expanded to show its 8 ‘child nodes’. The ‘Sources’ column shows how many
respondents made reference to each node, and the ‘References’ column shows how
many times the topic (that is, a reference coded by this node) appeared in total. So,
ANALYSIS 231
Figure 11.2 Part of a page from NVivo showing nodes and child nodes from the Usher
project (Ellis and Hodges, 2014)
for the ‘Communication’ node highlighted, 22 respondents spoke about something

which was coded into this category, while in total there were 63 references made to
Communication through all of the interviews.
NVivo has some interesting features that can add to an analysis. You can, for
example, make a ‘model’ to show relationships among nodes, in much the same
way as a theme map is drawn.
There is an NVivo ‘Getting Started’ guide, available as a PDF file, which you can
find online. This explains more of the program’s features.
Grounded theory
Grounded theory (Glaser and Strauss, 1967) is a commonly used term to describe the
process of drawing themes from qualitative data – or, even more broadly, the whole
process of interpretative research. Many people say that they are using a grounded
theory approach when what they mean, actually, is that they are using the constant
comparative method. Some people speak about grounded theory as if it is synony-
mous with interpretative inquiry – or even as if all of interpretative inquiry depends
on it. It is not, and it does not.
Perhaps I’m being a bit pedantic, but I find this scooping up of all things interpre-
tative and calling them ‘grounded theory’ a little irritating. Interpretative inquiry
is the Big Daddy; grounded theory is a set of techniques for use in interpretative
inquiry. In fact, many of the assumptions behind grounded theory – for example,
about grounded theory enabling prediction – seem inappropriate and past their
sell-by date now, and I have explored this elsewhere (Thomas and James, 2006).
Lincoln and Guba (1985: 339) make a criticism similar to mine, suggesting that
constant comparison is the kernel of grounded theory worth preserving.
What is helpful about grounded theory is that it offers a neat encapsulation of
the essence of interpretative inquiry – in that it puts a heavy emphasis on the way
that the ideas (the ‘theory’) emerge from your immersion in a situation. This is in
contrast, of course, to the notion that you can go into a case study with fixed ideas
(fixed ‘theory’) about what may be happening. Again, though, it is important to
note that many commentators have queried how far one can clear one’s mind of
existing ideas and theory and allow these to condense, untainted, out of the data
(again, see Thomas and James, 2006). One should, in other words, acknowledge the
extent to which already-established ideas and theory may contribute to the illumi-
nations, interpretations and understandings that come through case study inquiry,
and this is why it is so important to discuss one’s positionality in the methodology
chapter of the case study – in order that the reader can have some understanding of
the provenance of those interpretations.
Coding in grounded theory

So that you can get your head around some of the different words used to describe
similar processes of coding and theme-making in interpretative inquiry, it is worth
mentioning that grounded theorists such as Strauss and Corbin (1990) have par-
ticular terminology for the coding process as it moves towards the establishment of
themes. They talk, in particular, of open, axial and selective coding:
• Open coding is the first stage, going through the data, examining them, comparing
one part with another and beginning to make categorisations. This is, if you like,
the part where you are using your coloured highlighters to mark the text: it is the
identifying of the ‘temporary constructs’ I talked about when discussing the con-
stant comparative method.
ANALYSIS 233
• Axial coding is the second stage, in which you begin to make sense of the open
coding. It is where you ask yourself ‘What goes with what? What can we call
this set of comments? How are these ideas connected to those?’ Here you can
come up with labels for your codes – labels similar to the ‘nodes’ used in NVivo,
or the ‘second-order constructs’ I noted in the constant comparative method
above.
• Selective coding is the final part where the main themes are drawn. As Strauss and
Corbin put it (1990: 116), it is ‘The process of selecting the core category, system-
atically relating it to other categories, validating those relationships, and filling in
categories that need further refinement and development’. It is the stage where
you may be drawing the theme map I discussed above.
Strauss and Corbin also talk about memoing, which is just a posh way of saying tak-
ing notes. As you go through the coding and constant comparative processes, you
are taking notes (memos), in whatever way you choose, about associations, poten-
tial themes, ideas and so on. It may be useful to keep an example of your memos
for use in an appendix to your case study write-up.
Thick description
If you are confident enough, you may wish to forgo the systemisation offered by
the constant comparative method and its derivatives and rely on ‘thick descrip-
tion’. The term ‘thick description’ was popularised in interpretivist circles by the
great anthropologist Clifford Geertz (1975). It refers to understanding a piece of
behaviour – a nod, a word, a pause, for example – in context, and using one’s
‘human knowing’ to interpret it when one describes it.
Geertz (1975: 9) avers that in interpreting meanings you cannot be simply a
‘cipher clerk’. You cannot, in other words, just report things ‘thinly’ without con-
text. The example he uses is the one originally used by philosopher Gilbert Ryle
(from whom he borrowed the idea of thick description), of three boys moving their
eyes. With one, the movement is a twitch, with another a wink, and with the third
a parody of the second one’s wink. What turns the twitch into a wink, and what
makes the mickey-take a parody? As Geertz puts it, ‘a speck of behaviour, a fleck
of culture, and – voilà! – a gesture’. The trick is, in reporting your observations, to
make clear what is going on in a social situation to turn the twitch into the wink, or
into the parody of the wink. How do you know which is which? The interpretative
researcher has to tell the reader.
When you are offering thick description you are also offering an analysis. You are
doing this by intelligently reflecting on the scene, imagining, putting yourself in
another person’s shoes, and genuinely interpreting what the other person is doing.
You are doing this with the knowledge you have not just of people, but also of life
and of the contexts that people inhabit – the stages on which they act. If you opt
for doing this kind of thick description, it helps, then, to know something of the
situation that you observe.
Good examples of this making use of your ready-made knowledge – the knowl-
edge of being human – are found in the great sociologist Erving Goffman’s (1956)
book The Presentation of Self in Everyday Life. Here, Goffman unpacks a storehouse
of examples of how people behave in social life as if they were on a stage, always
conveying some meaning with their behaviour. In doing this, he is employing thick
description.
Word clouds
Think about using something such as a ‘word cloud’ to represent the verbal data
that you collect. I pasted the first few pages of this chapter into a program called
Wordle (available at: www.wordle.net and see Figure 11.3 – other word cloud gen-
erators are available; search for ‘free online word cloud generator’). While the words
from this chapter produce a rather dull cloud, I have to admit, the program could
produce some very interesting comparisons if you were, say, contrasting the tran-
scripts from long, unstructured interviews in a comparative case study. You can
choose the shape of your word cloud – I rather like the idea of the footprint used
for this one. While a word cloud almost certainly will not suffice as a method of
analysis on its own, it can often be a good starting point to get you thinking – one
can view it as a pilot study of an interview.
Figure 11.3 Using a word cloud can give a visual impression of an interview
ANALYSIS 235
Discourse analysis
Discourse analysis is the study of language in social use and it may be particularly
useful in the study of documents, speeches, essays, dialogue, newspaper articles
and so on. As Arnott and Ozga put it, these often overtly or between the lines ‘carry
definitions of problems, reference particular kinds of evidence and argument, and
produce “knowledge” of particular kinds’ (2010: 339).
The field of critical linguistics has sought to reveal the power of various kinds of lan-
guage, in texts and talk, and how readers or listeners are influenced via the use of particular
words and terms and argumentative strategies. It looks at the way in which choices
are made about people to cite, events to exemplify and words used to describe them.
Rather confusingly, discourse analysis is spoken about in different ways in differ-
ent branches of the social sciences. Psychologists think of discourse as the language
that goes on between people, tending to focus on small units of language such as
the choice of individual words and intonations (‘micro-analysis’). Sociologists, by
contrast, tend to think of discourses as forms of language use that define social
relations, and particularly power relations between and among people (‘macro-
analysis’). You will understand from these very different starting points that there
is no one method to discourse analysis.
Fairclough (1995) describes what he calls a ‘critical discourse analysis’ which is
useful for us in the applied social sciences because it combines both psychological
and sociological traditions. Fairclough puts it this way:
a piece of discourse is embedded within sociocultural practice at a num-

ber of levels; in the immediate situation, in the wider institution or
organization, and at a societal level; for example, one can read an inter-
action between marital partners in terms of their particular relationship,
relationships between partners within the family as an institution, or gen-
der relationships in the larger society. The method of discourse analysis
includes linguistic description of the language text, interpretation of the rela-
tionship between the (productive and interpretative) discursive processes
and the text, and explanation of the relationship between the discursive
processes and the social processes. (1995: 97, emphasis in original)
Fairclough’s (2003) three-dimensional model of discourse looks at how texts are …
1. ‘read’ in terms of the language itself

2. examined for the dissemination and consumption of the text
3. interrogated for how they relate to broader social and political themes.
There is no set methodological formula for doing this. For simplicity’s sake, take
it that the general method in ‘gutting’ a document, an interview or some other
sample of language use is broadly the same as in the constant comparative method
outlined in this chapter. The difference is in the focus of the discourse analyst.
Rather than being at the first level on ideas, it is on the use of particular words,
phrases, idioms, similes, metaphors, kinds of rhetoric and so on. How are these
used to construct notions such as ‘education’, ‘health’, ‘order’, etc.? In each case
the discourse analyst will look to see how notions are constructed by the choice of
words and language forms used in a discourse, whether that be an interview tran-
script or a document.
Sociograms
Sociograms are useful for mapping relationships among people. They are often
used in case study as they offer interesting insights into the dynamics of a group.
Notably, they were used by Paul Willis in Learning to Labor (1981). Using a variety of
analytical means, Willis untangled how the students at the ‘Hammertown’ school
in which he was observing developed an antagonism towards school. They devel-
oped what Willis calls a ‘counter school culture’. He intertwined the development
of the theoretical narrative about counter-culture with observations and illustra-
tions from the case study itself. As part of this, he used sociometry.
The sociogram given in Figure 11.4 is adapted from research I undertook for the
children’s charity Barnardo’s on the inclusion of children with physical difficulties
in mainstream schools. The children in one primary school class were asked to say
in confidence who they would like to sit next to or play with.
Sociometry involves asking respondents who they would like to sit next to or
who they would like to work with – or the opposite: who they would not like to sit
next to, and so on. For each choice an individual makes an arrow is drawn to show
the direction of choice. If a choice is returned, the arrow becomes double-headed.
Various features may emerge from the sociogram:
• one-way choices, where someone chooses another but this choice is not reciprocated
• reciprocated choices, where individuals choose each other
• cliques, tight-knit groups of three or more whose choices are all made within the
group
• isolates, those with no choices from other individuals
• clusters, more general groupings
Sophia
Emma Olivia Isabella Nathan Tom
Maya Zoe
Oliver
Jacob
Ava
Noah Max
Lily Jack
Madeline
Leo
Charlie
George
Charlotte
Chloe Harry
Joshua
Henry Dylan
Abigail Luke
Grace
Daniel
Layla
Ellie Ben
Adam
Amelia
Emma
Hayley
Figure 11.4 A sociogram of children’s seating choices

• stars, individuals who have a large number of choices

• power-brokers, individuals who have choices from the stars but not necessarily
many others.
Sociometry involves some significant ethical issues. Serious thought needs to be

given to the wording of questions and how material is presented to those par-
ticipating in your project. You must assure your research participants absolute
confidentiality and anonymity and take great care not to allow participants to see
the papers of others. Pseudonyms are essential, of course, in any write-up.
Systems thinking
I discussed the close ties between systems thinking and the case study in Chapter 3.
Because the aim of systems thinking is to try and avoid breaking up a complex web
of social activity, it fits naturally with the holistic emphasis of the case study. It is
especially useful if there is a problem or puzzle to be solved, which there might be
in an action research project, for example. Given that action research is likely to be
about your own specific situation, the case study will be particularly appropriate.
I have noted that there have been several models developed for using systems
thinking. These have originated in science and technology, but one of the most use-
ful for our purposes has been devised for use in social systems, where the problems
are often taken to be ‘messier’ than those in, say, engineering. It is known as the soft
systems methodology and was conceived by Peter Checkland (1981).
Checkland’s soft systems method is particularly useful in what might be called
‘purposive’ systems – that is, systems with a purpose. You can probably imagine a
broad range of purposive systems in your home or university or work life, such as
a system for taking out the rubbish on bin day, a university system for ensuring
speedy feedback to students on their work, a system for involving parents in their
children’s school. Almost anything where a purpose is involved and people are
expected to do something to achieve the purpose requires a system.
Checkland’s soft systems methodology involves seven stages:
1. The problem is outlined, though unstructured, using as many means as possible to

achieve this – brainstorming, checklists, Lewin’s field analysis (see page 59) and so
on. At this stage, however, there is no attempt to make judgements or analyses. It
is simply about the facts, in as raw a state as possible.
2. Some attempt is made at organising the problem into its constituent elements.
Here a ‘rich picture’ is drawn – literally a picture. It will be a cartoon of the various
ANALYSIS 239
elements and how they fit together and include, for example, relationships and
attitudes – the ‘atmosphere’. Draw stick diagrams of the people, with thought bub-
bles showing their fears and hopes, agreements and conflicts.
3. This is perhaps the most important stage and involves moving into systems thinking –
a bit like going through the back of the wardrobe and into Narnia. Here we have to
write ‘root definitions’, which will say something like ‘a system to do A by means of
B in order to C in the context of D’. The root definition involves six features of the
situation as a system:
a) customers, those who benefit from the system
b) actors, people who transform inputs into outputs
c) transformation from this to that, A to B, inputs to outputs
d) Weltanschauung, the broader context and worldviews
e) owners, people who ‘own’ the problem and want it resolved
f) environment, the constraints that the environment sets up.
Checkland gives this set of considerations the acronym CATWOE, created from the
first letters of the names of the six features listed above. It can be a bit confusing
since, in a simple system, the same person or people can occupy several of the
roles (an actor could be the same as the owner, for example). The root definition
(see above) enables us to think in systems terms about the roles being occupied by
people, the values and objectives that they may hold, who holds power and how it
operates in the system.
4. In this stage, you can think about how things might change. You develop what
Checkland calls ‘conceptual models’. You can be as creative as you like as you do
this, imagining how things would work. You should think about what is a ‘must
have’ in the system and what is simply desirable.
5. You now come back through the wardrobe door and return to the real world. You
compare the ideal, coming from the conceptual models, with the real world. You
ask why the actual and the imagined are not the same.
6. You look to see what changes may be desirable and feasible, on the basis of the
comparison you made in stage 5.
7. Action is proposed to improve the problem situation.
This all perhaps looks a little rigid and mechanical, but I prefer to think of it as
a framework for thinking, using the ‘let’s look at the whole picture’ of systems
thinking. Dick (2002) provides a useful summary of the approach. He says that
Checkland’s method has at its heart:
a comparison between the world as it is, and some models of the world
as it might be. Out of this comparison arise a better understanding of
the world (‘research’), and some ideas for improvement (‘action’) …
[T]he researchers begin with a real-world problem. They study the sys-
tems which contain the problem. Following this, they develop some
models of how those systems might work better … The ‘ideal’ mod-
els are then compared to the actual situation. Differences between the
models and reality become the basis for planning changes.
Checkland provides examples and it is useful to look at his book if you wish to go
into this method in any further detail. Figure 11.5 gives a summary of the method
and Figure 11.6 gives a worked example.
1 The problem 7 Action to improve

situation, the problem situation 6 Feasible,
unstructured desirable
changes …
2 The problem 5 Comparison of 4

situation, expressed with 2
3 Root definitions of 4 Conceptual

relevant systems models
4a Formal system 4b Other systems

concept thinking
Figure 11.5 Peter Checkland’s (1981) soft systems methodology
Drawing storyboards – the nuts and bolts

On page 31, I discussed how to use a storyboard in the development of your case
study, but storyboards are more than merely pictures. Do you remember the scene
in The Terminator where the cyborg reconstructs itself from all of its broken bits and
melted pieces? Well, that’s what the storyboard does for you. It is a bit like a silicon
chip with the instructions on it for the construction of a case study (rather than a
cyborg). It enables you to see all of the disparate elements encased in the case study
wrapper and helps you to understand the connections between them so that they
can be assembled into a meaningful whole. You will be left with not only a mean-
ingful whole but also a whole that is more than the sum of its parts.
ANALYSIS 241
1 The problem No 7 Action to improve

one will take out the the problem situation
rubbish. Dad resents Put into practice Devote time
the fact that teenage changes in 6 to discussion
6 Feasible,
son doesn’t desirable Establish a rota
contribute I’m in the changes …
middle Agree rewards
and sanctions
2 5 Comparison of 4
with 2 Discuss the
comparison
Why me? I’m
Why not being paid
should
I do it?
He’s
younger 4 Conceptual
and er model
than me 3 Root definitions
Customers all in the family Bins emptied,

Appreciate need sharing load
Actors family members
to share load
Transformation full bins to
empty bins
Weltanschauung view that
Task
emptying bins is necessary;
shared
sharing the task is fair
Owner Mum and Dad Bins need to
Environment weekly be emptied
emptying weekly
So root definition is ‘A
family system to empty
bins regularly by relevant
family members in line
with weekly bin collection’
Figure 11.6 Soft systems with a worked example
I have given plenty of examples of storyboards earlier in the book, so I will not
give another now. Here are some tips, however, about the nitty-gritty of drawing
storyboards in Word. Do remember, though, it is the ideas that are more important
than the drawings themselves and they do not need to be done in Word – you could
do them just as effectively with sticky notes.
How to draw a storyboard in Word

You’ll first of all need Drawing Canvas. This lets you treat the canvas as a page, it
keeps everything together and it enables you to draw in connectors that will stay
with your drawing objects.
1. Place the cursor where you want the Drawing Canvas for your storyboard to be.
2. Click the Insert tab at the top of the Word page, then look for the Illustrations
section, click the little arrow under Shapes, scroll to the bottom, and click New
Drawing Canvas.
3.
When you put your cursor
inside the Canvas, you’ll
find that the menu at the
top of the page adds ‘Shape
Format’ at the right-hand
side. (If the Shape Format
menu doesn’t immediately
appear, click Shape Format.)
This gives you options about
shapes to insert.
4. On the Insert Shapes tab, click Draw Text Box.

5. Using the + cursor, drag the text box to make it the shape you want.
6. Write in the text box.
7. Repeat 4, 5, and 6 as necessary for new text boxes for new ideas.
8. To draw connectors between the boxes, go back to the Insert Shapes box (top
left of page) and click one of the connector lines. Use it to draw between the text
boxes. If you make sure that each end of the connector ‘magnetically’ attaches to
the middle blobs of the text boxes, the connectors will stay connected, wherever
you move the text box to.
9. Drag boxes around as you wish, change their shapes or whatever. You’re in
charge.
If you feel life is too short for this, you can draw it by hand and/or use sticky notes,
but it’s not as hard as it sounds. If you commit to devoting a couple of hours of your
life to it, you’ll find it quite rewarding.
Developing your theory

I am including developing your theory in the toolkit for analysing and thinking
because I think of theory as a tool (Thomas, 2007). It is a tool that can be used in
any design frame, not just the case study, but is especially important to it since it
acts as the glue that holds the whole thing together.
ANALYSIS 243
Glue? Tool? Yes, I am mixing metaphors a bit. I think of theory as both, really, but
let’s start with theory as glue.
Theory as glue
I have said that your case study is like a wrapper containing all the bits of your
inquiry. The inquiry is defined by a moment in time, a person, place or whatever,
and your job in this is usually to drill down to the ‘how’ and the ‘why’. In doing
this, you will amass a treasure trove of data that you need to make sense of. For
all of the data to provide useful pieces of sensemaking, you need to connect them
somehow, which is where theory comes in – it makes the connections between X,
Y and Z and holds them together.
It is the ‘holding together’ bit that is important. Perhaps rather than glue, it is
best to think of theory as the sinews and tendons of your study – holding one part
to another, yet allowing them to articulate with each other.
Theory as a tool
The great sociologist Pierre Bourdieu was the person who encouraged me to think
of theory as a tool. He said of his own work (Bourdieu, in Wacquant, 1989, cited in
Jenkins, 1992: 67, emphasis in original):
There is no doubt a theory in my work, or, better, a set of thinking tools

visible through the results they yield, but it is not built as such …
It is a temporary construct which takes shape for and by empirical
work.
For Bourdieu, then, theory exists for a purpose. It is not the aim of the inquiry, it
is not an end in itself, it has to do a job, which is to help explain your findings. It
comes and goes as you need it, so it may be fleeting, like a shooting star, giving you
a flash of inspiration.
This idea of transitoriness is at variance with the idea of theory that is sometimes
held in social science. Sometimes, I think, people assume that theory is the end
point of an inquiry, but Bourdieu’s advice – excellent advice in my opinion – is
that it is disposable. It is the explanation the theory provides that is important, not
the theory itself. When I shave with a disposable razor, the important part of the
process is my silky smooth skin (I wish) at the end, not the razor itself. I don’t, on
completing my ablutions, carry the razor to work with me to show off to my col-
leagues, and, when it’s blunt, I dispose of it.
Another sociologist, Howard Becker (1998:

Drawing out theory is not just about
making links to ‘grand theory’ such 3), suggests that theory is a collection of ‘gener-
as that of Marx or Freud. It is about alising tricks’ – tricks that help in interpretation
making connections, identifying
and sensemaking. This idea may be a little con-
issues and offering reasoned
explanations. fusing in the context of a case study, for I have
been at pains to point out that you cannot gen-
eralise from a case study. By adding the word ‘tricks’, however, I think Becker means
something rather different from the usual meaning of generalisation. He is talking
about our ability to suggest meaningful explanations concerning our findings. How
do the parts fit together? What insights emerge? What patterns show up? Here, gen-
eralising means finding patterns or distilling or finding links between. It is about finding
bridges between ideas.
This can be hard to do if you have not had much experience to give you a
broad resource of different viewpoints and ideas to make bridges between. Using
some of the methods that I have outlined in this book, however (such as story-
boards, systems thinking, constantly comparing your data), you can show that
you have weighed up ideas, looked at them critically and tried to form intercon-
nections in the narrative that you are developing; you have tried to develop
themes. Always try to extend the process of theorising about what you have done
and what you are finding. By doing this you will demonstrate something of not
only what you have discovered in your research but also your ability to be crea-
tive and insightful.
The process is about not just seeing connections in the data you are collect-
ing yourself, but also bringing in and assimilating to your story your reading and
knowledge of the world. What use can you make of your own experience, for
example on a visit to a different country? Always try to see the relevance of one
thing to another and garner and collate thoughts, brainwaves and inspirations as
you do so.
In short, theory is about:
• seeing links between ideas

• noticing where patterns exist
• abstracting ideas from your data and offering explanations
• connecting your own findings with those of others
• having insights
• thinking critically about your own ideas and those of others.
Here’s an example:
ANALYSIS 245
Treatment of the news: The development of

theory from cases
Kayla is following a course in media studies at an American university and on
an exchange programme allowing her to study for a year in the UK.
After a while in the UK, Kayla observes what she considers to be a marked
difference in the way that the media in the USA and in the UK handle routine
national news stories. In the UK, she surmises, there seems to be a reflexively
critical, almost routinely deprecatory tone, with shades of adversarialism
always thrown into the mix.
Her prima facie questions are: is there a difference in the tone of criticality and
disapproval between UK journalism and that in North America, what might be
some reasons for this, and what might the consequences be for public discourse?
She begins in her literature review by:
• contrasting intellectual histories of journalism on each side of the

Atlantic, looking at the influences of the nations’ sizes and political alle-
giance in such different environments
• examining the influence of UK newspaper magnates such as Beaverbrook
and Rothermere on the intensely concentrated market of the UK with its
very high density of population
• comparing this with the historically more distributed and local state of
affairs in the USA and Canada
• looking at the changes likely to be effected henceforth by the advent of
mass communication
• noting the advice given by key figures in the field – people such as Harold
Evans (2000), who cautioned against a reflexive jump to the adversarial
to attempt to inject spice into a story.
Already, then – in this review of the literature and her reflections on it,
informed and bolstered by her own informal observations – Kayla is bring-
ing a theoretical element to her original questions. Already the beginnings
of answers are emerging – a little crystal of an idea forming here and there,
ready to be followed up and built on. These are embryonic theories, ready
for further investigation.
Kayla’s prima facie questions did not need much reformulation, but she
modified them more precisely to address her interests thus:
(Continued)
• Are there differences between the ways that newspapers report general
interest news stories in North America and in the UK?
• If so, how are these manifested in discourse?
• Are differences in discourse likely to impact the nature of journalism and
public discourse more generally?
The first of these three questions cannot be answered by a case study since it
calls for some kind of representative sample being taken of similar stories in
matched newspapers on both sides of the Atlantic. So, such a comparison of
stories in matched newspapers using formal discourse analysis is what Kayla
did to answer this question in the first part of her inquiry. We will not delve
any deeper into this since it is not relevant to our subject of case studies.
The second and third questions, however, are suitable for exploring using
a case inquiry, so Kayla decided to conduct a comparative case study of two
newspapers – the Guardian and the New York Times – over a period of two
months, comparing especially their treatment of similar or broadly equiva-
lent news stories.
Kayla was surprised at the degree to which the actuality bore out her
expectations. Here is one example of different treatments of stories about
something as seemingly innocuous and unadversarial as the weather. The
first is from the New York Times on 6 February 2010, about a particularly
fierce Washington, DC snowstorm:
The hard edges of Washington were softened as the snow recast the
capital of monuments and malls into a postcard town of soft ice cream
shapes that had been statues and aerodynamic blobs that had been
parked cars: the buried machines of a lost civilization.
Contrast this with a report from the Guardian about the unusually snowy and
cold winter in the UK exactly one month before the record of Washington’s
snow:
A Local Government Association spokesman insisted that councils,

which are responsible for gritting many A roads as well as minor and
town roads, were prepared: ‘As far as we are aware there are sufficient
supplies of salt in the country to deal with the current cold snap. How
much grit each council holds depends on local circumstances and how
much bad weather they tend to get.’
Clearly no general conclusion can be drawn from a simple comparison

of the two newspaper stories. Kayla could not suggest that these were
ANALYSIS 247
representative, but the stories do exemplify the general contrasts that

Kayla was noticing in the pieces she was looking at in depth over this
two-month period in these two national newspapers. She noticed what
she described as a ‘defensive, mean and irritable’ tone to much of the
UK coverage, seeking often to apportion blame, even for natural events
such as this. It was not ever-present, and the tone was to be found in the
US coverage as well, but her in-depth analysis certainly bore out the find-
ings of the inquiry in response to her first question – that is, differences
existed in the nature of the journalism undertaken in the two countries.
In making this general finding, Kayla was able to contrast excerpts such
as those given above. In doing this, she was able to point to the sense of
delight shown in the story in the American newspaper and compare this
with the inward-looking tone of the UK piece. She was also able to include
other news breaking in this period, such as the complaint by the director of
the Winter Olympic Games in Vancouver about the treatment of the Games.
His complaints centred exclusively on the UK press. Interestingly, one of
the allegedly caustic stories to which he pointed came from the Guardian,
which had claimed that the Games were in danger of being remembered
as ‘the worst in Olympic history’. The headline in the Vancouver Sun pro-
claimed, ‘Count on the British media to take the low road’, with its first
sentence being, ‘If there were a gold medal for premature Winter Olympic
whining, the British media would be perennial occupants of the middle
podium.’
The stories in these two newspapers were a good theme for forming the
basis of Kayla’s comparison. Using her core case comparison, she was thus
able to reflect on her earlier theorisation following her literature review and
accept or discard some of her ideas. She could also take some of them fur-
ther, accepting the idea of adversarialism having emerged from the tight
UK market wherein newspaper owners could use everyday news to promote
politically loaded messages.
This, then, was the second theoretical element, binding together origi-
nal and emergent ideas with empirical findings and coming up with new
explanatory ideas. These new ideas concerned the oppositional nature of
discourse in the two different environments on either side of the Atlantic,
even where this was not party-based. It affected, Kayla conjectured, all
kinds of discourse and extended even into a kind of anti-intellectualism,
such as the phrase ‘the buried machines of a lost civilization’ from the
New York Times, which, she guessed, would not have been tolerated in
the more cynical discursive environment of the UK. Her discussion came
to be about the extent to which this affected the intellectual climate more
generally (see Figure 11.7) and she sought to discuss this in the context
of further reading.
Theoretical element Effect Comparative, illustrative

of country size, intellectual Theoretical element
case study One UK story
histories – Rothermere and Why should these
about snow and one US
Beaverbrook (adversarial?), differences exist?
story. Textual analysis of
party allegiance, National publishing
each for words of
concentration of markets in UK v. more
confrontation and blame
local in USA?
Oppositional history
of journalism in UK?
Test general thesis about Anti-intellectual
difference in styles by trend in UK?
comparing 20 UK stories
and 20 US stories of the
same events
Idea Confrontation is the key to
broadcast media’s approach to
Idea UK and US treatment discussion. Use of language, ‘war’
of similar stories reveals and ‘battle’ rather than ‘disagreement’;
more of the ‘confrontational ‘admitted’ rather than ‘said’
cliché’ in UK ones
Ideas
Figure 11.7 Developing and using theory
‘Theorisation’ and ‘theory’, then, take a variety of shapes and forms as a study
progresses. There are two stages:
1. setting your analysis in the context of everything that has gone before, especially
your literature review
2. tying findings together – intertwining ideas, seeing patterns – and then knitting all
of this together into a fabric that is called ‘theory’.
In doing all of this, the important thing is to use your brain to try and understand
what is going on – to make sense as you progress.
Elliot Eisner (1991: 238) suggests that theorising is the ‘art of the eclectic’, wherein
we put things together, make multiple interpretations, build bridges and reason by
analogy. He makes a plea for ‘convincing insight rather than Truth’.
Using narrative
Throughout this book I have stressed the importance of the narrative – the
storyline – in a case study.
ANALYSIS 249
In the same way that a story has coherence, integrity and progression, so must a
case study. In looking at the whole you are eschewing the reductionist, fractionat-
ing methods of much social science inquiry – methods that attempt to dissolve the
connecting threads and fibres that hold social phenomena together. In a case study
you must make sense of the whole by retaining the fibres that bind the whole story
together. Those fibres concern time, place, meaning, intention and much more, all
interrelating. The interrelationship makes sense in much the way that a story does.
We cannot take one page of a novel and make very much sense of it. Nor can we
extract and analyse just the sentences that contain a character’s name to work out
his or her personality – we would probably form a very distorted picture of that
character if we did. Rather, each character is understandable only in relation to the
whole story. It is the same in a case study.
Not only this, but the structure of a story itself adds a great deal. In stories there
are assumptions about, for example, motives, intentions, jealousies and kindnesses.
These are things that we all understand from our daily experiences, our understand-
ing of life, and we use this experience and these understandings to deconstruct the
narrative of a case study.
The educator–philosopher Jerome Bruner (1997: 126) goes so far as to say that
narrative is at the heart of all meaningmaking, even if it is scientific meaningmaking:
The process of science making is narrative … we play with ideas, try

to create anomalies, try to find neat puzzle forms that we can apply to
intractable troubles so that they can be turned into soluble problems.
In his famous article in the journal Critical Inquiry, he calls this the narrative construc-
tion of reality (Bruner, 1991: 1). I give in the box below some key ingredients when
using narrative in case studies. In doing this I have borrowed heavily from Bruner’s
(1991: 3) ‘ten features of narrative’, taking away a few of them and adding a couple
of my own. It is an anatomy, if you like, of storymaking. I offer these ideas as ways to
extract every possible drop of juice from the story you are telling in your case study.
Questioning and surprise: Intelligent noticing

and serendipity
You will have ‘finds’ as you construct your narrative. Be prepared to use
these – to use surprise and serendipity. Samuel Johnson (1759/1963: 180)
(Continued)
(that’s the famous eighteenth-century Dr Johnson) put it well: ‘Our brightest

blazes of gladness are commonly kindled by unexpected sparks.’ We should
not ignore these just because they fall outside the bounds of some pre-
scribed method. When inventors and creative thinkers give us an insight into
the ways that they think, it is clear that the generalisation, if it is significant,
is secondary to metaphor, inspiration and imagination.
Storr (1997: 176) gives many examples of this kind of intuition in which
a solution suddenly appears – he quotes from mathematician Gauss, for
instance: ‘Like a sudden flash of lightning, the riddle happened to be solved.
I myself cannot say what was the conducting thread which connected what I
previously knew with what made my success possible.’ The same happens in
social inquiry, and a case study seems to be the ideal vehicle for this kind of
insight, as long as it is enabled by a spirit of curiosity and not snuffed out in
a relentless search for generality.
Heuristic and incremental chunking

Heuristic comes from the Greek heuriskein, meaning ‘to discover’. It is from
the same root as Archimedes’ ‘eureka’. Archimedes, of course, jumped out
of the bath with the solution to the problem of how to tell if his king’s new
crown was made of pure gold or had been adulterated with silver. We now
use the term ‘heuristic’ to mean an explanation that is assumed to be the
best in the circumstances and ‘for the time being’. Heuristics do not have to
be the best for evermore.
How do we get to that eureka moment? How, in other words, are the
elements woven together? What depends on what? Where are there con-
tradictions or paradoxes? This ‘eureka’ of intuition, if dissected, is similar
to Simon’s (1983) tacit processes of incremental chunking. It is the putting
together of related information to make a story. Be confident in doing this.
Narrative diachronicity
Narrative diachronicity? Not the most user-friendly term. ‘Diachronic’ means
changing over time. So, narrative diachronicity means ‘a story that changes
over time’.
Someone doing a case study should be acutely aware of change over
time – and not only in a diachronic study (see page 183). They notice
change as it happens and seek its antecedents and consequences. We
have to find the ‘sequence of steps’, as Becker (1992: 209) puts it, and
understand cause in relation to time, with ‘each step understood as pre-
ceding in time the one that follows it’. In doing this, we conjecture not
only about how one thing is related to another, but also about how cause
ANALYSIS 251
and effect change with time as other variables in a situation also change.
Becker (1992: 209) gives an example:
Causes may be seen to operate, but now it is possible to treat a given

causal variable as operating in different ways (or indeed not at all) at
different steps in the process. In an analysis of heroin addiction, race
might be a crucial variable in explaining exposure to the possibility of
using drugs, but once a person has started to use drugs, race might
play no further part in affecting whether people so exposed in fact use
drugs, or, having used them, become addicted to their use.
So, time may switch on or switch off the potency of a variable, and it is
important to be aware of this since its significance is one of the ready-made,
built-in advantages of the case study. While a reductionist inquiry shows the
patchy significance of a variable only with great difficulty, a case study ought –
by stressing the importance of the diachronic – to show how a variable such
as race (in Becker’s example) may ebb and flow in its significance.
Particularity
In her novel Under the Net, Iris Murdoch (2002) has one of her characters
come out with this line: ‘All theorising is flight. We must be ruled by the
situation itself and this is unutterably particular.’ This is perhaps at variance
with the advice I have been giving about the need for theorising, and there
is an apparent contradiction here. Certainly, if theory is about generalisation,
there is a tension to be resolved.
Without going into the ins and outs of this now, what Murdoch is getting
at is that there is a uniqueness to a particular situation and we should seek
to understand this without what the philosopher Michael Oakeshott (1967: 2)
calls the ‘irritable search for order’. We should not always judge this situation
and its significance by reference to others but by reference to the particular.
This may call for a special effort of will from the student or professional
researcher, given the ever-present desire to establish, develop and refer to a
certain kind of generalising theory among social scientists.
Intentional state entailment

Bruner notes that we should observe not just what people do, but what they
think and feel. It is their beliefs, intentions, hopes, desires and values that
are important. There is nothing different here from the thick description of
(Continued)
the interpretative inquirer (see page 233). It should be the unselfconscious

hallmark of those doing case studies.
Breaching the canon … and counter-factuality

Using the rather ugly word canonicity, Bruner (1991: 11) asks us to consider what
is usual, normal, in a line of reasoning or the unfolding of an argument or a story.
This is the ‘canon’ – it is the realm of what we expect. By suddenly throwing a
spanner in the works, we can jolt our readers into thinking differently – perhaps
differently enough to imagine an alternative explanation or a different state of
affairs.
As narratives, case studies have the function of letting us understand and
recognise how they differ from what is normal or expected. They let us guess
how and why this may be so. Feyerabend (1993) suggested that we may use
counterrules. These are hypotheses that contradict well-established thinking
of one kind or another. Kuhn (1970: 52) suggested something similar in what
he called the ‘awareness of anomaly’.
Such a breach can be engineered by the deliberate introduction of an
imaginary change of understanding, which is why I have taken the liberty
of adding ‘counter-factuality’ (Kahneman and Tversky, 1973; Mandel et al.,
2005) to the heading above using Bruner’s phrase.
Counter-factuality is the imagination of a different state of affairs. This
might exist if a particular event (usually, though not necessarily, a key one)
had not occurred or if there had been some other outcome from it. Counter-
factuality exists, in other words, in ‘what if …?’ questions.
This idea has been employed by historians (Ferguson, 1999), though it
has not been widely used as part of the method for carrying out a case study
in the social sciences. I would like to see it used more. Ragin (2007) has
discussed its use, as has Lebow (2007). Interestingly, Ruth Byrne (2005) dis-
cusses how central a role counter-factuality takes in our everyday reasoning,
showing how different kinds of reasoning – ‘what if …?’, ‘if only …’, ‘even
if …’ – can be refracted through the counter-factuality lens. If only more
case study inquiries (or, indeed, inquiries in general) used such aids to the
imagination …
Context sensitivity and negotiability

Perhaps in the same way that Barthes (1974) talks about ‘writerly’ texts,
in which the meaning is (perhaps confusingly) created by the reader (as
distinct from ‘readerly’ texts), so the assumption should be that the inter-
pretation of the case is embedded in the inquirer’s (and the reader’s) own
experiences.
ANALYSIS 253
Interpretation is personal. It is sensitivity to context that enables readers

to make sense of the narrative of the case and agree or disagree with the
researcher. As Bruner (1991: 17) puts it: ‘You tell your story, I tell mine, and we
rarely need legal confrontation to settle the difference.’
Analogy
We make sense of the unfamiliar by reference to the familiar, drawing
likenesses between one situation and another. We use our own knowl-
edge to do this – our ‘common sense’. Our understanding is based on
myriad personal interpretations we weave together into meaningful sto-
ries that help us to make sense of similar events and situations, similar
plotlines.
The narrative that a case study lets you draw can be the ideal frame for
enabling such analogy and metaphor. Tavor Bannet (1997: 655) discusses
analogy and, interestingly, says that it is a ‘method of reasoning from the
known to the unknown, and from the visible to the speculative’ by carry-
ing familiar terms and images across into unfamiliar territory. It is like a
form of translation, ‘a way of transporting something from place to place,
from old to new, from original to copy’, and we can therefore move from
one context to another. We bring together, juxtapose and see similarities
across contexts.
In discussing the importance of narrative in the case study, Abbott (1992) suggests
that we should always be seeking what he calls ‘causal narrativity’, but I personally
prefer to suggest that, in social science, we are, as with ‘theory’, making connec-
tions rather than trying to find causes. Becker (1998: 60–1) puts this very nicely:
he suggests that the use of the word ‘cause’ is misleading in social research, given
its complexity, so we should see ourselves as seeking narrative rather than a cause:
Assume that whatever you want to study has, not causes, but a history,
a story, a narrative, a ‘first this happened, then that happened, and
then the other happened, and it ended up like this’. On this view we
understand the occurrence of events by learning the steps in the process
by which they came to happen, rather than by learning the conditions
that made their existence necessary.
It is the case study that enables such creativity – allows you to suggest these bridges
and passages between ideas.
Think drama
In Chapter 3, I looked at the ways in which some sociologists, such as Erving
Goffman, have organised their views of the world around dramaturgy. This is a view
in which life is seen as a set of dramatic performances and has been used by not just
Goffman but also a range of interpretivist social scientists. They talk about actors,
roles and stages in their discussions of the interactions that take place between
people.
If the meat of your case study is about people’s interactions and you decide to
adopt an interpretivist approach (see page 159), how do you take it forward? I think
one of the best ways to make this happen is to extend the theatrical analogy. Think
of the substance of the case study as a play, a drama, complete with:
• a script, plot and leitmotif (provided by life and the world)

• a stage and scenery (the wider context)
• a dramatis personae (a list of characters, institutions or agents)
• a director (you)
• an audience (your readers).
The likening of a case study to a play is useful, I think – not just because it stresses
the bounded nature of this study (the boundaries in the case of the drama being
provided by the script of the play and the walls of the theatre) but also because it
emphasises the narrative of the study. In a play, a story is told – of a drama, with
a beginning a middle and an end (unless it is Waiting for Godot). There are argu-
ments and links intertwined through the story – a plot, with questions, premises,
intrigues, subterfuges. Characters come to life as they deal with the obstacles that
the plot puts in front of them. The case study is similar on all these counts.
The analogy is not perfect, however, since in a play the work is fictional and in a
case study the focus – the person, event, phenomenon or whatever – is there. It is
real. If it were not real we would not be interested in studying it.
So, there are differences and it is interesting to pursue the dramaturgical meta-
phor for those differences. In a case study, the script is not written by a playwright,
but is ‘written’ by life, by what has occurred. Anyone who has watched a modern-
day production of a Shakespeare play will realise, though, that a script can be inter-
preted in such a way that a familiar play becomes almost unrecognisable. Directors
examine scripts for meanings and impose their own interpretations on them. In the
same way, the script provided by a case study is interpretable. Actually, the ‘facts’
of the case are far more fluid than a play’s script since it will be for you (as director)
ANALYSIS 255
to decide how they are constituted and how the

Our social lives can be thought of as
various elements intertwine. a series of stages on which we act
They are, however, the raw materials and it is out different roles. This metaphor
can be useful in the analysis of a case
up to you to knit them together into a convinc-
study, thinking of the parts people
ing narrative. You become the narrator or the play in different contexts.
playwright, given the storylines and subplots
provided by life.
Running through the script will be the plot, which will comprise different ele-
ments involving the intentions of the characters and their responses to each other
and the twists and turns of events.
There will also be a leitmotif – the background milieu, the context, the themes
that provide the backdrop against which the action occurs.
We can also see the stage as the confines or boundaries provided by life. Here we
can return to the case study as a container, in the same way that the stage confines
the action of the play. There has to be a boundary to a case study and it is up to
you to set it.
The dramatis personae is perhaps the most important part – that is, the charac-
ters, whether they be people, institutions or natural phenomena, such as Hurricane
Katrina. What are these characters doing? What are their origins? What, if they are
people, gives them meaning and purpose? How far do particular characters react to
challenges or make their way around problems?
You are the director and, to an extent, the playwright – the latter because you put
the people and events together to tell the story as you see it unfolding (the raw
materials have been provided by the situation, but you are the one who sees how
things fit together). So, you develop and write the plot; you provide the connective
tissue.
In doing this, there is an onus on you not simply to be what the anthropologist
Clifford Geertz (1975: 9) called a ‘cipher clerk’. In other words, you have to use your
imagination for the raw materials available to you.
It is easy to forget the audience because its members are not direct participants in
the study. They are one of its most important elements, however. Most research –
action research excepted – is principally about communicating new information to
others. Many students fail to consider the readers, which is a great mistake, given
that the main readers of students’ theses or dissertations will be markers. Students
make the mistake of assuming that if all of the facts are down in more or less the
right order, this will suffice.
It won’t – and it particularly won’t for a case study. Even in a report for an experi-
ment in physics, the structure has to be clear and readers should be able to go
through without a double take – without, in other words, thinking, ‘Now, what on
Earth does this sentence/this paragraph/this chapter mean? How does it fit? What’s
it doing here?’ In my long life, I have read many dissertations and this, sadly, has
been an all-too-frequent set of questions in my mind while reading them.
For the readers, reading your report should be like eating a bowl of their favourite
ice-cream. It should be smooth, enjoyable, have nice gooey bits that are particularly
interesting and, when they have finished, it should leave them thinking ‘I enjoyed
that, mother. Is there any more?’
Actually, it should be better than a bowl of their favourite ice-cream because it
should get them thinking – and there is nothing better in life than thinking.
There are various techniques that can be adopted when writing up your case
study that will make it more interesting and more polished – I shall discuss a few
of these in Chapter 12.
Being intuitive and imaginative

[We need] ways of expanding the reach of our thinking, of seeing what
else we could be thinking and asking, of increasing the ability of our ideas
to deal with the diversity of what goes on in the world. (Becker, 1998: 7)
In Chapter 1, I noted that Einstein had said of the process of science, ‘There is no
logical path, but only intuition’ and this sentiment – with the obsession of social
science to be methodologically exact, rigorous and innovative – has sometimes
been forgotten. The philosopher Karl Popper (1968: 32) made a comment similar to
Einstein’s: ‘there is no such thing as a logical method of having new ideas, or a log-
ical reconstruction of this process. My view may be expressed by saying that every
discovery contains “an irrational element”, or “a creative intuition”.’ He proceeds
to liken this to Einstein’s comment (in an address on Max Planck’s 60th birthday)
that scientific advance depends on ‘intuition, based upon something like an intel-
lectual love (Einfühlung) of the objects of experience’.
So, it is intuition and intellectual love that we have to nurture. A tough call, per-
haps, since intuition and intellectual love are qualities that do not grow on trees,
nor do they lend themselves to bullet-point breakdowns. Given the elusive nature
of intellectual love, imagination and intuition, how should we consider them and
how should we foster them?
Where to start? Sadly, thinking about how we should foster an intuitive mindset
has not been a central consideration in the social sciences. The social sciences have
instead become obsessed with correctness, reliability and design exactitude, at the
expense of imagination and intuition.
ANALYSIS 257
With the paradigm wars that I spoke about on page 51 it is almost as though a
false opposition has been set up wherein one cluster of ideas – positivist, normative,
experimental, objective – has been set against another – interpretative, subjec-
tive, intuitive. This leads to a regrettable hunkering down and a defensiveness in
each of the positions. It leads ultimately to an inward turn, wherein there is more
concern for process than there is for questions, arguments and conclusions. The
consequence is that curiosity in inquiry is diminished and research becomes less
interesting.
As astute critic of social science, Stanislav Andreski (1972: 108), says that how we
achieve creativity
is just as much a mystery [today] as it was in the days of Socrates: all

that is known is that, in order to conceive fruitful original ideas, one
must have talent, must immerse oneself in the available knowledge,
and think very hard.
Is Andreski right about this? Can we really not offer useful advice on creativity and
intuition?
In large part I think he is right. I quoted the first line from Wittgenstein’s Tractatus
right at the beginning of this book, and he said such beautifully gnomic things that
I cannot resist quoting from him again. This time, I shall quote the last line, in the
context of Andreski’s advice (Kenny, 1993: 31): ‘Whereof one cannot speak, thereof
one must be silent.’
How true, though I think Wittgenstein was saying a little more than just
‘Keep your mouth shut if you have nothing to say’ (this is something I have
often wanted to stand up and shout out at conferences, but I have never had
the courage). Indeed, Frank Ramsey – philosopher, genius, admirer and transla-
tor of Wittgenstein – after translating the Tractatus (at the age of 19!) suggested
in a subsequent commentary that what Wittgenstein was getting at here was
‘What we can’t say we can’t say, and we can’t whistle it either.’ Ramsey died
at the age of 26. How tragic that someone so gifted and so funny should die so
young.
I cannot do better than Ramsey’s explanation about whistling, save to offer a
very ungnomic deciphering of what these superbrains were going on about. It
is this: some things are beyond our words; they exist in what the philosopher
Michael Polanyi (1958) has called ‘tacit knowledge’. Included in tacit knowledge
are your knowledge of how to ride a bicycle (if you can) and your intuition about
how to solve difficult puzzles. What the case study offers here, then, is the capac-
ity – or, better, a predilection – to see wholes rather than fractured parts and, with
this predilection, we can perhaps have some expectation of a greater emphasis on

intuition than is offered otherwise.
Assuming that you’re still with me and you don’t think I’ve bicycled off into
an academic’s cloud-cuckoo-land wearing odd socks, where does all this leave us?
It leaves us with a view of certain things being very difficult to teach – intuition
being one of them. If I were not writing a textbook, I would leave it at that, but,
since I am, I will proceed with one or two ideas that I have gleaned from Richard
Sennett.
Sennett (2009: 211) talks about intuition coming in three main stages:
1. from imagination, depending on looking back on sensations already experienced –

depending, in other words, on memory
2. from establishing adjacency, similar to analogy, where two unlike ideas are brought
together
3. from surprise, which comes from ‘dredging up tacit knowledge into consciousness
to do the comparing’.
All of this is similar to the ‘imagery’ that Becker (1998: 12) speaks about. He suggests
that, in studying society, we start with images and end with them. Our aim ‘is the
production and refinement of an image of the thing we are studying’. The rest of
Becker’s book, Tricks of the Trade, is excellent further reading on this.
Well, I haven’t done very well in advising you how to be intuitive, have I? No.
Perhaps, then, I should sum it up by saying that we all have intuition – it’s part
of the human condition – and when you are doing a case study, treasure it and
don’t let it become suffocated by concern over what you consider to be correct
procedure.

take this …
The analysis is the most important and enjoyable part of your case study project.
Here, you will be tying everything together, intertwining elements so that argu-
ments make sense and have integrity. You will be looking back to the original
questions that you posed, seeing how they changed in the context of your read-
ing of the literature and your preliminary inquiries and assessing how well your
project – using whatever methods of analysis – has answered those questions.
The richness of your analysis will be determined in large part by how much time
and thought you are able to devote to it, but I have offered in this chapter some
ANALYSIS 259
ways in which you can give integrity and coherence to your analysis. To review
some of these ways of making an analysis:
• the constant comparative method offers a means of drawing themes from inter-
pretative data
• theme mapping allows you to present in diagram form the ways in which themes
may interconnect
• packages such as NVivo enable the easier management of coding and theme-
making where large amounts of qualitative data are involved
• ‘grounded theory’ offers a specialised operationalisation of an interpretative
approach
• ‘thick description’ captures the essence of interpretative analysis in its synthesis
of reporting, reflecting and meaningmaking
• systems thinking lets you understand how separate bits of behaviour may be inter-
connected
• the storyboard offers a way of navigating along different potential routes to
develop your argument
• you have always in your analysis to be thinking of developing theory – it is about
seeing links, noticing patterns, abstracting ideas from your data and offering
explanations
• thinking about narrative – or how the events in your study can be seen drama-
turgically – can help to give structure to the events that unfold. In doing the
analysis, it is important to understand how you may be both using and develop-
ing your theory and how you can be intuitive and imaginative in making sense of
your findings.
REFLECTIVE ACTIVITY
In case study we are usually looking at data as words or images – heard, seen,
interpreted and (possibly) understood. Many of the data analysis strategies and
techniques outlined in this chapter involve the summarising of these words and
images and/or their transformation into a different form which makes them eas-
ier to understand, which interprets them, or which emphasises their relevance to
the topic being studied. Some methods are simple, while others are more com-
plex. One of the simplest starting points I have outlined in this chapter is the
word cloud, and while I would not recommend the word cloud as a sole means of
(Continued)
analysing data, it offers a good starting point for thinking about the use of words,
interpretations and themes.
Try pasting a page or more from some writing you have done – perhaps an
essay or other assignment, or a transcript from an interview – into a free online
word cloud generator. Does this give you any clues about the themes covered in
your writing or in your respondent’s interview response? What clues does the use
of words in the passage give you for further analysis?
Further Reading
Becker, H.S. (1998) Tricks of the Trade. Chicago: University of Chicago Press.
Becker’s take on research is refreshingly independent of formulae and bullet points.
He encourages readers to use their own resources of imagination and intuition to
make shape and sense out of their data.
Bruner, J. (1991) ‘The narrative construction of reality’, Critical Inquiry, 18 (1): 1–21.
The essence of a scientific approach, says Bruner, is in meaningmaking, while
narrative, with its ‘grammar’, its rules and its structure, provides a vehicle for
meaningmaking.
Eisner, E. (1991) The Enlightened Eye: Qualitative Inquiry and the Enhancement of
Educational Practice. New York: Macmillan.
Eisner has some interesting ideas on what constitutes theory in qualitative inquiry.
Fairclough, N. (2003) Analysing Discourse: Textual Analysis for Social Research.

London: Routledge.
Fairclough’s is a classic text on discourse analysis.
George, A.L. and Bennett, A. (2005) Case Studies and Theory Development in the Social
Sciences. Cambridge, MA: MIT Press.
Taking quite a traditional view on theory, the authors give a perspective on the case
study that is located principally in political science. It thus draws on a different dis-
ciplinary base from most commentary on the case study and offers some interesting
discussion on comparative case studies.
Jorgensen, D.L. (2020) Principles, Approaches and Issues in Participant Observation.

New York: Routledge.
ANALYSIS 261
Jorgensen provides an excellent overview and includes useful examples of up-to-

date analysis and iconic disputes in the field of qualitative inquiry.
Loader, D. (1997) The Inner Principal. London: Falmer.

Loader tells stories from his own experiences. He compares the outer and inner
influences on change.
Taylor, S.J., Bogdan, R. and DeVault, M.L. (2016) Introduction to Qualitative Research
Methods: A Guidebook and Resource. Hoboken, NJ: Wiley.
A key text, and excellent on the constant comparative method, with some fascinat-
ing examples of real-life qualitative inquiries.
Wright, T. (2010) ‘Learning to laugh: a portrait of risk and resilience in early child-
hood’, Harvard Educational Review, 80 (4): 444–63.
An excellent and moving study, told in a narrative form.
12
WRITING YOUR STUDY
Ultimately, your case study is written up as a research project, dissertation or thesis.
So, it needs to conform to what is expected of one of these. How, then, should it
be structured, and is it different from the balance and shape of the write-up of a
project that follows a different research design?
Structure
The following gives a breakdown of a typical dissertation or thesis. Be aware, how-
ever, that there are few ‘typical’ dissertations; all are different, and yours may be
rather different from this. For example, a literature review may comprise not one,
but two or more chapters. The findings and discussion may be conflated into one
chapter or expanded to form several, depending on the nature of your work. A case
study will often be unusual in its structure because of its typically narrative form,
but you should remember that there are essential elements to any project that must
be incorporated in the write-up. Here is a summary.
You need an introduction that explains how you came
to be interested in the topic of the study. Here, you will lay
Chapter 1
out your prima facie questions (see page 30).
Introduction
You need a literature review that looks at what other
inquiries have been done on this or related topics and helps
you to understand their contribution to your own ques-
Chapter 2 tion. As I explained in Chapter 2, this enables you to refine
Literature review
and reformulate your prima facie research question(s). It
may comprise more than one chapter if there are different
elements of the literature that you feel warrant their own
treatment.
WRITING YOUR STUDY 263
You then need a methodology chapter to explain why

Chapter 3 you have chosen to conduct your research in the way you
Method have. That is to say, why have you done a case study? You
also explain here why you have picked your chosen design
pathway (see Chapters 2 and 6–9) and why you have used
the data-gathering techniques and analytical methods that
you have (see Chapters 10 and 11). Here, you can also set
out details of the research setting in which you have cho-
sen to work: who, where, what, when, how many and so
on? Alternatively, given that the setting in which you are
studying is likely to be especially important in a case study,
you can make this into a separate, subsequent chapter all
on its own, rather than being incorporated in the method
chapter.
In this chapter you set out your findings – what you actu-
ally did and found. This and the next chapter really are the
Chapter 4
Findings ‘guts’ of your research and the write-up, and in a case study
they may be conflated. In other words, findings, analysis
and discussion may all occur together in one combined
chapter. If this is the case, you may wish to break up this
‘mega-chapter’ in some other way – perhaps by subject or
theme.
In the analysis and discussion chapter (which may be
Chapter 5
divided into two chapters) you offer an analysis of your
Analysis and
discussion findings using the tools for analysing and thinking out-
lined in Chapter 11. You discuss this analysis in the light of
your research questions and any issues highlighted in the
literature review.
Your final chapter, the conclusion, draws together
Chapter 6
Conclusion the threads and assesses how well the question has been
answered by your inquiry.
Figure 12.1 shows the rough proportions (in terms of numbers of words) given to
each of these sections, though I should warn you that this is only indicative – some
case study write-ups will have a very different balance. As I have noted, your field-
work, findings, analysis and discussion may, in an interpretative case study, all be
merged together because you will be reporting observations almost simultaneously
with discussing them.
Is the structure of a case study different from that

of other research?
In a case study write-up, it is especially difficult to decide on the balance of the ele-
ments I have just outlined. Because of the holistic nature of the worldview taken when
doing a case study, it is sometimes less than easy to divide a write-up into sections.
Conclusion Introduction
5% 5%
Analysis and Literature

discussion review
30% 30%
Findings Method
15% 15%
Figure 12.1 Elements of the write-up of a formal, traditional study and their rough
proportions
In a formal study, you can sometimes take a straightforwardly linear approach,

wherein findings precede the analysis, which precedes the discussion. For a case
study, however, you may not want to separate out sections in this way since all the
time you will be testing out your emerging findings against your thoughts. If this is
the case, it is inappropriate to impose a strict line between analysis and discussion
since one infiltrates and merges into the other. The differences between the differ-
ent kinds of study are shown in Figure 12.2.
Formal linear study Findings Analysis Discussion
Case study Findings, analysis and discussion
Figure 12.2 Different kinds of analysis and discussion

Writing up your case study

Why is research done at all? It is done in the spirit of inquiry – to find things out.
An essential part of this is the communication of the findings of that inquiry to
others – telling other people. There is not much research that is of benefit to only
the person undertaking it. Perhaps simply keeping the findings to yourself is all
right with certain kinds of action research, but nearly always the communication
of findings to others is an essential part of the inquiry process.
Certainly this communication is a necessary part of any university-based research,
when the people you will be communicating your findings to will be principally your
tutor and other markers. The audience may be wider than this, though. It may include
fellow researchers, other students, colleagues and other interested professionals – even
the general public or local politicians.
Because it is about communication, your write-up has to be interesting. Indeed,
interest is the key and you will not establish it simply by dredging up as many facts
as possible and flinging them down in front of your readers in more or less the
right order. This is true for any kind of research reporting, but with a case study it
is especially true.
A case study is not about finding facts but gathering evidence. This evidence,
though, is gathered for the purpose of developing an argument, and there are two
important aspects to this. One is the way in
which your argument is constructed and the Avoid the ‘mushy heap syndrome’
other is about the critical reasoning that you when writing up your case
study – that is to say, avoid an
demonstrate. I shall come to these in a sec- undifferentiated list of quotations
ond, but let us pause for a moment to consider and reportage with little
the structure of a case study, for, alongside its commentary or analysis to glue it
together. Remember, you need
undoubted strengths, there are some signifi- narrative, explanation and theory.
cant risks associated with adopting it as a frame Make it all hang together.
for your work.
The risks come from your ‘freedom to roam’ when doing a case study. All research
has fences around it – barbed wires that define the territory and its integrity, but
also have the effect of holding you in place. With some research there are so many
fences, enclosures, and enclosures within enclosures, there is hardly any room for
the researcher to roam at all. The fences structure the research so that it meets the
standards expected of it so that its conclusions can be said to be valid. For exam-
ple, your choices in terms of how to conduct an experiment are constrained by
the limited number of established designs; the conduct of surveys is circumscribed
by expectations about samples and the nature of questioning. So, these kinds of
research have formidably sturdy fences around them.
The methods that may be used inside the case study wrapper may be constrained by
the same kinds of fences, but the case study itself is defined more by your own imagina-
tion than methodological boundaries. You thus have freedom to roam, without fences.
Marvellous. With this freedom, though, the onus is on you, the researcher, to provide
the structure and meaning for the work. With a case study, at its worst, this can lead to
the mushy heap syndrome to which I referred earlier (see page 265). Sometimes, what
goes under the name ‘case study’ can amount to not much more than an undifferenti-
ated collection of thoughts and quotations from interviews with very little in the way
of glue to hold the whole thing together. This is what you must avoid.
The point I am trying to make is that the quality of a case study cannot be guar-
anteed by simply following a procedure correctly. Instead, it is determined by your
imagination, the way in which you construct your narrative and theory, your criti-
cal reasoning and the way in which you develop your argument using evidence.
The whole of this book, really, has been about your imagination, so I will not dwell
on this now and I have discussed narrative and theory in Chapter 11, but it is
important to say a little more about critical reasoning and argument.
Employ critical reasoning

Your findings, the original knowledge that you produce – the guts of your research –
are, of course, important. Almost as important, though – particularly when doing a
case study as a student – is your attitude to knowledge.
In any kind of inquiry, not just a case study, it is not so much about the amount of
knowledge that you can show you possess, but, rather, whether or not you can show
how you have acquired evidence and analysed it and argued from it with a critical
frame of mind. Your approach to knowledge should always be as a sceptic – you
should show appropriate suspicion and doubt. You will be expected to be aware that
there are always different ways to interpret an observation, different ways to argue
a case, different interests at play in any argument. In short, you will be expected to
recognise – and demonstrate that you recognise – that truth is hard to come by.
Critical awareness, however, is not just about spotting bias. It is about an aware-
ness that knowledge is frail; it is about recognising that you should approach
everything you read and hear with a questioning mind, always asking yourself
whether something could have been done differently. When you read anything –
any piece of research or any scholarly argument – you should be asking yourself
questions such as this:
• Are there any assumptions, overt or hidden, that underpin the research? Does it
seem to start from a particular position? For example, much research in the area of
special education has started from the position that ‘children with special needs’
are an identifiable group or set of groups. This approach is now taken to be sim-
plistic. If you read around, you will get a rounded picture of the area that you are
studying, taking in the controversies and disagreements.
• Are there any vested interests at play? Was the research funded, for example, by
an organisation that may have had an interest in particular findings being made?
• Might the writers’ objectives in undertaking the research sway their reasoning in
some way?
• Would different methods have yielded different findings?
• What sources of information are being drawn upon – is there evidence of balance,
or are sources being ‘cherry picked’?
• What is the quality of the data being drawn upon – are they from good primary
sources?
• Is the writer’s reasoning sound – so, if you were arguing with the writer, what would
you say? (But ask yourself also how much validity your own criticisms would have,
and whether you yourself are likely to be swayed by tradition, sentiment or vested
interest.)
You will need to be tentative about conclusions that you feel you are able to make.
Try to avoid phrases such as ‘this proves’ and, instead, express yourself in phrases
such as ‘this indicates’ or ‘the evidence suggests’ or ‘points towards’. Try to use
phrases that moderate the strength of an assertion, such as ‘tends to’ or ‘one might
conclude that’. Academic writing is an area where it pays to be tentative: no one
will give you high marks for making unwarranted claims about your own research
or for innocently swallowing the reports of others.
Doubt everyone’s findings, even your own. Many great minds have attested to
the importance of this. René Descartes (1647/1996) said that ‘Doubt is the ori-
gin of wisdom’, summing this up in the aphorism ‘De omnibus dubitandum’
(meaning ‘Doubt everything’), which Karl Marx took as his own motto for life.
Similarly, the great biologist J.B.S. Haldane concluded his famous essay Possible
Worlds (1928: 224) with the words ‘science has owed its wonderful progress very
largely to the habit of doubting all theories’. Jeremy Paxman summed up the
sentiment a little more demotically in his personal rallying cry, ‘Why is this lying
bastard lying to me?’ Actually, Paxman claims that what he meant was that he
always approaches people with ‘a degree of scepticism, asking “why are they say-
ing this?” and “is it likely to be true?” … Scepticism is a necessary and vital part
of the journalist’s toolkit’ (Wells, 2005). Scepticism is a necessary and vital part of
the researcher’s toolkit, too.
The great philosopher of education John Dewey (1920/2004) also championed

scepticism. He said that we should be suspicious of certain kinds of thinking,
particularly those arising from tradition and authority. We should be cautious in
trusting the reasoning of others; we should think for ourselves. We should also be
especially wary of any line of reasoning (in others or in ourselves) that comes from
a strongly held opinion. He promoted reflective thought, which is being sceptical
about our thoughts and always looking for evidence for a line of reasoning.
So, critical thinking is not just about approaching the views and reports of oth-
ers with scepticism. It is also about criticality in your own thinking, for example
when observing something or listening to someone’s account of an event. Critical
thinking is about:
• going beyond surface appearances and digging down, trying to understand the
meaning of an observation or an utterance
• including consideration of the context (what
is happening, what went before, what were
Be critical. Remember: ‘De omnibus the consequences?)
dubitandum’ or ‘Why is this lying
bastard lying to me?’ • uestioning your first assumptions, connect-
q
ing with the ideas of others.
Develop your argument

Much of this book has been about developing the argument of a case study, stress-
ing the importance of coherence, integrity and the need to weave a storyline – often
with a storyboard – from the information that you have collected. While I do not
want to go into that again, I do want to stress something that I have not touched
on so far, which is the way in which your argument is actually built – the chassis, if
you like, of your argument.
There are two main elements to this chassis that, in my experience, are often
weak in research project write-ups:
1. establishing the problem to be solved

2. establishing the argument itself.
Let us look at these in turn and see what each should contain.
The problem
The problem must be the sandwich filling between two elements: the context and
the response:
Opening context Problem or ‘angle’ stated Response
The opening context sets out some common ground on which everyone can
agree. The problem or ‘angle’ has two parts:
• some missing evidence or contradictory reasoning

• the consequences of not having an answer.
The response concerns your promise of a solution.

An example of how this works, presented in Table 12.1, is an 18-month case
study that I undertook for the children’s charity Barnardo’s (Thomas et al., 1998).
The case studied was of a school closing and moving all of its staff and students to
the local secondary and primary schools. This forms the foundation stone of the
entire thesis and it is important that it is stated clearly. I have borrowed the broad
structure from Booth et al.’s (2016) excellent book The Craft of Research.
Table 12.1 Locating an angle and promising a solution
Context … Common Inclusion of children with special needs in mainstream education is

ground happening increasingly. The move towards inclusion, backed by anti-
discrimination legislation, is occurring principally in response to concerns
over the loss of social and educational opportunities for those who are
segregated in special schools … and so on
Problem or angle Little is known about the experiences of students who transfer to
stated … What are mainstream schools. Concerns have been expressed about:
the consequences
• the ability of mainstream teachers adequately to meet the needs of
of not knowing? Who
young people with serious difficulties
cares?
• the readiness of mainstream students to help accommodate special
so what?
school students
• … and so on
Much comparative research has been undertaken [brief outline of what it
is prior to a full explanation in the literature review], but little work has been
undertaken to examine the quality and ‘texture’ of students’ experiences
in the new environment and how this changes over time – improving or
deteriorating – as the process of inclusion becomes ‘bedded in’ … and so on
Without information on these issues, the policy to include children with
special needs risks failure … and so on
Response … A diachronic case study focusing intensively on the experiences of the
Promise a solution students of a recently closed special school promises to add to the
knowledge about students’ social and educational adaptation over time and
will offer insight into the means by which such closures are effected
Beyond this, your literature review will help you to refine and revise the issue that
you wish to research. After doing this, you will begin your research using your case
study and the methods that you have chosen to conduct it. Your case study wrapper
encases all of the methods and instruments you choose to address the issue.
So, in the example of the special school that was closing, my fellow researchers
and I had started with some notions about the ways in which inclusion operated,
and these notions were further informed by our reading of the literature. We then
had ideas – a theory, if you like – about the challenges and opportunities that the
project might face as it proceeded. We made observations, held interviews with staff
and students, kept diaries, conducted documentary analysis and drew sociograms,
and these helped to ‘thicken’ or, alternatively, tone down or soften the various
strands of theory that we had developed.
The argument
The evidence that emerges as you proceed contributes all the time to the question
that you originally posed. Does the question still seem valid in the light of the new
information? Do you want to change course?
Remember that evidence is nothing on its own: it is data in support of a propo-
sition or a claim, so you will be modifying and adapting the claims that you are
making as you continue with your empirical work.
So, the process of argument involves recursion (retracing your footsteps), sum-
mary, synthesis, putting ideas (yours and those of others) next to each other to see
how they shape up together.
This is the way that your argument develops. It involves a revisiting and reas-
sessment of your original ideas and arguments to see which seem more or less
valid. Some ideas may be rejected or put into abeyance. Tell readers about this –
tell them about your process of acceptance, rejection and reformulation and the
conclusions you have come to. Booth et al. (2016) suggest that the process is
like that set out in Figure 12.3. I’ve added the bin to their diagram – I hope they
don’t mind.
Two examples of good analysis, argument

and writing
In the previous chapter as well as this one, I have noted that analysis and argu-
ment are difficult subjects on which to offer advice. It may help to provide a couple
of examples of first-rate analysis, and there are many examples of such quality.
Claim
Supports
Reasoning
Explains
Supports
Evidence Doesn’t support
Figure 12.3 The relationship between evidence and reason supporting a claim
(adapted from Booth et al., 2016)
I give several examples of such quality elsewhere (see Thomas and Myers, 2015).
As you explore the methodological literature about the case study, you will repeat-
edly find the same iconic studies used as examples. Again and again, you will find
case studies done as part of the sociological tradition, such as Howard Becker et al.’s
(1961/1980) Boys in White and William Foote Whyte’s (1955/1993) Street Corner
Society. You will probably also find reference to The Polish Peasant in Europe and
America by Thomas and Znaniecki (1927/1958).
I think it is a pity, though, to treat the case study as the sole prerogative of the
interpretative sociologist, so, in seeking examples of good analysis I have tried to
take a broader disciplinary approach to it. In the points that I make in this section,
I should like – for the sake of freshness – to use different examples of the genre.
They are interesting not only in terms of what they have to offer as case studies,
but also because their authors do not flag the fact that they have undertaken case
studies. Perhaps they don’t even care that they’re doing case study research. As I said
earlier, those who research in disciplines outside the social sciences seem delight-
fully unselfconscious about the methods they use.
I will concentrate on two sets of case studies – those done by a neurologist and a
biologist – for the insights that they have to offer.
Oliver Sacks – A neurologist who writes case

studies
Oliver Sacks (1996) was a neurologist by professional training, but he dis-
carded some of his neurological education to assume the clothes of a
storyteller and researcher in his book An Anthropologist on Mars. His book
relates how he talked with and observed people who exhibited unusual
kinds of behaviours – for example, in the syndromes that lead people to
display the characteristics of autism and Tourette’s.
In doing his case studies, Sacks offers a set of sparkling insights into
and understandings of the worlds of a number of people who behave
differently. Such insights have been largely curtained off from us by the
understandings offered by the traditional kinds of analysis found in medi-
cine and psychology.
Sacks gives reasons for eschewing many of the procedural and meth-
odological habits of his own discipline, neurology, in making his analysis.
Neurologists focus on helping people who are, for whatever reason, uncom-
fortable, unhappy, disaffected, unable or unwilling to ‘fit in’. Sacks’ insight is
that the methods which have been used to examine this discomfort or disaf-
fection, while they can be successful up to a point, fail to address the real
issues at stake – human issues. He says (Sacks, 1996: xvii–xviii):
The exploration of deeply altered selves and worlds is not one that
can be fully made in a consulting room or office. The French neurolo-
gist François Lhemitte is especially sensitive to this, and instead of just
observing his patients in the clinic, he makes a point of visiting them
at home, taking them to restaurants or theatres, or for rides in his car,
sharing their lives as much as possible. (It is similar, or was similar, with
physicians in general practice. Thus, when my father was reluctantly
considering retirement at ninety, we said, ‘At least drop the house
calls.’ But he answered, ‘No, I’ll keep the house calls – I’ll drop every-
thing else instead.’)
A case study is like keeping the house calls. Sacks spent days and weeks with
people with different kinds of behavioural issues – ones that caused them or
those around them concerns of one kind or another. He made every effort,
however, to put those behaviours in context, not seeing them as disembod-
ied, decontextualised medical or psychiatric conditions but, rather, as valid
ways to deal with the world, sometimes very successfully. For example, an
autistic woman who found great difficulty relating with other people but
who felt great tenderness for animals, combined the latter with her skill in
drawing and planning. She used this to help in the design of more humane
slaughterhouses that minimised, as far as possible, any potential distress
to the animals. Rather than defining a condition or trying to exemplify or
generalise from a condition with a supposedly typical example, Sacks used
the intimate knowledge he was gathering to illustrate the idiosyncrasy, the
particular story, of her case. He showed the individual at work and at home.
He showed her autonomy, strength, individuality and the integrity of her life.
This is not something that could be said of most medical or psychological
writing about autism.
Sacks makes no attempt to generalise from this or from any of his
vignettes, and his method is almost invisible since he lays out no mental
map at the beginning of his work. Rather, readers are left to make up their
own minds; they are left to piece together conclusions for themselves.
The conclusions – conclusions, note, rather than generalisations – are
about the ways in which difference can be respected rather than pigeon-
holed. The pigeonholing comes from the obvious categories of clinical and
administrative convenience and the emotions associated with those stereo-
typical categories leading too easily into hackneyed emotions of pity or
admiration so familiar to those with the conditions concerned. By using the
case study, Sacks expects to move away from the rut of categories and offers
us instead rich analyses of people’s lives.
The second set of case studies is from the work of well-known biologist Jared
Diamond (2005). While Sacks leaves us to make what we will of his case studies,
with each one being a little jewel in and for itself, Diamond develops from the case
studies a complex argument and thesis.
He looks at a range of communities and societies that have collapsed or are cur-
rently in the process of collapsing. Each one – the Montana of today, the Easter Island
of the seventeenth century, Pitcairn, the Mayans, the Vikings, modern Haiti – is very
different from the others, but they are connected by decline and collapse.
Diamond’s thesis is that the decline of a society happens because of a number
of potential factors, ranging from hostile neighbours to environmental change
and the society’s response to that change. His first point, then, was to establish a
connection between these communities. The second was to look for threads that
might form the basis of the connection.
Jared Diamond – Using a case study to tell a

story about societal collapse
Let’s look briefly at one of Diamond’s case studies. Easter Island has always
been a mysterious phenomenon to explorers because of its huge, ghostly
statues gazing out to sea. Yet, today, aside from these striking megaliths,
there is virtually nothing there. These sculptures were clearly the product of
a centuries-old, highly sophisticated civilisation, but why did it evaporate,
leaving nothing behind other than these remarkable statues?
It is fanciful to suggest that Diamond had a ‘method’ for answering these
questions since his inquiries consisted of not much more than reading widely
about the island, using existing research – botanical, cultural, historical (he
wasn’t fussy) – visiting it for an extended period and thinking. He looked at
Easter Island’s history and geography. He looked at what the people ate,
drawing from research using pollen and charcoal remains that gave clues
about now-extinct crops. Using archaeological evidence and oral history, he
looked at the likely structure of the society, people’s attitudes to death and the
afterlife, how the statues might have been sculpted, transported and erected.
The case study is about using many and varied sources of evidence. It is
about imagining and using your intuition in arguing a case.
Most importantly, he looked at how all of this might be connected, in the
way that the islanders sought and used resources – from the intensification
of their agriculture using windbreaks and pits for better growing to the divi-
sion of the island into 11, eventually competing, territories. All of this he
pieced together using intelligent questioning and intelligent answering –
often called the Socratic method. This method was formalised by Plato and
involved the posing of questions by Socrates to a pupil or other person.
Diamond does this, but he does it with himself and an imaginary interlocu-
tor. Here’s an example (Diamond, 2005: 99):
How did all those Easter Islanders, lacking cranes, succeed in carving,
transporting, and erecting those statues? Of course we don’t know for
sure, because no European ever saw it being done to write about it.
But we can make informed guesses from oral traditions of the island-
ers themselves … from statues in the quarries at successive stages of
completion, and from recent experimental tests of different transport
methods.
For me, two things are interesting about this passage. One is the self-
questioning that I have just mentioned. The other is the humility and
deference with which Diamond approaches knowledge. ‘Of course we
don’t know for sure’ should be the motto of anyone doing a case study. I’ll
go further: it should be the motto of any inquirer. We don’t know anything
for sure – we don’t even know for sure if the sun is going to rise tomorrow
morning.
The second thing is that Diamond goes on to point out that we make
informed guesses, which is a point I have made throughout about the
case study. Guessing – using our imagination on the basis of the available
evidence – is one of the main tools of the researcher, not just someone
working on a case study.
Diamond’s method of informed questioning, reasoning and case con-
struction works powerfully – and intelligent guessing is at the centre of all
this. It happens not just as a result of combining inert hard facts (which are,
of course, never as hard as we might like them to be) but also a chemis-
try, a bubbling cauldron in our brains. Thus, while he is making his case,
Diamond is imagining, linking here, there and everywhere, from prehistory
to Hollywood. He tells a story and develops an argument.
We are not all as imaginative or knowledgeable or energetic as Diamond,
but we can replicate his process to an extent. We can weave a narrative,
using the information that we garner not just as part of our formal inquiries
but also from our everyday lives – as professionals and ordinary people.
Some rules for writers

Writing is one of those crafts that just has to be learned on the job, but most of us –
even if we don’t think that we’re writers – are writing all the time in some shape or
form, so the good news is that we are always learning how to write.
Writing up a case study, however, requires a particular set of awarenesses and
skills and I think these can be split into two broad types:
• those concerning writing as an imaginative construction

• those to do with simply getting the thing completed.
Thus, it is about the creative and the pragmatic. The Guardian (2010) asked 20
leading novelists and writers to offer their own ‘rules for writers’ and these, inter-
estingly, split into the creative and the pragmatic. In fact, it was so interesting that
I have extracted some of the ones most pertinent to our own focus here and taken
the liberty of offering my own comments in Table 12.2.
A last thought about writing up a research report: you do have to consider register.
‘Register’ means the accepted form and style of writing for a particular audience
or forum – to write your dissertation using text-speak, for example, would obvi-
ously be unacceptable. A student of mine told me that she had tried to write like I
do in another of my books. Flattering as this was, I had to remind her that in the
book I was writing for a particular audience – that is, students who wanted to learn
something and who are sometimes intimidated by academic style and jargon – so
my register was as light as I could make it without its actually blowing away. The
register you need for a report is, sadly, more formal.
Table 12.2 Rules for writers, taken from the Guardian (2010)
Author My comment
Diana Athill
• Read it aloud to yourself Then you will hear as the reader will hear. Does it
sound right? Does it make sense? Will the reader
understand?
• Cut (perhaps that should be CUT): only Do cut out words that you think sound impressive.
by having no inessential words can every Avoid – like the plague – pretentious words such as
essential word be made to count ‘hegemony’ and ‘inscribe’
Roddy Doyle
• Chances are the words that come into As above. Ordinary words are usually best, though
your head will do fine, e.g. ‘horse,’ ‘ran,’ you will sometimes need technical ones in a piece of
‘said’ academic work. Think about the possible words you
could use and judge which will be right
Helen Dunmore
• Reread, rewrite, reread, rewrite Excellent advice. You might add … and read it out
loud to someone else
Anne Enright
• Keep putting words on the page One of the ways to make sure you get the thing
finished is just to keep writing. Every day. Set
yourself a target number of words per day. Even if
you write rubbish, you are keeping the momentum
going and you can cut out anything that doesn’t
work later
Jonathan Franzen
• The reader is a friend, not an adversary, Do think of your readers. Your readers are not some
not a spectator inert, bodyless phenomena. They are human beings
who breathe, burp, yawn and make sense of your
work (or don’t), depending on how you write it.
Imagine your readers
Author My comment
Michael Morpurgo
• Record moments, fleeting impressions, Particularly in a case study, you make sense of
overheard dialogue, your own sadnesses others’ intentions, feelings, words and emotions.
and bewilderments and joys Keeping a record of your own may help you to
interpret the emotions and ideas of others
Andrew Motion
• Honour the miraculousness of the Look for what is special in ordinary moments … the
ordinary gaze that is a second too long, the turn of the head,
the uncomfortable moment of silence … What can
you glean from anyone doing a case study?
• Think big and stay particular Could be the motto of anyone doing a case study.
While seeing the big picture – the model you are
developing – always root it in the concrete, the
particular
Will Self
• Always carry a notebook. And I mean Self-explanatory and very true
always. The short-term memory only
retains information for three minutes;
unless it is committed to paper you can
lose an idea for ever
Jeanette Winterson
• Turn up for work. Discipline allows She means ‘get down to work – and regularly’.
creative freedom. No discipline equals no Nothing gets written if you don’t write
freedom
• Never stop when you are stuck As above – it’s the momentum thing again
Remember you have an audience

In writing up a case study, probably more than in writing up any other kind of
research, it is important to remember the reader. Here, more than anywhere else,
the narrative, and your ‘voice’ need to be heard. So, make sure that you can ‘hear it
in your head’. In other words, read it out to yourself in your head (or, even better,
out loud to a friend). Does it sound right? Does it make sense? Is it interesting and
engaging? Will the reader understand?
For any kind of interpretative work, which case study is, you should sound like
yourself; you should put yourself in the picture. I say this with the proviso that
there is register to consider. ‘Register’ is the accepted form of writing for a particular
audience or forum, and there is a certain expectation about a degree of formality
in the write-up of academic work. However, this does not stop you from using your
own words and sounding like yourself in your writing. Indeed, it is to be welcomed
and admired. George Orwell, in his brilliant essay Politics and the English Language
(1946), makes the point that there is bad writing that finds its roots in imitation of
others (that is, imitating them and therefore not sounding like yourself) and he sug-
gests that this bad, imitative writing actually leads to sloppy thinking.
While your writing may well be published, for most write-ups your audience will be
other students, colleagues at work and, perhaps most importantly, your markers (in
the plural, since there will nearly always be at least two). Try to remember that your
markers are human, and want to read something that engages them, that they find
coherent, logical and interesting. It is good to have an idea in mind of what your mark-
ers are like so you can have them in your mind when you do your writing. Wallace
and Poulson (2005: 8) offer a useful guide, which I have liberally adapted in Table 12.3.
Table 12.3 A sense of audience: what your reader (and marker) looks like. (Liberally
adapted from Wallace and Poulson, 2005)
Age 25–70+
Appearance Anyone’s guess – from smart casual to mad-professor-look with odd socks. (If male
and wearing a tie, ask to see their credentials)
Lifestyle Busy, and likes reading – so appreciates writing with a logical structure and clear focus
Attitudes Fair and respectful
Sceptical
Open-minded
Favourite What you’re writing about. Highly knowledgeable about the area in general, but
subject possibly not au fait with the details of your specific topic
Likes Books, broadsheet newspapers and scholarly periodicals. So … likes carefully
reading … constructed, well-argued, balanced, detailed, reflective writing
Pet hates Waffle
Irrelevance
Avoidable errors … of punctuation, spelling, referencing – or anything that careful
proofreading would have picked up
Most likely How is this relevant?
to say Keep to the point
Be more critical
Read around the subject more

take this …
Think carefully about the shape of your written-up study. It needs to conform to
certain expectations. Those expectations concern structure (it needs a beginning,
middle and end), quality of argument and style. It also has to be good to read. Do
think about your readers – put yourself in their shoes. Remember:
• A certain structure is expected in any research report, of whatever kind. Before

embarking on the findings and analysis of the case study itself, you have to make
it clear why you thought this was a good topic to study, and why you considered
case study to be an appropriate design frame for your research. You also need to
show that you have contextualised your own work in that of others and show that
you have read around the area.
• In a case study, findings, analysis and discussion will usually be conflated.
• The heart of the case study is in the reasoning to support the argument being
made; readers should be able to see not only rigour in your collection of evidence,
but also imagination and originality in the conclusions you draw. For inspiration,
look at some iconic case studies, such as those discussed briefly in this chapter,
and in the further reading below.
• There are some basic rules of writing: continually proofread and edit your work;
think about the reader; be disciplined. Perhaps most important for case study:
‘think big and stay particular’.
… and this – a final thought

This is the end of the book and I’ll finish with a story since I’ve been stressing the
importance of narrative and meaning since the beginning.
I once saw a television programme in which a theoretical physicist was describ-
ing how difficult it was for him and his colleagues to build accurate models and
theories of the Universe, given the fragmentary data to which they had access. He
described it as trying to work out the rules of chess by watching a game with your
view restricted to the four squares at the bottom left-hand side of the board, seeing
pieces emerge and disappear at different angles and at different times, seemingly
without rhyme or reason. I guess he would have given his eye teeth for more infor-
mation about the physical universe – for a view of all 64 squares.
The irony is that, in social research, while we will never, ever work out all the
rules of social life, we can gain access to much more than the bottom left-hand
corner of the chess board. No one is restricting our view and, while we cannot see
everything we want to see, we can see a great deal.
More importantly, there is no need deliberately to restrict our own view.
Sometimes, of course, it might be sensible to structure the ways in which we organ-
ise our scrutiny of the worlds in which we are interested. We should not imagine,
however, that there is any requirement to split the world into manageable-sized
chunks since, by that splitting, we may force ourselves to sidestep scrutiny of some
key aspect of the whole.
I have cautioned in this book against the ‘mushy heap syndrome’ in case studies – a
syndrome in which a case is loosely focused without care for the elements of design
and conduct that I have stressed. At its worst, a case study can be a bit like an
unmade bed, with an untidy collection of quotes and observations without object,
direction, argument or conclusion. At its best, though, it can provide sparkling
insight and analysis that is unrivalled by any other kind of research. I hope that
in this book you will have found advice that helps you to conduct such research.
Enjoy your case study research, and good luck.
REFLECTIVE ACTIVITY
While a case study done as a piece of research has to conform to certain expecta-
tions, particularly if it is done as a piece of assessed academic work, it is, more
than the write-ups from research structured according to the expectations in
other design frames, a piece of writing. It should be interesting, stimulating,
thought-provoking. It is not just a write-up: it is writing, and the story, the narra-
tive, has to come through. This is why I have included advice from outstanding
writers at the end of this chapter. Try extracting from their advice, in three bullet
points, the key points for you.
Further Reading
Becker, H.S. (2007) Writing for Social Scientists (2nd edn). Chicago: University of
Chicago Press.
Good advice on writing, but it has to be read, like most of Becker’s work, a bit like a
novel. Very good on the idea that there is no single way of writing something well.
You can say things in a variety of ways. Lots of good examples. No bullet points.
Becker, H.S. (2014) What About Mozart? What About Murder? Reasoning from Cases.
More wisdom from Becker, this specifically on case study.
Booth, W.C., Colomb, G.C. and Williams, J.M. (2016) The Craft of Research (4th
edn). Chicago: University of Chicago Press.
I have mentioned this book in relation to its advice about establishing a rationale
(Chapter 7). It is equally good on the question of writing – not just how to con-
struct good sentences but also how to build good argument.
Cohen, M. (2015) Critical Thinking Skills for Dummies. London: Wiley, For Dummies.
Don’t be fooled by the ‘for dummies’ of the title. This is the most thoughtful and
readable book about critical thinking that I know of. It is as entertaining as it is
informative.
Cottrell, S. (2017) Critical Thinking Skills: Effective Analysis, Argument and Reflection.
London: Macmillan.
Good, basic advice on reflecting on your own work.
Diamond, J. (2005) Collapse: How Societies Choose to Fail or Survive. London: Penguin.
This book is a collection of case studies that can be read separately or all together.
About the collapse of societies due to environmental degradation, together the
studies offer an integrated thesis that Diamond unrolls as he proceeds through this
multiple case study.
Goodwyn, A. and Stables, A.W. (eds) (2004) Learning to Read Critically in Language
and Literacy. London: Sage.
While this book is nominally about criticality in language and literacy, the first two
chapters, by Louise Poulson and Mike Wallace, are specifically on criticality gener-
ally and what is meant by ‘thinking critically’ in higher education. They are as good
as anything else I have read on the topic, before or since.
Kamler, B. and Thomson, P. (2014) Helping Doctoral Students Write: Pedagogies for
Supervision. London: Routledge.
Useful advice throughout on writing in academic life, and particularly on the rela-
tionship of writing to research.
Mewburn, I., Firth, K. and Lehmann, S. (2019) How to Fix Your Academic Writing
Trouble. London: Open University Press.
Good practical advice on writing, offering a range of techniques and strategies to
achieve a ‘scholarly voice’. Begins each chapter with examples of where things can
go wrong.
Myers, K. (2011) ‘Contesting certification: mental deficiency, families and the

state’, Paedagogica Historica: International Journal of History of Education, 47 (6):
749–66.
Two examples are given in this chapter of good analysis and writing. This is another
one. The study is of a working-class family in the 1930s which successfully resisted
the pressure of the local authority to send the youngest child of the family to a
residential special school. It raises a range of issues, among them eugenics, the
employment of psychometrics, the use of authority and professional power, and
the changing role of parents. It brings these together to provide a rich analysis of
the ways in which forces interplay to develop, enact or resist policy.
Sacks, O. (1996) An Anthropologist on Mars. London: Picador.

In this classic collection of case studies, Sacks is focusing on difference and differences –
the differences that exist between people, which are sometimes magnified by a disability
or condition of some kind. Each is a vignette in itself, but, as a whole, the book offers a
multiple case study par excellence, since the whole is more than the sum of its parts. Each
study is, in a way, refracted through the others.
Thomas, G. (2012) ‘Changing our landscape of inquiry for a new science of educa-
tion’, Harvard Educational Review, 82 (1): 26–51.
Thomas, G. (2016) ‘After the gold rush: questioning the “gold standard” and
reappraising the status of experiment and randomized controlled trials in educa-
In both of these articles I talk about the power of case study as a form of social inquiry
vis-à-vis other forms of inquiry. They were written in the context that the use of
case study research has waned as policymakers have sought ‘evidence-based’ policy
informed by randomised controlled trials. They seem unable to understand the frailties
of such trials in social research and the potential contribution of case study evidence.
Thomas, G. (ed.) (2013) Case Study in Education. Volumes 1–4. London: Sage.
This collection gives many examples of iconic case studies and less well-known but
equally good ones. Some of these provide excellent examples of good writing.
In this book, Kevin Myers and I try to work out what is going on ‘inside’ the case
study. In Chapter 6 we offer ten examples of what we consider to be good analysis.
Wolcott, H. (1992) ‘Posturing in qualitative inquiry’, in M.D. LeCompte, W.L.

Milroy and J. Preissie (eds), The Handbook of Qualitative Research in Education.
New York: Academic Press, pp. 3–52.
Wolcott, H. (2009) Writing Up Qualitative Research (3rd edn). London: Sage.

Wolcott is good not just on writing but also positioning yourself as a researcher.
Other Reading (not mentioned elsewhere

in the book)
Bendix, R. (1963) ‘Concepts and generalization in comparative sociological studies’,
American Sociological Review, 28: 532–9.
Bradshaw, Y. and Wallace, M. (1991) ‘Informing generality and explaining unique-
ness: the place of case studies in comparative research’, International Journal of
Comparative Sociology, 32: 154–71.
Feagan, J., Orum, A.M. and Sjoberg, G. (eds) (1991) A Case for the Case Study. Chapel
Hill, NC: University of North Carolina Press.
Kurtz, M.J. (2000) ‘Understanding peasant revolution: from concept to theory and
case’, Theory and Society, 29 (1): 93–124.
Rosenblatt, P.C. (1981) ‘Ethnographic case study’, in M.B. Brewer and B. Collins
(eds), Scientific Inquiry and the Social Sciences: A Volume in Honor of Donald T.
Campbell. San Francisco: Jossey-Bass, pp. 194–225.
Snyder, R. (2001) ‘Scaling down: the subnational comparative method’, Studies in
Comparative International Development, 36 (1): 93–110.
Stoeker, R. (1991) ‘Evaluating and rethinking the case study’, Sociological Review, 39:
88–112.
Tobin, J. (2005) ‘Scaling up as catachresis’, International Journal of Research & Method
in Education, 28 (1): 23–32.
REFERENCES
Abbott, A. (1992) ‘What do cases do? Some notes on activity in sociological analy-
sis’, in C.C. Ragin and H.S. Becker (eds), What Is a Case? Exploring the Foundations
of Social Inquiry. Cambridge: Cambridge University Press.
Alderson, P. (2004) Ethics, Social Research and Consulting with Children and Young
People (2nd edn). London: Barnardo’s.
Althusser, L. (1979) For Marx. London: Verso.
Andreski, S. (1972) Social Sciences as Sorcery. London: André Deutsch.
Arnott, M. and Ozga, J. (2010) ‘Education and nationalism: the discourse of educa-
tion policy in Scotland’, Discourse: Studies in the Cultural Politics of Education, 31:
335–50. doi:10.1080/01596301003786951
Back, S. (2002) ‘The Aristotelian challenge to teacher education’, History of Intellectual
Culture, 2 (1) (available at: www.ucalgary.ca/hic/files/hic/back_forum.pdf).
Ball, S. (1981) Beachside Comprehensive: A Case-Study of Secondary Schooling. Cambridge:
Barker, R.G. (1968) Ecological Psychology. Palo Alto, CA: Stanford University Press.
Barrow, J.D. (1997) ‘The values of science’, Amnesty International Science Lecture,
Oxford University.
Barthes, R. (1974) S/Z (R. Miller, trans.). New York: Hill & Wang.
Bateson, G. (1972/1999) Steps to an Ecology of Mind. Chicago: University of Chicago Press.
Becker, H.S. (1992) ‘Cases, causes, conjunctures, stories, imagery’, in C.C. Ragin
and H.S. Becker (eds), What Is a Case? Exploring the Foundations of Social Inquiry.
Becker, H.S. (1998) Tricks of the Trade. Chicago: University of Chicago Press.
Becker, H.S. (2014) What About Mozart? What About Murder? Reasoning from Cases.
Becker, H.S., Geer, B., Hughes, E.C. and Strauss, A.L. (1961/1980) Boys in White.
Edison, NJ: Transaction Publishers.
Belmont Report (1979) Ethical Principles and Guidelines for the Protection of
Human Subjects of Research: The National Commission for the Protection of
Human Subjects of Biomedical and Behavioral Research. Washington, DC: US
Department of Health & Human Services (available at: www.hhs.gov/ohrp/
humansubjects/guidance/belmont.html).
Berger, P.L. and Luckmann, T. (1979) The Social Construction of Reality: A Treatise in
the Sociology of Knowledge. London: Penguin.
REFERENCES 285
Biletzki, A. (2009) ‘Ludwig Wittgenstein’, in E.N. Zalta (ed.), Stanford Encyclopedia

of Philosophy (summer 2009 edition). Stanford, CA: Metaphysics Research Lab,
Center for the Study of Language and Information, Stanford University (avail-
able at: http://plato.stanford.edu/archives/sum2009/entries/wittgenstein/).
Blumer, H. (1992) Symbolic Interactionism: Perspective and Method. Berkeley, CA:
University of California Press.
Booth, W.C., Colomb, G.C. and Williams, J.M. (2016) The Craft of Research
(4th edn). Chicago: University of Chicago Press.
Bronfenbrenner, U. (1979) The Ecology of Human Development. Cambridge, MA:
Harvard University Press.
Bronfenbrenner, U. and Morris, P.A. (1998) ‘The ecology of developmental pro-
cesses’, in R.M. Lerner (ed.), Handbook of Child Psychology, Vol. 1: Theoretical
Models of Human Development. New York: Wiley, pp. 993–1028.
Bruner, J. (1991) ‘The narrative construction of reality’, Critical Inquiry, 18 (1): 1–21.
Bruner, J. (1997) The Culture of Education. Cambridge, MA: Harvard University Press.
Bryant, G. and Monk, P. (2001) ‘Summary of the final report of the investiga-
tion into the North Leicestershire cluster of variant Creutzfeldt-Jakob disease’,
Leicestershire Health Authority (available at: www.cbsnews.com/htdocs/pdf/
vcjd.pdf).
Burgess, R.G. (1982) Field Research: A Sourcebook and Field Manual. London: Routledge.
Byrne, R.M.J. (2005) The Rational Imagination: How People Create Alternatives to
Reality. Cambridge, MA: MIT Press.
Caldwell, J.C. (1986) ‘Routes to low mortality in poor countries’, Population and
Development Review, 12 (2): 171–220.
Checkland, P. (1981) Systems Thinking, Systems Practice. Chichester: Wiley.
Collingwood, R.G. (1946/1994) The Idea of History. Oxford: Oxford University Press.
Cremin, H., Mason, C. and Busher, H. (2011) ‘Problematising pupil voice using
visual methods: findings from a study of engaged and disaffected pupils in an
urban secondary school’, British Educational Research Journal, 37 (4): 585–603.
Cremin, H., Thomas, G. and Vincent, K. (2005) ‘Working with teaching assistants:
three models evaluated’, Research Papers in Education, 20 (4): 413–32.
Denzin, N. (2017) Sociological Methods: A Sourcebook. London: Routledge.
Descartes, R. (1647/1996) Meditations on First Philosophy. Cambridge: Cambridge
University Press.
Dewey, J. (1920/2004) How We Think. Whitefish, MT: Kessinger Publishing.
Diamond, J. (2005) Collapse: How Societies Choose to Fail or Survive. London: Penguin.
Dick, B. (2002) Soft Systems Methodology, Session 13 of Areol – action research and eval-
uation on line. Southern Cross University and Southern Cross Institute of Action
Research (available at: www.aral.com.au/areol/areol-session13. html).
Dickens, W.T. and Flynn, J.R. (2001) ‘Heritability estimates versus large environ-
mental effects: the IQ paradox resolved’, Psychological Review, 108: 346–69.
Eisenhart, M. (2009) ‘Generalization from qualitative inquiry’, in K. Ercikan and
W.M. Roth (eds), Generalizing from Educational Research. Abingdon: Routledge.
Eisner, E. (1991) The Enlightened Eye: Qualitative Inquiry and the Enhancement of
Educational Practice. New York: Macmillan.
Ellis, L. and Hodges, L. (2014) ‘Life and change with Usher: the experiences of diag-
nosis for people with Usher syndrome’, University of Birmingham (available at:
www.birmingham.ac.uk/Documents/college-social-sciences/education/projects/
final-report-on-life-and-change-with-usher.pdf).
Evans, H. (2000) Essential English for Journalists, Editors and Writers. London:
Pimlico.
Fairclough, N. (1995) Critical Discourse Analysis. Harlow: Longman.
Fairclough, N. (2003) Analysing Discourse: Textual Analysis for Social Research.
London: Routledge.
Ferguson, N. (1999) Virtual History: Alternatives and Counterfactuals. New York: Basic
Books.
Feyerabend, P. (1993) Against Method (3rd edn). London: Verso/New Left Books.
Field, A. (2013) Discovering Statistics Using SPSS (4th edn). London: Sage.
Fish, S. (1989) Doing What Comes Naturally. Oxford: Clarendon Press.
Flynn, J.R. (1987) ‘Massive IQ gains in 14 nations: what IQ tests really measure’,
Psychological Bulletin, 101: 171–91.
Flynn, J.R. (1998) ‘IQ gains over time: toward finding the causes’, in U. Neisser (ed.),
The Rising Curve: Long-Term Gains in IQ and Related Measures. Washington, DC:
American Psychological Association.
Flynn, J.R. (1999) ‘Searching for justice: the discovery of IQ gains over time’,
American Psychologist, 54 (1): 5–20.
Flynn, J.R. (2003) ‘Movies about intelligence: the limitations of g’, Current Directions
in Psychological Science, 12 (3): 95–9.
Flyvbjerg, B. (2001) Making Social Science Matter: Why Social Inquiry Fails and How It
Can Succeed Again. Cambridge: Cambridge University Press.
Flyvbjerg, B. (2006) ‘Five misunderstandings about case-study research’, Qualitative
Inquiry, 12 (2): 219–45.
Foucault, M. (1981) ‘Questions of method: an interview with Michel Foucault’,
Ideology and Consciousness, 8 (Spring): 3–14.
Geertz, C. (1975) The Interpretation of Cultures. London: Hutchinson.
Geoghegan, T. (2007) ‘History lessons: for Americans, Herodotus has better ones to
offer than Thucydides’, The American Prospect, 12 March (available online only
at: https://prospect.org/article/history-lessons-1).
REFERENCES 287
Georgiades, N.J. and Phillimore, L. (1975) ‘The myth of the hero-innovator and
alternative strategies for organizational change’, in C.C. Kiernan and F.P.
Woodford (eds), Behaviour Modification with the Severely Retarded. Amsterdam:
Associated Scientific Publishers.
Glaser, B.G. and Strauss, A.L. (1967) The Discovery of Grounded Theory: Strategies for
Qualitative Research. New York: Aldine.
Goffman, E. (1956) The Presentation of Self in Everyday Life. Edinburgh: University of
Edinburgh Press.
Grinyer, A. (2002) ‘The anonymity of research participants: assumptions, ethics and
practicalities’, Social Research Update, 36: 1–6.
Guardian (2010) ‘10 rules for writers’, Guardian, Guardian Review, 20 February: 2–5.
Haig, B.D. (1995) ‘Grounded theory as scientific method’, in A. Neiman (ed.),
Philosophy of Education Yearbook, 1995: Current Issues. Champaign, IL: Philo-
sophy of Education Society, University of Illinois Press at Urbana-Champaign,
pp. 281–90.
Haldane, J.B.S. (1928) Possible Worlds and Other Essays. London: Chatto & Windus.
Hammersley, M. (2005) ‘Assessing quality in qualitative research’, Paper presented
to ESRC TLRP seminar series: Quality in Educational Research, University of
Birmingham, 7 July.
Hammersley, M. (2007) ‘The issue of quality in qualitative research’, International
Journal of Research and Method in Education, 30 (3): 287–306.
Hammersley, M. and Gomm, R. (2000) ‘Introduction’, in R. Gomm, M. Hammersley
and P. Foster (eds), Case Study Method. London: Sage.
Hearnshaw, L.S. (1979) Cyril Burt: Psychologist. London: Hodder & Stoughton.
Holton, G. (1995) ‘The controversy over the end of science’, Scientific American, 273
(4): 168.
Houser, N., Kloesel, C. and the Peirce Edition Project (eds) (1992) The Essential
Peirce, 2 vols. Bloomington, IN: Indiana University Press.
James, W. (1890/1981) The Principles of Psychology. Cambridge, MA: Harvard
University Press.
Jenkins, R. (1992) Pierre Bourdieu. London: Routledge.
Jiménez, R.T. and Gersten, R. (1999) ‘Lessons and dilemmas derived from the liter-
acy instruction of two Latina/o teachers’, American Educational Research Journal,
36 (2): 265–302.
Johnson, S. (1759/1963) ‘The Idler, no. 58, Universal Chronicle’, in W.J. Bate, J.M.
Bullitt and L.F. Powell (eds), Works of Samuel Johnson, Vol. 2. New Haven, CT: Yale
University Press.
Junghans, C., Feder, G., Hemingway, H., Timmis, A. and Jones, M. (2005) ‘Recruiting
patients to medical research: double blind randomised trial of “opt-in” versus
“opt-out” strategies’, British Medical Journal, 331: 940 (available at: www.bmj.
com/cgi/reprint/331/7522/940.pdf).
Kahneman, D. and Tversky, A. (1973) ‘On the psychology of prediction’, Psychological
Review, 80 (4): 237–51.
Kamprad, I. with Toreleull, B. (2000) Leading by Design: The IKEA Story. London:
HarperCollins.
Kellett, M. (2005) How to Develop Children as Researchers. London: Paul Chapman
Publishing.
Kelly, D.H. (1993) ‘Case: grammar and terminology’, The Classical World, 87 (1):
35–9.
Kenny, A. (ed.) (1993) The Wittgenstein Reader. Oxford: Blackwell.
Kossyvaki, L. and Papoudi, D. (2016) ‘A review of play interventions for chil-
dren with autism at school’, International Journal of Disability, Development and
Education, 63 (1): 45–63. doi: 10.1080/1034912X.2015.1111303
Kounin, J. (1970) Discipline and Group Management in Classrooms. New York: Holt,
Rinehart & Winston.
Kuhn, T.S. (1970) The Structure of Scientific Revolutions. Chicago: University of
Chicago Press.
Lebow, R.N. (2007) ‘Counterfactual thought experiments: a necessary teaching
tool’, The History Teacher, 40 (2) (available at: www.historycooperative.org/
journals/ht/40.2/lebow.html).
Levitt, S.D. and List, J.A. (2009) ‘Was there really a Hawthorne effect at the Hawthorne
plant? An analysis of the original illumination experiments’, NBER Working
Paper No. 15016. Cambridge, MA: National Bureau of Economic Research.
Lewin, K. (1951) Field Theory in Social Science: Selected Theoretical Papers. New York:
Harper.
Lewin, K. (1997) Resolving Social Conflicts: Field Theory in Social Science. Washington,
DC: American Psychological Association.
Lewis, A. and Lindsay, G. (2000) Researching Children’s Perspectives. Buckingham:
Open University Press.
Lewis, A. and Porter, J. (2004) ‘Interviewing children and young people with learn-
ing disabilities: guidelines for researchers and multi-professional practice’, British
Journal of Learning Disabilities, 32 (4): 191–7.
Lincoln, Y.S. and Guba, E.G. (1985) Naturalistic Inquiry. Beverly Hills, CA: Sage.
MacIntyre, A. (1985) After Virtue: A Study in Moral Theory. London: Duckworth.
Mandel, D.R., Hilton, D.J. and Catellani, P. (2005) The Psychology of Counterfactual
Thinking. Abingdon: Routledge.
Milgram, S. (1963) ‘Behavioral study of obedience’, Journal of Abnormal and Social
Psychology, 67 (4): 371–8.
REFERENCES 289
Milgram, S. (1981) ‘Behavioral study of obedience’, Social and Behavioral Science,

9: 18 (available at: www.garfield.library.upenn.edu/classics1981/A1981LC3330
0001.pdf).
Miller, S. and Fredericks, M. (1999) ‘How does grounded theory explain?’, Qualitative
Inquiry, 9: 538–51.
Mitchell, J.C. (2006) ‘Case and situation analysis’, in T.M.S. Evens and D. Handelman
(eds), The Manchester School: Practice and Ethnographic Praxis in Anthropology.
Oxford: Berghahn Books.
Mogey, J.M. (1955) ‘The contribution of Frédéric Le Play to family research’, Marriage
and Family Living, 17 (4): 310–15.
Moreno, M.A., Goniu, N., Moreno, P.S. and Diekema, D. (2013) ‘Ethics of social
media research: common concerns and practical considerations’, Cyberpsychology,
Behavior and Social Networking, 16 (9): 708–13 (available at: www.ncbi.nlm.nih.
gov/pmc/articles/PMC3942703/).
Mouzelis, N. (1995) Sociological Theory: What Went Wrong? London: Routledge.
Murdoch, I. (2002) Under the Net. London: Vintage.
Norton, J.D. (2004) ‘Einstein’s investigations of Galilean covariant electrodynamics
prior to 1905’, Archive for History of Exact Sciences, 59: 45–105 (available at: www.
pitt.edu/~jdnorton/papers/Einstein1905.pdf).
Oakeshott, M. (1967) Rationalism in Politics and Other Essays. London: Methuen.
Oakley, A. (1999) ‘Paradigm wars: some thoughts on a personal and public trajec-
tory’, International Journal of Social Research Methodology, 2 (3): 247–54.
Ohno, T. (2006) ‘Ask “why” five times about every matter’, Toyota Motor Corporation
(available at: www.toyota-global.com/company/toyota_traditions/quality/mar_
apr_2006.html).
Omaar, R. (2009) Race and Intelligence: Science’s Last Taboo. London: Channel 4 Television.
Ortlipp, M. (2008) ‘Keeping and using reflective journals in the qualitative research
process’, The Qualitative Report, 13 (4): 695–705 (available at: https://nsuworks.
nova.edu/tqr/vol13/iss4/8).
Orwell, G. (1946) Politics and the English Language. London: Penguin.
Parlett, M. and Hamilton, D. (1987) ‘Evaluation as illumination’, in R. Murphy and
H. Torrance (eds), Evaluating Education: Issues and Methods. Milton Keynes: Open
University Press.
Patrick, J. (1973) A Glasgow Gang Observed. London: Methuen.
Polanyi, M. (1958) Personal Knowledge: Towards a Post-critical Philosophy. London:
Routledge & Kegan Paul.
Popper, K.R. (1968) The Logic of Scientific Discovery. London: Hutchinson.
Popper, K.R. (1977) ‘On hypotheses’, in P.N. Johnson-Laird and P.C. Wason (eds),
Thinking: Readings in Cognitive Science. Cambridge: Cambridge University Press.
Prosser, J. (1992) ‘Personal reflections on the use of photography in an ethnographic

case study’, British Educational Research Journal, 18 (4): 397–411.
Prosser, J. (1998) Image-Based Research. London: Routledge.
Prosser, J. and Loxley, A. (2008) ‘Introducing visual methods’, ESRC National Centre for
Research Methods Review Paper NCRM/010. Southampton: ESRC NCRM (available
at: http://eprints.ncrm.ac.uk/420/1/MethodsReviewPaperNCRM-010.pdf).
Putnam, R. (2000) Bowling Alone: The Collapse and Revival of American Community.
New York: Simon & Schuster.
Ragin, C.C. (1992) ‘Introduction: cases of “What is a case?”’, in C.C. Ragin and H.S.
Becker (eds), What Is a Case? Exploring the Foundations of Social Inquiry. Cambridge:
Ragin, C.C. (2007) ‘Comparative methods’, in W. Outhwaite and S.P. Turner (eds), The
Sage Handbook of Social Science Methodology. Thousand Oaks, CA: Sage, pp. 67–81.
Roethlisberger, F. and Dickson, W. (1939) Management and the Worker. Cambridge,
MA: Harvard University Press.
Rutter, M. and the English and Romanian Adoptees (ERA) Study Team (1998)
‘Developmental catch-up, and deficit, following adoption after severe global
early privation’, Journal of Child Psychology and Psychiatry, 39: 465–76.
Sacks, O. (1996) An Anthropologist on Mars. London: Picador.
Schatzman, L. (1991) ‘Dimensional analysis: notes on an alternative approach to
the grounding of theory in qualitative research’, in D.R. Maines (ed.), Social
Organization and Social Process: Essays in Honor of Anselm Strauss. New York:
Aldine, pp. 303–14.
Schwandt, T.A. (1996) ‘Farewell to criteriology’, Qualitative Inquiry, 2: 58–72.
Schwandt, T.A. (2001) Dictionary of Qualitative Inquiry (2nd edn). Thousand Oaks,
CA: Sage.
Schwartz, D. (1992) Waucoma Twilight: Generations of the Farm. Washington, DC:
Smithsonian Institution Press.
Sen, A. (1999) Development as Freedom. New York: Alfred A. Knopf.
Sennett, R. (2009) The Craftsman. London: Penguin.
Silverman, D. (2010) Doing Qualitative Research (3rd edn). London: Sage.
Simon, H. (1983) Reason in Human Affairs. Oxford: Basil Blackwell.
Simons, H. (1989) ‘Ethics of case study in educational research and evaluation’, in
R. Burgess (ed.), The Ethics of Educational Research. Lewes: Falmer, pp. 114–40.
Smail, D. (1993) The Origins of Unhappiness. London: HarperCollins.
Smith, J.K. and Deemer, D.K. (2000) ‘The problem of criteria in the age of relativ-
ism’, in N.K. Denzin and Y.S. Lincoln (eds), Handbook of Qualitative Research
(2nd edn). Thousand Oaks, CA: Sage.
REFERENCES 291
Solzhenitsyn, A. (1963/2000) One Day in the Life of Ivan Denisovich. London: Penguin.
Stake, R.E. (1995) The Art of Case Study Research. Thousand Oaks, CA: Sage.
Stake, R.E. (2005) ‘Qualitative case studies’, in N.K. Denzin and Y.S. Lincoln (eds),
The Sage Handbook of Qualitative Research (3rd edn). Thousand Oaks, CA: Sage.
Stenhouse, L. (1979) ‘Can research improve teaching? Summary of a lecture given
by Lawrence Stenhouse at the Scottish National Conference on Curriculum
and Evaluation for P.E. Teachers’, January (available at: www.uea.ac.uk/docu
ments/4059364/4994243/Stenhouse-1979-Can+Research+improve+teaching.
pdf/54a14fb3-6932-40f4-8570-942944ef0d63).
Storr, A. (1997) Feet of Clay: A Study of Gurus. London: HarperCollins.
Strauss, A. and Corbin, J. (1990) Basics of Qualitative Research: Grounded Theory
Procedures and Techniques. Newbury Park, CA: Sage.
Tavor Bannet, E. (1997) ‘Analogy as translation: Wittgenstein, Derrida, and the law
of language’, New Literary History, 28 (4): 655–72.
Taylor, S.J. and Bogdan, R. (1984) Introduction to Qualitative Research Methods: The
Search for Meanings. New York: Wiley.
Thomas, G. (1992) Effective Classroom Teamwork: Support or Intrusion? London:
Routledge.
Thomas, G. (2007) Education and Theory: Strangers in Paradigms. Maidenhead: Open
University Press.
Thomas, G. (2009) ‘“What works” as a sublinguistic grunt, with lessons from cata-
chresis, asymptote, football and pharma’, Research Intelligence, 106: 20–2.
Thomas, G. (2010) ‘Doing case study: abduction not induction; phronesis not the-
ory’, Qualitative Inquiry, 16 (7): 572–82.
Thomas, G. (2011a) ‘A typology for the case study in social science following a
review of definition, discourse and structure’, Qualitative Inquiry, 17 (6): 511–21.
Thomas, G. (2011b) ‘The case: generalisation, theory and phronesis in case study’,
Oxford Review of Education, 37 (1): 21–35.
Thomas, G. (2012) ‘Changing our landscape of inquiry for a new science of educa-
Thomas, G. (2016) ‘After the gold rush: questioning the “gold standard” and reap-
praising the status of experiment and randomized controlled trials in education’,
Harvard Educational Review, 86 (3): 390–411.
Thomas, G. (2017) How to Do Your Research Project (3rd edn). London: Sage.
Thomas, G. (2020) ‘Experiment’s persistent failure in education inquiry, and why it
keeps failing’, British Educational Research Journal, doi:10.1002/berj.3660
Thomas, G. and James, D. (2006) ‘Reinventing grounded theory: some questions
about theory, ground and discovery’, British Educational Research Journal, 32 (6):
767–95.
Thomas, G. and Loxley, A. (2007) Deconstructing Special Education and Constructing

Inclusion (2nd edn). Maidenhead: Open University Press.
Thomas, G., Walker, D. and Webb, J. (1998) The Making of the Inclusive School.
London: Routledge.
Thomas, W.I. and Znaniecki, F. (1927/1958) The Polish Peasant in Europe and America
(2nd edn). New York: Dover.
Vaughan, D. (1992) ‘Theory elaboration: the heuristics of case analysis’, in
C.C. Ragin and H.S. Becker (eds), What Is a Case? Exploring the Foundations of
Social Inquiry. Cambridge: Cambridge University Press.
Vincett, K., Cremin, H. and Thomas, G. (2005) Teachers and Assistants Working
Together. Maidenhead: Open University Press.
von Bertalanffy, L. (1950) ‘The theory of open systems in physics and biology’,
Science, 111: 23–9.
Wacquant, L.D. (1989) ‘Towards a reflexive sociology: a workshop with Pierre
Bourdieu’, Sociological Theory, 7: 26–63.
Wallace, M. and Poulson, L. (2005) ‘Critical reading for self-critical writing’, in
A. Goodwyn and A. Stables (eds), Learning to Read Critically in Language and
Literacy. London: Sage, pp. 3–38.
Walton, J. (1992) ‘Making the theoretical case’, in C.C. Ragin and H.S. Becker (eds),
What Is a Case? Exploring the Foundations of Social Inquiry. Cambridge: Cambridge
University Press.
Wells, M. (2005) ‘Paxman answers the questions’, Guardian, 31 January (available at:
http://guardian.co.uk/media/2005/jan/31/mondaymediasection.politicsand
themedia).
White, H.C. (1992) ‘Cases are for identity, for explanation, or for control’, in
C.C. Ragin and H.S. Becker (eds), What Is a Case? Exploring the Foundations of
Social Inquiry. Cambridge: Cambridge University Press.
Whittemore, R., Chase, S.K. and Mandle, C.L. (2001) ‘Validity in qualitative
research’, Qualitative Health Research, 11: 522–37.
Whyte, W.F. (1955/1993) Street Corner Society: Social Structure of an Italian Slum.
Whyte, W.F. (1985) Learning from the Field: A Guide from Experience. Beverly Hills,
CA: Sage.
Wieviorka, M. (1992) ‘Case studies: history or sociology?’, in C.C. Ragin and
H.S. Becker (eds), What Is a Case? Exploring the Foundations of Social Inquiry.
Wiles, R., Crow, G., Heath, S. and Charles, V. (2008) ‘The management of confiden-
tiality and anonymity in social research’, International Journal of Social Research
Methodology, 11 (5): 417–28.
REFERENCES 293
Willis, P.E. (1981) Learning to Labor: How Working Class Kids Get Working Class Jobs.
New York: Columbia University Press.
Wood, R. (2019) Inclusive Education for Autistic Children: Helping Children and Young
People to Learn and Flourish in the Classroom. London: Jessica Kingsley Publishers.
Yin, R.K. (2018) Case Study Research: Design and Methods (6th edn). Thousand Oaks,
CA: Sage.
INDEX
Abbott, A., 76, 79, 253 bias, 209, 266
ABC of behaviour, 182 biography, 73
abduction, 74–9, 84, 129, 283 Blumer, H., 55
accounts, 206–7, 215 Booth, W., 143, 269, 270, 271, 280
accumulation of evidence, 76 Bourdieu, P., 75, 161, 243
action research, 45–6, 53, 58, 125, 194, Bradshaw, Y., 283
238, 255, 265 brainstorming, 33, 37, 116, 135, 238
adopted Romanian children, 111–3 British Educational Research Association
Alderson P., 93, 103 (BERA), 100, 101
Althusser, L., 75 Bronfenbrenner, U., 59, 61, 62, 194
analogy, 189, 191, 248, 253–4, 258 Bruner, J., 31, 131, 249–53, 260
analysis, 203–35 Bryman, A., 222
analytical category, 12, 147 Burgess R., 65, 172, 217
analytical frame, 12–17, 21–4, 33–4, 39, Burt, C., 70–1
41, 44, 223
Andreski, S., 257 Caldwell, J.C., 118–25
anecdotal evidence, 4 Campbell, D., 283
anomalies, 74, 249 capability, Sen’s notion of, 121
anonymisation, 89, 96 cartoons, 217
anonymity, 87–8, 91, 96, 104–5, 238 causal narrativity, 253
anorexia nervosa, 94, 182 causation, 9, 11, 46
Anthropologist on Mars, 272, 282 cause, and Becker’s ‘sequence of steps’, 250
anthropology, 159, 172 census data, 219
approaches to research, 44 Checkland, P., 60–1, 65, 238–40
archival records, 178 chefs’ recipes, 188, 195
Aristotle, 51, 52, 78, 79 classrooms, 133, 192, 194, 195
artefacts, 211 closed questions, 212
ATLAS.ti, 230 coding, 225–7, 232–3, 259
attitude to knowledge, 266 axial, 232
attitudes, 212–4, 239 open, 232
audiences, of research, 254–5, 265, 276–8 selective, 232
audio, 96, 207, 211, 225, 230 Coffey, A., 104
audit society, 70 Cohen, M.F., 25, 222, 281
authority, problems arguing from, 268 collecting data, 1, 171, 203–222
collective case study. See multiple study
Bacon, F., 51 Collingwood, R., 80, 178
Ball, S., 47, 48 comparative study. See multiple study
Barker, R., 29, 58, 60, 64 comparative research, 45, 46, 219, 269, 283
Barthes, R., 252 concealment, 88, 93
Bassey, M., 125 conceptual model, in soft systems, 239
Bateson, G., 58, 64, 65 confidentiality, 87–91, 96, 105, 205, 238
Beachside Comprehensive, 47, 48 consent, 86, 89, 90–8, 101–2, 112
Becker, H., 5, 9, 26, 44, 48, 65, 84, 142, 244, constant comparative method, 194, 224–5,
250–3, 256, 258, 260, 271, 280 228, 230–3, 236, 259, 261
Belmont Report, 86, 88 contextual antecedents, 136
Bendix, R., 283 control group, 168
BERA, 100, 103 correlation, 118
Berlin Wall, 32–8, 179 craft knowledge, 51, 78, 188
INDEX 295
creativity in designing research, 253, 257 ethics, 85–104, 130, 205

criteriology, 69, 81, 84 and concealment or deception, 93
criterion- v norm-referenced, 218 and confidentiality, 88–90
criticality, in thinking and reasoning, 131, and consent, 90–8
209, 244, 265–8 and risk, 92, 94
cross-case analysis. See multiple study and social media, 98
Cuban Missile Crisis, 4, 94 and vulnerable groups, 86–7
cultural expectations, 116 protocols, 101
cumulation. See accumulation of evidence ethnographic approach, 159, 173, 206, 217
curiosity, 128, 250, 257 ethnography, 18, 84, 104, 159, 160, 173
European Social Survey, 219
data evaluation, 22, 45, 46, 81, 104, 109, 120–31
analysis, 81, 223–60 evaluative study, 129–31
gathering, 48, 203–222 Evans, H., 7, 245
secondary, 222 evidence
Data Protection Act, 90 collecting, 203–20
data security and stewardship, 90, 103 analysing, 223–58
DBS (Disclosure and Barring Scheme) examples of case studies
checks, 95–6 adopted Romanian children (as local
debriefing for participants, 88, 96 knowledge case study), 111–3
deception, 88 Berlin Wall (designing a case study),
deduction and deducing, 77, 121 32–8, 179
Denzin, N., 26, 72 dynamic hip screw (as illustrative–
Department of Health, 101 demonstrative case study), 158
dependent variable, 170 dyslexia (as exploratory case study), 136
Descartes, R., 267 Easter Island (as multiple or comparative
design of research, 27–48 case study), 273–4
Dewey, J., 268 employment in schools (as explanatory
diachronic study. See types of case study case study), 132
diaries, 47, 178, 204, 207–10, 215, 221–2, hospital infection with MRSA
225, 270 (as theory-testing case study), 150
disability, 86, 196, 282 Hostos–Lincoln Academy (from idea to
discourse analysis, 234–5 question), 42–4, 111
discussion, 260, 262, 264 hung parliament (as analytical frame,
dissertations, 255, 256, 262 or object), 17
proportions of words in, 262 Hurricane Katrina (as key case study), 111,
writing up, 265 115–7, 122, 255
document interrogation, 211, 221 IKEA as a business (defining boundary of
documents, 178, 203–4, 211–2, 230, 235 case), 4, 19–21, 25
dramaturgy, 54–6 IQ and ethnicity (from idea to question),
39, 40, 41
Kerala, longevity in (as outlier case study),
ecological model, 58
111, 112, 118–25, 219
ecological psychology, 29, 56–9, 61, 64
Korean War (and analytical frames), 15
ecological systems theory, 61–3
media styles, contrasting US and UK
ecology of the classroom, 29, 56–7, 61,
(drawing theory from), 245–7
64–5, 167
park design (as sequential case study), 197
Education and Training Statistics, 219
quiet child (and distinction of subject and
Einstein, A., 7, 9, 44, 54, 69, 70, 256
object), 16–19, 23, 25
Eisenhardt, K., 172 school governor (as interpretative case
Eisner, E., 248, 260 study), 163
empirical findings, 247 teaching assistants (as experimental case
engagement, in class 89, 170 study), 132, 134, 168
epidemiological inquiry, 121 terrorism (as theory-building case study),
ethical clearance, 88, 93, 99–101, 184 146–8, 175
experiment, 165–70 holism, 49–64

experiments, 8, 26, 85, 165–7, 183 Holton, G., 7
exploratory study, 135–41 Hostos–Lincoln Academy, 42–4, 111
expressions, facial, 207 Husserl, E., 55
Facebook, 98, 220 identifying your case, 109–24

facilitator in interviews, 210–1 IKEA, 4, 19, 20, 21, 25
Fairclough N., 235, 260 illumination, 21
Feagan, J., 283 illustrative case studies, 152–9
feedback to participants, 87, 96, 157, 238 image–based methods, 204, 217, 221
Ferguson, N., 252 imagery, 218, 258
Feyerabend, P., 252 imagination, 250, 252, 256–8, 260,
field theory (Lewin), 59 266, 275, 279
fieldwork, 141, 160, 203, 217, 263 implied consent, 91, 92, 97, 98
films, 217 incremental chunking, 250
Fish, S., 77, 79 in–depth analysis, 42, 114, 196, 247
Flynn effect, 40 induction, 74–9, 84, 283
Flynn, J., 40, 41, 43 inductive reasoning, 75
Flyvbjerg, B., 6, 51, 65, 154 information sheet, 95, 98, 99
focus groups, 186, 204, 210, 221 informed consent, 87, 91, 98
focus materials, 186, 211 See also, consent
forms of case study, 21–4 insight, 79, 129, 244, 248, 272
See also, types of case study institutional review, 100
Foucault, M., 5, 72 Institutional Review Board. See, IRB
Freud, S., 244 instrumental study, 23, 128
intellectual love, Einstein’s notion of, 256
GDPR. See General Data Protection Regulation intelligent noticing, 249
Geertz, C., 84, 163, 172, 233, 255 interaction, 55, 58, 235
gender, 53, 73, 235 interdisciplinarity, 60
General Data Protection international comparison, 121
Regulation (GDPR), 90 interpretation, 55, 71–9, 144, 163–5, 217,
generalisation and generalisability, 5, 7, 9, 235, 244, 252
11, 24, 51, 52, 54, 65, 73–8, 80, 82, 84, interpretative inquiry, 69, 71, 73, 77, 80, 159,
125, 129, 244, 250, 251, 283 160, 161–65, 171–3, 186, 205, 206, 209,
generalising tricks, Becker’s notion of, 244 224–35, 252–9, 271
George, A., 55, 126, 172, 260, 278 interpretivism, 55, 160
Georgiades, N., 193 interviewing, 94, 146, 210, 217
Gestalt psychology, 52, 59, 64, 179 interviews, 205–6, 207, 210–11
Glaser, B., 77, 231 group, 204, 210, 221
Glasgow Gang Observed, 6, 26 semi-structured, 186, 194, 204, 206,
Goffman, E., 54, 55, 65, 122, 153, 154, 172, 210, 221
216, 234, 254 structured, 206
Gomm, R., 10, 11, 25 unstructured, 130, 156, 196, 204–7,
graffiti, 34, 36, 217 208, 211
grounded theory, 217, 232, 259 intrinsic study, 128
group interview, 204, 210, 221 intuition, 7, 9, 250, 256–60, 274
guesses, informed, 274, 275 IRB (Institutional Review Board), 100, 102
isolation as a researcher, 8, 53, 176
Haig, B., 75
Haldane, J., 267 James, W., 166
Hammersley, M., 10, 11, 25, 26, 75, 80, journalism, 245–6
81, 84
Herodotus, 80 Kahneman, D., 252
heuristics, 75 Kellett, M., 93
INDEX 297
key case, 22, 38, 68, 109, 111–2, 115, normal distribution curve, 40, 41
124, 142 NVivo, 230–1, 233, 259
kinds of case study. See types of case study
Köhler, W., 53 object of the case study, 15–24, 31, 109,
Korean War, 15 127–41, 145, 181, 223, 256
Kounin, J., 167 observation, 23, 44, 47, 72, 82, 93–4, 96,
Kuhn, T., 252 166, 170, 178, 186, 203, 204, 208, 215–7,
Kurtz, M., 283 221–2, 266, 268
structured, 204, 215–6, 221
Le Play, F., 4, 187 unstructured, 204, 208, 215, 216–7,
legal case studies, 159, 175 221, 225
Lewin, K., 53, 58–9, 238 Office for National Statistics, 219
Lewis, A., 87, 93 official statistics, 44, 204, 221
literature review, 23, 30–3, 41, 47, 185, 203, Omaar, R., 39–44, 48, 203
245, 247–8, 262–3, 269–70 One Day in the Life of Ivan Denisovich.
Loader, D., 261 See Solzhenitsyn, A., 179
local knowledge case, 22, 68, 111, 112, online, 98, 101, 103, 211, 212,
124, 125 optical character recognition (OCR)
longitudinal study, 72 software, 212
Loxley, A., 39, 217–8 opt-in versus implied consent, 91
Luker, K., 172 Orwell G., 278
outlier case, 4, 22, 24, 42, 43, 48, 68, 109,
MacIntyre, A., 52, 65, 74, 75, 77, 78 111, 112, 118–25
macrosystem. See ecological systems theory
Malinowski, B., 160, 172 paradigm shift, 61
Marx, K., 244, 267 paradigm wars, 51, 257
Mead, G.H., 55 paradigms, 83
measurements and tests, 218–9 parallel and sequential studies, 178
memos, as part of coding, 233 Parlett, D., 8
mesosystem. See ecological systems theory participant observation, 204, 208, 216–7, 221
metaphors, use of, 154, 188, 190, 191, 236 Patrick, J., 6, 26
methodology, 23, 73, 84, 99, 110, 232, 263 pattern-finding, 53
methods Paxman, J., 267
of data collection, 203–222 Peirce, C., 74, 75, 77
of analysis, 223–260 personal experience, using, 16, 68, 78, 111, 188
microsystem. See ecological systems theory PGCE qualification, 113
Milgram, S., 85, 88 photo-elicitation, 217
Miller, S., 75 photographs, 148, 178, 186, 203, 211, 217
Mitchell, J., 14, 76 phronesis, 78–9, 84, 129, 153, 283
Mouzelis, N., 77 piloting, 44, 67, 102, 117, 195, 234
multiple study, 186–191, 192, 195, 197 Plato, 51, 79, 274
mushy heap syndrome, 29, 265, 266, 280 plausibility, 69–71
Polanyi, M., 78, 257
narrative, 79, 131, 178–80, 248–50, 253–66 polyhedron of intelligibility, 5, 72
narrative diachronicity, Bruner’s Popper, K., 53, 256
concept of, 250 population, 5, 40, 66–7
natural sciences, 7, 50–2, 70, 74–5, 166 positionality, 66, 69, 73, 83, 209, 232
nested case studies, 38, 42–4, 174, 191–5, potential explanations, 119, 131, 132, 137,
197, 198 141, 185
network analysis, 224 practitioners, 154, 159
newspaper stories, 178, 246 prediction, 52, 76, 232
Newton, I., 3, 54 presentation, 74, 91, 191
Nietzsche, F., 6 prima facie question, 30–2, 39, 41, 44, 48,
nomothetic research, 52 203, 245, 262
proofreading, 278 Sacks, O., 272–3, 282

Prosser, J., 217, 218, 222 Salkind, N., 222
pseudonyms, 89, 178 samples and sampling, 5, 10, 11, 66–9,
psychology, 52–61, 68, 80, 112, 181, 188 246, 265
psychometrics, 68, 282 See also, selecting your case
public health, 136–7, 141 Schatzman, L., 75
Putnam, R., 177 school governor (interpretative case study),
162, 165
quality, in research, 66–83 school inclusion policy (nested case study),
qualitative research, 83, 84, 98, 228, 231, 192–3
259, 260 Schwandt, T., 69, 81, 84, 187
See also, interpretative inquiry scientific method, 7–9, 74
quantitative research, 83, 84, 125, 142 second-order constructs, 226, 233
questionnaires, 47, 204, 212–5, 220–2 security, of data, 90
questions selecting your case, 109–24
kinds of, 44–6, 165 Sen, A., 121
rating-scale, 214 Sennett, R., 154, 164, 187, 188–91,
research, 27–29, 32, 40, 45, 110, 167, 199, 258
262–3 serendipity, 12, 249
quotations, 224, 226, 228, 230, 265, 266 simile in multis, 51
Simons, H., 10, 26, 93, 104
Ragin, C., 11, 26, 84, 252 single case, versus multiple case, 175–8, 198
Ramsey, F., 257 Smail, D., 62
reasoning from, 253 Smith F., 68, 158, 222
reductionism, 50–2 snapshot case study, 23–4, 43–4, 178–83, 198
reflective journals, 204, 207–9, 222 Snyder, R., 283
reflective thought, 268 social capital, 177
See also, criticality, in thinking social media, 92, 98, 220–2
and reasoning social sciences, 5, 7, 8, 41, 51, 70–5, 80, 126,
register of writing, 276, 277 166, 172, 235, 252, 256, 272
reliability, 66–9, 71, 77, 80, 82, 256 sociograms, 224, 270
repeated measures design, 9, 168, 169 sociology, 55, 59
representativeness of data, 11 sociometry, 236
research Socrates, 51, 257, 274
approach, 45 soft systems, 60–1, 65, 238, 240
design, 27–48, 45–7, 119, 165, 262 Solzhenitsyn, A., 179, 180
idea, 32 special needs, 87, 267, 269
problem, 27 special schools, 269
question, 27–29, 32, 40, 45, 110, 167, Spradley, J., 173
262–3 Sri Lanka, 118, 119
retrospective study, 178, 179 Stake, R., 9, 10, 14, 26, 128, 129, 186
rigour. See quality statistics, 140, 167, 170, 172, 219, 222, 224
risk, assessing for ethics 92, 94 eyeballing, 170
assessment, 95 official, 204
management, 96 Stoeker, R., 283
risky shift phenomenon, 210 Storr, A., 250
Roethlisberger, F., 183 storyboards, 48, 82, 117, 142, 240–1, 244
room management, 169–70 Street Corner Society. See Whyte, W.
root definition, in soft systems, 239 structured reflection, 77
Rorty, R., 7 subject and object (of case study), 14–24
Rosenberg Self Esteem Scale, 214 subject of the case study, 14–15, 22–4, 43, 48,
Rosenblatt, P., 283 141, 171, 181, 203, 223
Rutter, M., 112, 113 key case as subject, 22, 38, 68, 109, 111–12,
Ryle, G., 233 115–17, 124, 142
INDEX 299
local knowledge case as subject, 22–3, 38, types of case study

68, 109, 111–14, 124–5 diachronic case study, 178, 183, 269
outlier case as subject, 42, 43, 48, 68, 112, evaluative case study, 218
118–22 experimental case study, 73, 186
subjectivity, 73 explanatory case study, 135
survey, 10, 212, 220 exploratory case study, 135–7, 145
SurveyMonkey, 212, 220 illustrative case study, 152, 155, 157, 217
symbolic interactionism, 55, 57, 63 instrumental case study, 129
synthesis, 31, 259, 270 interpretative case study, 73, 15–60, 162,
systems thinking, 59, 60–1, 64, 194, 238–9, 205–6, 263
244, 259 intrinsic case study, 128
multiple case study, 186–9, 191, 199, 281–2
tacit knowledge, 78, 189, 257–8 nested case study, 43
taking notes, 114, 233 parallel case study, 195
teachers, 193–5, 218, 230, 269 sequential case study, 195–6
teaching assistants, 132, 134, 168 single case study, 23, 38, 175, 178
temporary constructs, 225–7, 232 snapshot case study, 179, 182
tests, 41, 68–9, 78, 172, 204, 218–9, 221 typology of case study, 21–5, 38, 109, 123,
textbooks, 67, 225 143, 198
The Craftsman. See Sennett, R.
theme mapping, 23, 259 uniqueness, 4, 10, 34, 38, 122, 251, 283
themes, from constant comparison, 31, 137,
148, 149, 209, 224–35, 244, 255, 259, 260 validity, 66, 68–71, 78, 80, 82, 142
theorisation, 38, 161, 162, 164, 247 variables, 8, 11, 29, 50, 53, 57, 63, 69, 160,
theorising, 164, 244, 248, 251 166, 167–8, 170, 184, 224, 251
theory, 76–9, 144–71, 242–48 verbal and non–verbal cues, 206
as glue, 242–3, 265, 266 verisimilitude, Bruner’s concept of, 131
building, 41, 142–8 video, 134, 148, 170, 179, 186, 207, 211, 217,
developing, 242–48 220, 225, 230
grand, 161, 162, 244 von Bertalanffy, L., 60
practical, 78 vulnerable groups, 86
testing, 122, 150, 151 Vygotsky, L., 188
thick description, 6, 23, 172, 209, 233–4,
251, 259 walk-throughs, 196–7
thinking critically. See criticality, in thinking websites, 98, 103, 219, 222
and reasoning Weltanschauung, 239
Thomas, G., 8, 21, 26, 39, 72, 77, 84, 111, Wertheimer, M., 53
126, 143, 167, 173, 194, 199, 204, 222, wheelchair users, 196–7
224, 230, 232, 242, 269, 271, 282, 283 White, H.,, 3, 122
Thomson, A., 281 Whyte, W., 160, 161, 173, 271
thought experiments, 44, 70 Wieviorka, M., 11, 12, 15–17, 147
Thucydides, 80 Windelband, W., 52
time, as a dimension of the case study, Wittgenstein, L., 13, 257
178–86 Wolcott, H., 282
timeline, 180 World War II, 11, 15, 223
Tobin, J., 84, 283 write-up, 86, 99, 194, 233, 238, 262–83
transcripts, 83, 203, 224, 225, 234 ‘writerly’ and ‘readerly’ texts, 252
triangulation, 71–3, 121 writing, quality of, 270–78
Trobriand Islanders, 160, 172
Twitter, 98, 220 Yin, R., 191, 199

Case Study

Uploaded by

Copyright:

Available Formats

Case Study

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Case Study

Uploaded by

Copyright:

Available Formats

HOW TO DO YOUR

Los Angeles | London | New Delhi | Singapore | Washington DC | Melbourne

Editor: Jai Seaman

STAGE 1 GETTING YOUR BEARINGS xviii

1 What is a case study? 3

2 The case study and research design 27

3 The whole is more than the sum of the parts: seeing a

Dramas, theatres and stages 54

4 Rigour and quality in your case study: what is important? 66

STAGE 2 CHOOSING A CASE AND STRUCTURING YOUR STUDY 106

6 Different kinds of case studies: selecting a subject for

7 Your purpose: thinking about the object of your study 127

8 Your approach: theory testing or theory building;

Nested case studies 191

STAGE 3 COLLECTING EVIDENCE, ANALYSING AND WRITING IT UP 200

11 Analysis: a toolkit for analysing and thinking in case study 223

12 Writing your study 262

So, this book is organised by these stages. Here’s

What a case is study all about

STAGE 2. CHOOSING A CASE AND

How case studies are categorised

STAGE 3. COLLECTING EVIDENCE,

Ways to collect data for a case study

STAGE 1. GETTING YOUR BEARINGS

What a case study is all about

STAGE 2. CHOOSING A CASE AND

How case studies are categorised

STAGE 3. COLLECTING EVIDENCE,

Ways to collect data for a case study

• s ome ground rules for the conduct and design of

• s ome of the ways in which the success of a case

• the ethical dimensions of case study.

Singleness is the watchword with a case study.

Sometimes, this is called an idiographic approach. An idiograph is a little figure,

practical rationality and judgment evolve and operate primarily by vir-

is why … Richard Rorty [says]: ‘the way to re-enchant the world … is to

Is the case study scientific?

isolation and manipulation of variables and carefully controlled experiments.

Case study is not a methodological choice but a choice of what is

Case study is an in-depth exploration from multiple perspectives of the

Case study Experiment Survey

To Hammersley and Gomm’s table I would add …

Case study Experiment Survey

The … case-oriented approach places cases, not variables, centre stage.

For a ‘case’ to exist, we must be able to identify a characteristic unit …

Wieviorka’s insight is an important one, stressing the ‘analytical category’ – the

In summary, definitions of case study stress

Table 1.2 Meanings given for ‘case’

Container Situation Argument

• capsa, meaning ‘box’, ‘container’ and, thus, ‘case’

The case as container

The case as situation, event

The case as argument

The subject and object of a case study

How the subject and object connect

• be a good example of that analytical frame

Table 1.3 What is and is not a case?

Subject (the ‘practical What would not (on its own)

Figure 1.2 Putting the two parts together

A case study of a quiet child: Bringing subject

STAGE 1 GETTING YOUR BEARINGS xviii

1 What is a case study? 3

2 The case study and research design 27

Dramas, theatres and stages 54

4 Rigour and quality in your case study: what is important? 66

STAGE 2 CHOOSING A CASE AND STRUCTURING YOUR STUDY 106

7 Your purpose: thinking about the object of your study 127

Nested case studies 191

STAGE 3 COLLECTING EVIDENCE, ANALYSING AND WRITING IT UP 200

11 Analysis: a toolkit for analysing and thinking in case study 223

12 Writing your study 262

• s ome ground rules for the conduct and design of

• s ome of the ways in which the success of a case

• the ethical dimensions of case study.