Business Research Methods Unit I
Business Research Methods Unit I
Business Research Methods Unit I
UNIT I
INTRODUCTION
The Advanced Learner‟s Dictionary of Current English lays down the meaning of research as,
“a careful investigation or inquiry specially through se rch for new facts in any branch of
knowledge”.
SOME DEFINITIONS
Redman and Mory define research as a,” Systematized effort to gain new knowledge”. Some
people consider research as a movement, a movement from the known to the unknown.
According t Clifford woody, research comprises defining and redefining problems, formulating
hypothesis or suggested solutions collecting, organising and evaluating data, making deductions and
reaching conclusions; to determine whether they fit the formulating hypothesis.
OBJECTIVES OF RESEARCH
3. To determine the frequency with which something occurs or with which it is associated
with something else. (studies with this object known as diagnostic research)
4. To test a hypothesis of a causal relationship between variables. (such studies are
known as hypothesis testing research)
TYPES OF RESEARCH
It is imperative that a marketer has to have a broad understanding of the various types of research,
in general. There are eleven types of research depending on whether it is primarily “fundamental”
or “applied” in nature. They are as follows:
1. Applied research, also known as dec s o al research, use existing knowledge as an aid to
the solution of some given problem or s t of problems.
2. Fundamental research, frequently called basic or pure research, seeks to extend the
boundaries of kno ledge in a given area with no necessary immediate application to existing
problems.
3. Futuristic research: Futures research is the systematic study of possible future conditions.
It includes analysis of how those conditions might change as a result of the implementation of
policies and actions, and the consequences of these policies and actions.
4. Descriptive research includes surveys and fact-finding enquiries of different kinds. It tries
to discover answers to the questions who, what, when and sometimes how. Here the researcher
attempts to describe or define a subject, often by creating a profile of a group of problems, people,
or events. The major purpose of descriptive research is description of the state of affairs as it exists
at present
7. Analytical research: The researcher has to use facts or information already available, and
10. Conceptual research: Conceptual research is that related to some abstract idea(s) or
theory. It is generally used by philosophers and thinkers to develop new concepts or to reinterpret
existing ones.
11. Empirical research: It is appropriate when proof is sought that certain variables affect
other variables in some way. Evidence gathered through experiments or empirical studies is today
considered to be the most powerful support possible for a give hypothesis.
Several authors have attempted to enumerate the steps involved in the research process, however,
inconclusive. Nevertheless, the research process broad y onsists of the following steps and
predominantly follows a sequential order as depicted in fig re 1.1.
1. Problem formulation
3. Research Design
5. Sampling techniques
First there is initiating or planning of a study, which comprises the initial four steps in our model:
determining (1) problem formulation, (2) development of an approach to the problem (3) Research
design (4) selection of data collection techniques (5) sampling techniques.
Third, there is (7) analysis and interpretation of the data and (8) report preparation and
presentation.
PROBLEM IDENTIFICATION
The starting point of any research is to formulate the prob em and mention the objectives before
specifying any variables or measures. This involved defining the problem in clear terms. Problem
definition involves stating the general roblem and identifying the specific components of the
research problem. Components of the research problem include (1) the decision maker and the
objectives (2) the environment of the problem (3) alternative courses of action (4) a set of
consequences that relate to courses of action and the occurrence of events not under the control of
the decision maker and (5) a state of doubt as to which course of action is best. Here, the first two
components of the research problem are discussed whereas others are not well within the scope,
though, not beyond.
Problem formulation is perceived as most important of all the other steps, because of the fact that
a clearly and accurately identified problem would lead to effective conduct of the other steps
involved in the research process. Moreover, this is the most challenging task as the result yields
information that directly addresses the management issue, though, the end result is for the
management to understand the information fully and take action based on it. From this we
understand, that the correctness of the result depends on how well the research takes on, at the
starting point.
Problem formulation refers to translating the management problem into a research problem. It
involves stating the general problem and identifying the specific components of research problem.
This step and the findings that emerge would help define the management decision problem and
research problem.
Problem formulation starts ith a sound information seeking process by the researcher. The decision
maker is the provider of information pertaining to the problem at the beginning of the research
process (problem formulation) as well as the user of the information that germinates at the end of
the research process. Given the importance of accurate problem formulation, the research should
take enough care to ensure that information seeking process should be well within the ethical
boundaries of a true research. The researcher may use different types of information at the
problem formulation stage. They are:
1. Subjective information termed as those based on the decision maker‟s past experiences,
expertise, assumptions, feelings or judgments without any systematic gathering of facts. Such
information is usually readily available.
2. Secondary information are those collected and interpreted at least once for some specific
situation other than the current one. Availability of this type of information is normally high.
3. Primary information refers to first hand information derived through a formalised research
process for a specific, current problem situation.
In order to have better understanding on problem formulation, the researcher may tend to
categorise the information collected into four types. The categorisation of the information is done
based on the quality and complexity of the inform tion collected. They are:
1. Facts are some piece of information w th very high quality information and a higher degree
of accuracy and reliability. They could be absolutely observable and verifiable. They are not
complicated and are easy to unde stand and use.
2. Estimates are information hose degree of quality is based on the representativeness of the fact
sources and the statistical procedures used to create them. They are more complex than facts due
to the statistical procedures involved in deriving them and the likelihood of errors.
3. Predictions are lower quality information due to perceived risk and uncertainty of future
conditions. They have greater complexity and are difficult to understand and use for decision-
making as they are forecasted estimates or projections into the future.
The outputs of the approach development process should include the following components: (i)
Objective/theoretical framework (ii) analytical model (iii) Research questions (iv) hypothesis.
(i) Objective/theoretical framework: Every research shou d have a theoretical framework and
objective evidence. The theoretical framework is a conceptual scheme containing:
a set of statements that describes the s tuations on which the theory can be applied
Operationalising the concept gives more understanding on the meanings of the concepts specified
and explication of the testing procedures that provide criteria for the empirical application of the
concepts. Operational definition would specify a procedure that involves say, for example, a
weighing machine that measures the weight of a person or an object.
(ii) Analytical model: An analytical model could be referred to as a likeness of something. It
consists of symbols referred to a set of variables and their interrelationships represented in logical
arrangements designed to represent, in whole or in part, some real system or process. It is a
representation of reality making explicit the significant relationships among the aspects. It enables
the formulation of empirically testable propositions regarding the nature of these relationships. An
empirical model refers to research that uses data derived from actual observation or
experimentation.
(iii) Research Questions: Research questions are refined statements of the specific components
of the problem. It refers to a statement that scert ins the phenomenon to be studied. The research
questions should be raised in an u ambiguous manner and hence, would help the researcher in
becoming resourceful n dentifying the components of the problem. The formulation of the
questions should be strongly guided by the problem definition, theoretical framework and the
analytical model. The knowledge gained by the researcher from his/her interaction with the
decision maker should be borne in mind as they sometimes form the basis of research questions.
The researcher should exercise extreme caution while formulation research questions as they are
the forerunner for developing hypothesis. Any flaw in the research questions may lead to flawed
hypothesis. The following questions may be asked while developing research questions:
a) Do I know the area of investigation and its literature?
c) What are the areas that are not explored by the previous researchers?
f) Is my study a new one thus contributing to the society or has it been done before?
(iv) Hypothesis: Hypothesis could be termed as tentative answers to a research problem. The
structure of a hypothesis involves conjectural statements relating to two or more variables. They
are deduced from theories, directly from observ tion, intuitively, or from a combination of these.
Hypothesis deduced from a y of the means would have four common characteristics. They should
be clear, value-free, specific and amenable to empirical testing.
Hypothesis could be viewed as stat m nts that indicate the direction of the relationship or
recognition of differences in groups. However, the researcher may not be able to frame hypotheses
in all situations. It may be because that a particular investigation does not warrant a hypothesis or
sufficient information may not be available to develop the hypotheses.
UNIT II
INTORDUCTION
With the completion of the initial phase of the research process, the researcher turns to designing a
research design to formally identify the appropriate sources of data. This is done in order that any
researcher who embarks on a research project should have a blueprint of how he is going to
undertake scientifically the data collection process. The framework developed to control the
collection of data is called research design.
Research design is an absolute essentiality in rese rch irrespective of the type of research (e.g.,
exploratory or descriptive), as it ensures that the data collected is appropriate, economical and
accurate. This also ensures that the r s arch project conducted is effectively and efficiently done. A
sufficiently formulated research design would ensure that the information gathered is consistent
with the study objectives and that the data are collected by accurate procedures. Since, research
designs germinate from the objectives, the accuracy and adequacy of a research design depends
on the unambiguous framing of the objectives.
Exploratory design; and Conclusive design. (Descriptive researh and casual research)
It is appropriate when the research objective is to provide insights into (i) identifying the problems
or opportunities (ii) defining the problem more precisely, (iii) gaining deeper insights into the
variables operating in a situation (iv) identifying relevant courses of action (v) establishing
priorities regarding the potential significance of a problems or opportunities (vi) gaining additional
insights before an approach can be developed and (vii) gathering information on the problems
associated with doing conclusive research has been of an exploratory nature; emphasising on
finding practices or policies that needed changing and on developing possible alternatives.
Exploratory research could also be used in conjunction with other research. As mentioned below,
since it is used as a first step in the research process, defining the problem, other designs will be
used later as steps to solve the problem. For instance, it could be used in situations when a firm
finds the going gets tough in terms of sales volume, the researcher may develop use exploratory
research to develop probable explanations. Analysis of data generated using exploratory research
is essentially abstraction and generalization. Abstraction refers to translation of the empirical
observations, measurements etc. into concepts; generalization means arranging the material so that
it focuses on those structures that are common to all or most of the cases.
The exploratory research design is best characterised by its flexibility and versatility. This is so,
because of the absence of the non-imperativeness of a structure in its design. It predominantly
involves imagination, creativity, and ingenuity of the researcher. Examples of exploratory research
are:
survey of experts to validate an instrument; pilot studies conducted to perform reliability check on
a questionnaire; use of secondary data in order to analyse it in a qualitative way; qualitative
research.
It involves providing information on evaluation of alternative courses of action and selecting one
from among a number available to the researcher. As portrayed in the figure 4.1, conclusive
research is again classified as:
(i) Descriptive Research: It is simple to understand as the name itself suggests that it
involves describing something, for example:
(c) estimate the percentage of customers in a particular group exhibiting the same purchase
behaviour;
Majority of research studies are descriptive studies. As research studies involve investigating the
customers/consumers, collection of data includes interrogating the respondents in the market and
data available from secondary data sources. However, it cannot be concluded that descriptive
studies should be simply fact-gathering process. Descriptive study deals with the respondents in
the market and hence, extreme caution has to be exercised in developing this study. Much
planning should be done, objectives sho d be clear than exploratory studies.
In descriptive research, the data is collected for specific and definite purpose and involves analysis
and interpretation by the researcher. The major difference between exploratory and descriptive
research is that desc iptive r search is characterised by the formulation of specific objectives. The
success of desc iptive studies depends on the degree to which a specific hypothesis acts as a guide.
While designing a descriptive research, the researcher should also have sufficient knowledge on
the nature and type of statistical techniques he/she is going to use. This will greatly help to have
the right design in place. Mostly descriptive studies are conducted using questionnaire, structured
interviews and observations. The results of description studies are directly used for marketing
decisions.
(a) Longitudinal
Panel data is analytical and possess advantages with respect to the information collected in the
study. They are also considered to be more accurate than cross sectional data because panel data
better handle the problem associated with the errors that arise in reporting past behaviour and the
errors that arise because of the necessary interaction between interviewer and respondent.
(b) Cross-sectional research is the most predominantly and frequently used descriptive
research design in marketing. It involves a sample of elements from the population of interest. The
sample elements are measured on a number of characteristics. There are two types of cross-
sectional studies:
It may appear that field studies and surveys are no different but the same. However, for practical
reasons, they are classified into two categories cross sectional research. The fundamental
difference lies in the depth of what these research cover. While survey has a larger scope, field
study has greater depth. Survey attempts to be representative of some known universe and filed
study is less concerned with the generation of large representative samples and is more concerned
with the in-depth study of a few typical situations.
Cross sectional design may be ith r single or multiple cross sectional design depending on the
number of samples d awn from a population. In single cross sectional design, only one sample
respondents is drawn whereas in multiple cross sectional designs, there are two or more samples
of respondents. A type of multiple cross sectional design of special interest is Cohort analysis.
Cohort analysis consists of a series of surveys conducted at appropriate time intervals, where the
cohort serves as the basic unit of analysis. A cohort is a group of respondents who experience the
same event within the same time interval.
(a) Case Study: This study involves intensive study of a relatively small number of cases. In
this method, much emphasis is on obtaining a complete description and understanding of factors
in each case, regardless of the number involved. It could be used significantly, particularly when
one is seeking help on a problem in which interrelationships of number of factors are involved,
and in which it is difficult to understand the individual factors without considering them in their
relationships with each other. As in the case of exploratory research, case method is also used in
conjunction with exploratory research as first step in a research process. It is of prime value when
the researcher is seeking help on a market problem in which the interrelationships of a number of
factors are involved, and in which it is difficult to understand the individual fa tors without
considering them in their relationships with each other.
This is an important type of s a ch useful for marketers as this allows marketers to base their
decision on assumed causal relationships. Causal research is done in the following situations:
(a) To identify hich variables are the cause and which are the effect. In statistical terms, causal
variables are called independent variables and effectual variables are called dependent variables.
(b) To determine the nature of the relationship between the causal variables and the effect to
be predicted.
Causal research requires a strong degree of planning on the design as its success depends on the
structure of the design.
THE MEASUREMENT PROCESS
Measurement is defined as the assignment of numbers to characteristics of objects or events
according to rules. The definition of measurement clearly states that the researcher should know
that the measurement scale measures the characteristics of the objects or event and not the objects
or events.
Further, to make the measurement process effective, the relationships existing among the objects
or events in the empirical system should directly correspond to the rules of the number system. If
this correspondence is misrepresented, measurement error has occurred. The term number
indicates the application of numbers to various aspects measured in the measurement process.
Data analysis is a statistical process done on the data generated using scales. Hence, all measures
should be converted into quantitative terms by applying numbers. However, the definition of
measurement imposes certain restrictions on numerical manipulations admissible.
The numerical application on all measurements and the analysis of numbers using mathematical
or statistics involve one or more of the four characteristics of number system. Measurement of
any property could be fitted into any of these characteristics.
LEVELS OF MEASUREMENT
a) Nominal scale
b) Ordinal scale
c) Interval scale
d) Ratio scale
Nominal Scale
Nominal scale are categorical scales used to identify, label or categorise objects or persons or
events. A familiar example is the use of alternative numbering system by our Physical Education
Teacher in our school days to engage us in a game. The teacher as a result would form two groups
one labelled 1 and the other 2. The numbers 1 and 2 are assigned to two groups and the members
belonging to group 1 would exclusively be a part of group 1 and the members belonging to group
2 would exclusively be a part of group 2. However, assigning the numbers does not indicate any
order or position to the group it represents. Interchanging the numbers otherwise would also result
in the same effect in that, the order r p siti n would not change.
Nominal scales are the lowest form of measurement. The simple rule to be followed while
developing a nominal scale: Do not assign the s me numerals to different objects or events or
different numbers to the same object or eve t. In marketing nominal scales are used substantially in
many occasions. For example, nom nal scale is used to identify and classify brands, sales regions,
awareness of brands, wo king status of women etc.,
On data generated using nominal scale, the type of statistical analysis appropriate are mode,
percentages, and the chi-square test. Mode alone could be used as a measure of central tendency.
Mean and median could be employed on nominal data since they involve higher level properties of
the number system. Researchers should be careful enough to identify the type of scales before they
apply any statistical technique. The researcher may not be able to make any meaning inference
from the mean or median value obtained from nominal data.
Ordinal Scale
Ordinal scale is a ranking scale that indicates ordered relationship among the objects or events. It
involves assigning numbers to objects to indicate the relative extent to which the objects possess
some characteristic. It measure whether an object or event has the same characteristic than some
other object or event. It is an improvement over nominal scale in that it indicates an order.
However, this scale does not indicate on how much more or less of the characteristic various
objects or events possess. The term how much refers to ranks that it do not indicate if the second
rank is a close second or a poor second to the first rank.
Data generated using ordinal scale appears as ranks where the bje t which has ranked first has
more of the characteristic as compared to those objects ranked second or third. Hence, the
important feature of ordinal scale over nominal sc le is th t it indicates relative position, not the
magnitude of the difference between the objects. In research, ordinal scales are used to measure
relative attitudes, opinions, perceptions etc., Most data collected by the process of interrogating
people have ordinal properties To illustrate, a marketer may be interested in knowing the
preference of the customers across various brands. The customers may be requested to rank the
products in terms of their preference for the products.
The numbers assigned to a particular object or event can never be changed in ordinal scales. Any
violation of this principle would result in confounding results by the researcher. Mean is not an
appropriate statistic for ordinal scale.
Interval Scale
Interval scale is otherwise called as rating scale. It involves the use of numbers to rate objects or
events. It interval scales, numerically equal distances on the scale represent equal values in the
characteristic being measured. Interval scale is an advancement over the ordinal scale that it has
all the properties of an ordinal scale plus it allows the researcher to compare the differences
between objects. It also possesses the property of equality of difference between each levels of
measurement. The feature of this scale is that the difference between any two scale values is
identical to the difference between any other two adjacent values of an interval scale. Examples of
interval scales are the Fahrenheit and Celsius scales.
Interval scales also place restriction on the assignment of values to the scale points. The zero that
could be assignment is a arbitrary zero rather than a natural zero. Arbitration involves freedom to
place the zero value on any point. There is a constant or equal interval between scale values.
In research, most of the research on attitudes, opinions and perceptions are done using scales
treated as interval scales. All statistical techniques th t re employed on nominal and ordinal scales
could also be employed on data ge erated using interval scales.
Ratio Scales
Ratio scales differ from interval scales in that it has a natural/absolute zero. It possesses all the
properties of the normal, ordinal and interval scales. Data generated using ratio scales may be
identified, classified into categories, ranked and compared with others properties. It could also be
expressed in terms of relativity in that one can be expressed in terms of a division of the other.
Hence, it may be called as relative scales.
Ratio scales have great many number of application in research. They include sales, market share,
costs, ages, and number of customers. In all these cases, natural zero exists. All statistical
techniques can be applied on ratio data.
PERFECT MEASUREMENT
Research should always be based on absolutely correct, defect less and errorless measuring
instruments, tools or procedures of measurement. For this purpose the acceptability of a measuring
instrument should be tested on the principles of adherence to the standards of perfect reliability,
confirmed practicality and verified validity. The reliability of an instrument can be ensured when it
conforms to certain prescribed norms. It is not the physical form or shape but it is the accuracy of
the prescribed standard content of the instrument that leads to acceptability. An instrument should
be conveniently usable with verifiable validity. Perfection in measurement can be achieved if a
researcher, at the outset, carries out appr priately, the prescribed tests of reliability, practical
acceptability and validity of his tools of measurement.
Errors in Measurement
Errors in the course of measurement can be traced to a number of factors such as carelessness,
negligence, ignorance in the usage of the instruments. If appropriate and defectless instruments are
used and care is taken in the p ocess of measurement, only then can accuracy in research be
ensured.
In regard to survey- ork, where the researcher obtains information through interviews, it is
necessary, to judge as to whether the respondent is providing accurate facts or is biased. As
situational factors also influence measurement, it is imperative that the researcher adopts his
measuring procedures accordingly.
Research findings and conclusions can be reliable and acceptable if they are based on sound
analysis carried out through appropriate procedures of error-free and perfect measuring tools.
SCALING TECHNIQUES
A well-designed research problem constitutes a well designed measurement process. The process
of measurement is a fundamental aspect of any research. This is the step where you actually try to
find out the reality by measuring it. Decision makers are more interested as the steps prior to this
step are purely descriptive, and, this is the step where actual quantification happens.
Developing effective measures of marketing is not an easy task. The measures should be devoid of
measurement errors. There may be disastrous situations where the arketer may be confused with
the findings of the data. If he is well aware of the confounding results, then he may discard the
findings the emerge from the data analysis. This req ires ot of wisdom and knowledge in
identifying if the data that resulted from the me surement is consistent, unambiguous etc., But
unfortunately, marketers may not be interested in knowing or rather would not know the type of
scales used to measure the aspects involved n the marketing problem. Any decision made based on
the findings would lot of negative implications on the organisation. Hence, it is very imperative
that the researcher is ise enough to develop measurement scales that capture the right property
with appropriately.
The scaling techniques employed in research could be broadly classified into comparative and non
comparative scale. Comparative scales as its name indicate derive their name from the fact that all
ratings are comparisons involving relative judgements. It involves direct comparison of stimulus
objects. It contains only ordinal or rank order properties. It is also otherwise called non metric
scales in that it does not allow any numerical operations on it against all that could be applied on
interval and ratio scales. Comparative scales involve the direct comparison of stimulus objects.
d) Q-sort.
Paired comparison scaling as its name ind cates i volves presentation of two objects and asking the
respondents to select one according to some criteria. The data are obtained using ordinal scale. For
example, a respondent may be asked to indicate his/her preference for TVs in a paired manner.
Paired comparison data can be analysed in several ways. In the above example, the researcher can
calculate the percentage of respondents who prefer one particular brand of TV over the other.
Under the assumption of transitivity, data generated using paired comparison technique could be
converted to a rank order. Transitivity of preference implies that if a respondent prefers brand X
over brand Y, and brand Y is preferred to Z, then brand X is preferred to Z. This may be done by
determining the number of times each brand is preferred by preference, from most to least
preferred.
Paired comparison technique is useful when the number of brands is limited, as it requires direct
comparison and overt choice. However, it is not so, that possible comparison could not be made,
but comparisons would become so much unwieldy.
The most common method of taste testing is done by paired comparison where the consumer may
be, for example, asked to taste two different brands of soft drinks and select the one with the most
appealing taste.
Data generated using this technique are employed with conjoint analysis because of the
discriminatory potential of the scaling, stimulating the consumers to discriminate one brand from
the other.
Under the assumptions of transitivity, rank order can be converted to equivalent paired comparison
data, and vice versa.
This technique allows the respondents to allocate a constant sum of units, such as points, rupees or
among a set of stimulus objects with respect to some criterion. The technique involves asking the
respondents to assign 10 points to attributes of a sports utility vehicle. If the attribute is
unimportant, then the respondents would want to enter zero.
The attributes are scaled by counting the points assigned to each one by al the respondents and
dividing the number of respondents. This predominantly uses ordinal because of its comparative
nature and the resulting lack of generalisability. Constant sum scaling has advantage in that it
allows for discrimination among stimulus objects without requiring too much time. Its advantage
involves allocation of more or fewer units than those specified.
Q-Sort
Q-sort refers to discriminating among a relatively large number of objects quickly. This technique
uses a rank order procedure in which objects are sorted into piles based on similarity with respect
to some criterion. A typical exam le quoted in Malhotra (2004) is as follows:
Respondents are given 100 attitude stat m nts on individual cards and asked to place them into 11
piles, ranging from „most highly ag d with‟ to „least highly agreed with‟. The number of objects
to be sorted should not be less than 60 nor more than 140: 60 to 90 objects is a reasonable range.
The number of objects to be placed in each pile is pre-specified, often to result in a roughly normal
distribution of objects over the whole set.
Non-comparative scales or otherwise called as nomadic scales because only one object is
evaluated at a time. Researchers use this scale allowing respondents to employ whatever rating
standard seems appropriate to them and not specified by the researcher. The respondents do not
compare the object being rated either to another object or to some specified standard set by the
researcher. Non-comparative techniques use continuous and itemised rating scales.
In such scales, each object is scaled independently of the other objects in the stimulus set, the
resulting data is generally assumed to be interval or ratio scale.
This is also otherwise called as graphic rating scale. This is a type of scale that offers respondents
a form of continuum (such as a line) on which to provide a rating of an object. Researchers
develop continuous rating scale allowing the resp ndents to indicate their rating by placing a mark
at the appropriate point on a line that runs from one end of the criterion variable to the other or a
set of predetermined response categories. Here the respondents need not select marks already set
the researcher.
There are several variations that are poss ble. The line may be vertical or horizontal; it may be
unmarked or marked; if marked, the divisions may be few or as many as in the thermometer
scale; the scale points may be in the form of numbers or brief descriptions. Three versions are
Please evaluate the service quality of a restaurant by placing an x at the position on the
horizontal line that most reflects your feelings Empathy
Continuous rating scales are easy to construct, however, the scoring may be cumbersome and
unreliable. With the advent of computers in research, they are increasingly used, though, they
otherwise provide little new information.
This scale is similar to the graphic scale in that the individuals make their judgement
independently, without the benefit of direct comparison. The respondents are provided with a scale
that has a number or brief description associated with each category. This scale allows the
respondents to choose from a more limited number of categories, usually five to seven, although
10 or more are rarely used. The categories are ordered in terms f scale position; and the
respondents are required to select the specified category that best describes the object being rated.
The categories are given verbal description, lthough this is not absolutely necessary. These scales
are widely used in research a d owadays, more complex types such as multi-item rating scales are
used. There are f w var ants among itemised rating scales. They are Likert, Semantic differential
and stapel scales
Likert Scale
This scale is named after Renis Likert. This is the most widely used scale in research, in particular,
in testing models. Several research studies are done using Likert scale. The respondents require to
indicate a degree of agreement of disagreement with each of a series of statements about the
stimulus objects. Example of a portion of a popularly used Likert scale to measure tangibility of
service is given below.
Listed below are the tangibility of service rendered by a bank is given below. Please indicate how
strongly you agree or disagree with each by using the following scale
1 = Strongly disagree
2 = Disagree
4 = Agree
5 = strongly agree
To analyse the data generated using this scale, each statement is assigned a numerical score,
ranging either from -2 to +2 through a zero or 1 to 5. The analysis can be conducted item wise or a
total score (summated) or a mean can be calculated for each respondent by summing or averaging
across items. It is important in Likert scale that a onsistent scoring procedure so that a high score
reflects favourable response and a low score reflects unfavourable response. Any deviation in the
form of reverse coding where the lowest value is given to a favourable response and highest value
is given to an unfavourable response should be clearly specified by the researcher. Usually, reverse
coding is us d when the statements indicate a negative concept and when used with other
statements, reverse coding would give a positive effect.
Semantic Differential Scale
Semantic differential scale is a popular scaling technique next to Likert scale. In this scale, the
respondents associate their response with bipolar labels that have semantic meaning. The
respondents rate objects on a number of itemised, seven point rating scales bounded at each end by
one of two bipolar adjectives such as “Excellent” and “Very bad”. The respondents indicate their
response choosing the one that best describes their choice.
The points are marked either from - 3 to +3 through a zero or from 1 to 7. The middle value may
be treated as a neutral position. The value zero in the first type is the neutral point and 4 in the
second type is the neutral point. The resulting data are commonly analysed through profile
analysis. In such analysis, the means or median values on each rating scale are calculated and
compared by plotting or statistical analysis. This would help the researcher to determine the
overall differences and similarities among the objects.
To assess differences across segments of respondents, the researcher can compare mean responses
of different segments. This data generated using this scale could be employed with summary
statistics such mean, though, there is a controversy on the employment of mean on this scale.
Mean is typical of Interval and ratio scales whereas this scale theoretically is an ordinal scale.
However, looking beyond this objection by statisticians, researchers invariably apply all statistical
techniques on this scale. The following example illustrates semantic differential scales
1) Pleasant unplea
------------------------------------------------------- sant
2) Aggressive
submi
-------------------------------------------------------
ssive
3) Exciting unexci
-------------------------------------------------------- ting
Stapel Scale
This scale is named after Jan Stapel, who developed it. This is a unipolar rating scale with in
general 10 categories number from -5 to +5 without a neutral point (zero). This scale is usually
presented vertically and respondents choose their response based on how accurately or
inaccurately each item describes the object by selecting an appropriate numerical response
category. The higher number indicates more accurate description of the object and lower number
indicates lower description of the object. An example is given below:
+ 5
+4
+3
+2
+1
-1
-2
-3
-4
-5
The data generated using staple scale could be an lysed in the same way as semantic differential
scale. The main advantage of Stapel Scale s that it does not require a pretest of the adjectives or
phrases to ensure true bipolarity, and it can be administered over the telephone.
DATA COLLECTION
The next step in the research process after identifying the type of research the researcher intends to
do is the deciding on the selection of the data collection techniques. The data collection technique
is different for different types of research design. There are predominantly two types of data: (i)
the primary data and (ii) the secondary data.
Primary data is one a researcher collects for a specific purpo se f investigating the research
problem at hand. Secondary data are ones that have not been collected for the immediate study at
hand but for purposes other than the problem at hand. Both types of data offer specific advantages
and disadvantages.
a) Secondary data offer cost and time economies to the researcher as they already exist in various
forms in the company or in the market.
b) It is feasible for a firm to collect.
c) Since they are collected for some other purposes, it may sometimes not fit perfectly into
the problem defined.
d) The objectives, nature and methods used to collect the secondary data may not be
appropriate to the present situation.
SECONDARY DATA
Secondary data are the data that are in actual existence in accessible records, having been already
collected and treated statistically by the persons maintaining the records. In other words,
secondary data are the data that have been already collected, presented tabulated, treated with
necessary statistical techniques and conclusions have been drawn. Therefore, collecting secondary
data doesn't mean doing some origin l en meration but it merely means obtaining data that have
already been collected by some agencies, reliable persons, government departments, research
workers, dependable organisations etc. Secondary data are easily obtainable from reliable records,
books, gov rnment publications and journals.
When once primary data have been originally collected, moulded by statisticians or statistical
machinery, then it becomes secondary in the hands of all other persons who may be desirous of
handling it for their o n purpose or studies. It follows, therefore, that primary and secondary data
are demarcated separately and that the distinction between them is of degree only. It a person 'X'
collects some data originally, then the data is primary data to 'X' whereas the same data when used
by another person 'Y' becomes secondary data to 'Y'.
Though the given list of secondary data cannot be said to be thorough or complete, yet it can be
pointed out that it fairly indicates the chief sources of secondary data. Also, besides the above
mentioned data there are a number of other important sources, such as records of governments in
various departments, unpublished manuscripts of eminent scholars, research workers, statisticians,
economists, private organisations, labour bureaus and records of business firms.
Secondary data are of two types. Data that are originated from within the company are called as
internal data. If they are collected for some other purpose, they are internal secondary data. This
poses significant advantage as they are readily available in the company at low cost. The most
convenient example internal secondary data is the figures relating sales of the company. Important
internal source of secondary data is database marketing, Database marketing involves the use of
computers to capture and track customer profiles and purchase details. The information about
customer profile would serve as the foundation for arketing programmes or as an internal source
of information related to preference of ust mer‟s preference of a particular product.
Published external secondary data refers to the d ta v il ble without the company. There is such a
pool of published data available in the market that it is sometimes easy to underestimate what is
available and thereby bypass relevant nformation. Several sources of external data are available.
They are:
Directories are helpful for identifying individuals or organisations that collect specific data.
Indexes used to locate information on a particular topic in several different publications by using
an index.
Government Sources
Census data is a report published by the Government containing information about the population
of the country.
Other Government publications may be pertaining to availability of train tickets just before it
leaves.
Computerised Databases
Online databases are databases consisting of data pertaining to a particular sector (e.g.,
banks) that is accessed with a computer through a telecommuni ati n network
Internet databases are available in internet portals that can be accessed, searched, and
analysed on the internet.
Offline databases are databases ava lable in the form of diskettes and CD-ROM disks.
Directory databases provide information on individuals, organisations and service. E.g. Getit
Yellow pages.
External Data-syndicated In response to the growing need for data pertaining to markets,
consumer etc., companies have started collecting and selling standardised data designed to serve
the information needs of the shared by a number of organisations. Syndicated data sources can be
further classified as
(a) consumer data (b) retail data (c) wholesale data (d) industrial data (e) advertising evaluation
data and (f) media and audience data.
Consumer data relates to data about consumers purchases and the circumstances surrounding the
purchase.
Retail data rely on retailing establishments for their data. The data collected focus on the products
or services sold through the outlets and / or the characteristics of the outlets themselves.
Wholesale data refers to data on warehouse shipment data to esti ate sales at retail.
Industrial data refers to substantially more syndicated data services available to consumer goods
manufacturers than to industrial goods suppliers.
Advertising evaluation data refers to money s ent each year on media such as magazines and
television with the expectation that these expenditures will result in sales.
(i) Whether the organization that has collected the data is reliable.
(ii) Whether the appropriate statistical methods were used by the primary data enumerators
and investigators.
By primary data we mean the data that have been collected originally for the first time. In other
words, primary data may be the outcome of an original statistical enquiry, measurement of facts or
a count that is undertaken for the first time. For instance data of population census is primary.
Primary data being fresh from the fields of investigation is very often referred to as raw data. In
the collection of primary data, a good deal of time, money and energy are required.
QUESTIONNAIRE
a) It must translate the information needed into a set of specific questions that the
respondents can and will answer.
Questionnaire Components
a) Identification data
c) Instruction
d) Information sought
e) Classification of data
a) Identification data occupation is the first section of a questi nnaire where the researcher
intends to collect data pertaining to the respondent‟s name, address and phone number.
b) Request for cooperation refers to gaining res ondent‟s cooperation regarding the data
collection process.
c) Instruction refers to the comm nts to the respondent regarding how to use the
questionnaire.
d) The information sought form the major portion of the questionnaire. This refers to the
items relating to the actual area of the study.
This is another type of method used when the researcher feels that survey type of methods may not be
so relevant in data collection. In subjective issues, respondents need to be observed rather than asked
lest biases and prejudices happen in their response. Observation method may be either structured or
unstructured. Structured observation method involves having a set of items to be observed and
how the measurements are to be recorded. In unstructured observation, the observer monitors all
aspects of the phenomena that seem relevant to the problem at hand. In this context, the observer
may have an open mind to study the persons or object.
SAMPLING DESIGN
research does not exist without sampling. Every research study requires the selection of some kind
of sample. It is the life blood of research.
Any research study aims to obtain information about the characteristics or parameters of a
population. A population is the aggregate of all the elements that share some common set of
characteristics and that comprise the universe for the purpose of the research problem. In other
words, population is defined as the totality of all cases that conform to some designated
specifications. The specification helps the researcher to define the elements that ought to be
included and to be excluded. Sometime s, groups that are of, interest to the researcher may be
significantly smaller allowing the researcher to collect data from all the elements of population.
Collection of data from the entire population is referred to as census study. A census involves a
complete enumeration of the elements of a population.
Collecting data from the aggregate of all the elements (population) in case of, the number of
elements being larger, would sometimes render the researcher incur huge costs and time. It may
sometimes be a remote possibility. An alternative way would be to collect information from a
portion of the population, by taking a sample of elements from the population and the on the basis
of information collected from the sample elements, the characteristics of the population is inferred.
Hence, Sampling is the process of selecting units (e.g., people, organizations) from a population of
interest so that by studying the sample we may fairly generalize our results back to the population
from which they were chosen.
While deciding on the sampling, the researcher should clearly define the target population without
allowing any kind of ambiguity and inconsistency on the boundary of the aggregate set of
respondents. To do so, the researcher may have to use his wisdom, logic and judgment to define
the boundary of the population keeping with the objectives of the study.
Sampling techniques are classified into two broad categories f probability samples or non-
probability samples.
Probability samples are characterised by the fact that, the sampling units are selected by chance. In
such case, each member of the population has a known, non-zero probability of being selected.
However, it may not be use that all samples would have the same probability of selection, but it is
possible to say the probability of selecting any particular sample of a given size. It is possible that
one can calculate the probability that any given population element would be included in the
sample. This requires a precise definition of the target population as well as the sampling frame.
Probability sampling techniques differ in terms of sampling efficiency which is a concept that
refers to trade off between sampling cost and precision. Precision refers to the level of uncertainty
about the characteristics being measured. Precision is inversely related to sampling errors but
directly related to cost. The greater the precision, the greater the cost and there should be a trade-
off between sampling cost and precision. The researcher is required to design the most efficient
sampling design in order to increase the efficiency of the sampling.
Probability sampling techniques are broadly classified as simple random sampling, systematic
sampling, and stratified sampling.
This is the most important and widely used probability sampling technique. They gain much
significance because of their characteristic of being used to frame the concepts and arguments in
statistics. Another important feature is that it allows each element in the population to have a
known and equal probability of selection. This means that every element is selected independently
of every other element. This method resembles lottery method where a in a system names are
placed in a box, the box is shuffled, and the n mes of the winners are then drawn out in an
unbiased manner.
Simple random sampling has a definite proc ss, though not, so rigid. It involves compilation of a
sampling frame in which each element is assigned a unique identification number. Random
numbers are generated either using random number table or a computer to determine which
elements to include in the sample. For example, a researcher is interested in investigating the
behavioural pattern of customers while making a decision on purchasing a computer. Accordingly,
the researcher is interested in taking 5 samples from a sampling frame containing 100 elements.
The required sample may be chosen using simple random sampling technique by arranging the
100 elements in an order and starting with row 1 and column 1 of random table, and going down
the column until 5 numbers between 1 and 100 are selected. Numbers outside this range are
ignored. Random number tables are found in every statistics book. It consists of a
Randomly generated series of digits from 0 – 9. To enhance the readability of the numbers, a space
th th
between every 4 digit and between every 10 row is given. The researcher may begin reading
from anywhere in the random number table, however, once started the researcher should continue
to read across the row or down a column. The most important feature of simple random sampling
is that it facilitates representation of the population by the sample ensuring that the statistical
conclusions are valid.
Systematic Sampling
This is also another widely used type of sampling technique. This is used because of its ease and
convenience. As in the case of simple random sampling, it is conducted choosing a random
starting point and then picking every element in succession from the sampling frame. The sample
interval, i, is determined by dividing the population size N by the sample size n and rounding to
the nearest integer.
Consider a situation where the research intends to choose 10 elements from a population of
100. In order to choose these 10 elements, number the elements from one to 100. Within 20
population elements and a sample of size 10, the number is 10/100 = 1/10, meaning that one element
in 10 will be selected. The sample interval will, therefore, be 10. This means that after a random start
th
from any point in the random table, the researcher has to choose every 10 element.
Systematic sampling is almost similar to simple random sampling in that each population element
has a known and equal probability of selection. However, the difference lies in that simple random
sampling allows only the permissible samples of size n drawn have a known and equal probability
of selection. The remaining samples of size n have a zero probability of being selected
Stratified sampling
Stratified sampling is a two-way process. It is distinguished from the simple random sampling and
systematic sampling, in that:
a) It requires division of the parent population into mutually exclusively and exhaustive
subsets;
b) A simple random sample of elements is chosen independently from each group or subset.
Therefore, it characterises that, every population element should be assigned to one and only
stratum and no population elements should be omitted. Next, elements are selected from each
stratum by simple random sampling technique. Stratified sampling differs from quota sampling in
that the sample elements are selected probabilistically rather than based on convenience or on
judgemental basis.
Strata are created by a divider called the stratification variable. This variable divides the
population into strata based on homogeneity, heterogeneity, relatedness or cost. Sometimes, more
than one variable is used for stratification purpose. This type of sampling is done in order to get
homogenous elements within each stratum and, the elements between each strata should have a
higher degree of heterogeneity. The number of strata to be formed for the research is left to the
discretion of the researcher, though; researchers agree that the optimum number of strata may be
6.
c) it combines the use of simple random sampling with potential gains in precision;
d) estimates of the population parameters may be wanted for each sub-population and;
Non-probability sampling does not involve random selection. It involves personal judgement of
the researcher rather than chance to select sample elements. Sometimes this judgement is imposed
by the researcher, while in other cases the selection of population elements to be includes is left to
the individual field workers. The decision maker may also contribute to including a particular
individual in the sampling frame. Evidently, n n probability sampling does not include elements
selected probabilistically and hence, leaves an degree of „sampling error‟ associated with the
sample.
Sampling error is the degree to which a sam le might differ from the population. Therefore, while
inferring to the population, results could ot be reported plus or minus the sampling error. In non-
probability sampling, the d g to which the sample differs from the population remains unknown
However, we cannot come to a conclusion that sampling error is an inherent of non probability
sample.
Non-probability samples also yield good estimates of the population characteristics. Since,
inclusion of the elements in the sample are not determined in a probabilistic way, the estimates
obtained are not statistically projectable to the population.
The most commonly used non-probability sampling methods are convenience sampling, judgment
sampling, quota sampling, and snowball sampling.
Convenience Sampling
Convenience samples are sometimes called accidental samples because the elements included in
the sample enter by „accident‟. It is a sampling technique where samples are obtained from
convenient elements. This refers to happening of the element at the right place at the right time,
that is, where and when the information for the study is being collected. The selection of the
respondents is left to the discretion of the interviewer. The popular examples of convenience
sampling include (a) respondents who gather in a church (b) students in a class room (c) mall
intercept interviews
without qualifying the respondents for the study (d) tear-out questi nnaire included in magazines
and (e) people on the street. In the above examples, the peop e may not be qualified respondents,
however, form part of the sample by virtue of ssembling in the place where the researcher is
conveniently placed.
Convenience sampling is the least expens ve a d least time consuming of all sampling
techniques. The disadvantage with conv ni nce sampling is that the researcher would have no
way of knowing if the sample chosen is representative of the target population.
The distinguishing feature of judgment sampling is that the population elements are purposively
selected. Again, the selection is not based on that they are representative, but rather because they
can offer the contributions sought. In judgement sampling, the researcher may be well aware of
the characteristics of the prospective respondents, in order that, he includes the individual in the
sample. It may be possible that the researcher has ideas and insights about the respondent’s
requisite experience and knowledge to offer some perspective on the research question.
Quota Sampling
Quota sampling is another non-probability sampling. It attempts to ensure that the sample chosen
by the researcher is a representative by selecting elements in such a way that the proportion of the
sample elements possessing a certain characteristic is approximately the same as the proportion of
the elements with the characteristic in the population.
Quota sampling is viewed as two-staged restricted judgemental sampling technique. The first stage
consists of developing control categories, or quotas, of population elements. Control
characteristics involve age, sex, and race deified on the basis of judgement. Then the distribution
of these characteristics in the target population is determined. For example, the researcher may use
control categories in that, he/she intends to study 40% of men and 60% of women in a population.
Sex is the control group and the percentages fixed are the quotas.
In the second stage, sample elements are selected based on convenience or judgement. Once the
quotas have been determined, there is considerable freedom to select the elements to be included
in the sample. For example, the researcher may not choose more than 40% of men and 60% of
women in the study. Even if the researcher comes across qualified men after reaching the 40%
mark, he/she would still restrict entry of men into the sample and keep searching for women till
the quota is fulfilled.
Snowball Sampling
This is another popular non-probability technique widely used, especially in academic research. In
this technique, an initial group of respondents is selected, usually at random. After being
interviewed, these respondents are asked to identify others who belong to the target population of
interest. Subsequent respondents are selected based on the information provided by the selected
group members. The group members may provide information based on their understanding about
the qualification of the other prospective respondents. This method involves probability and non-
probability methods. The initial respondents are chosen by a random method and the subsequent
respondents are chosen by non-probability methods.
DATA ANALYSIS AND INTERPRETATION
The most commonly used tools are 'Mean, Median, Mode; Ge metric Mean, Measures of
Dispersion such as Range; Mean Deviation, Standard Deviation and also other measures such as
Coefficient of Correlation, Index Numbers etc. It is necess ry to note that technical interpretation
of data has to be combined with a high degree of sound judgement, statistical experience, skill and
accuracy. After all figures do not l e, they are innocent. But figures obtained haphazardly,
compiled unscientifically and analyz d incompetently would lead to general distrust in statistics by
the citizens. It should be understood that "statistical methods are the most dangerous tools in the
hands of an expert".
In spite of a careful collection by a researcher, there may be a possibility for errors of omission
and commission arising and it is for this purpose that the process of editing becomes necessary.
The editor, while examining certain responses of the respondents, may find some mistakes in the
form of incomplete, vague or irrelevant answers. Such inconsistent answers have to be eliminated
or suitably and reasonably modified. Further, there should be no room for fictitious data to creep
in. Hence the editor has to take care to see that the authenticity of the data is in a perfect shape.
For the purpose of classification of the data into meaningful and useful classes, the procedure of
coding has to be used. This procedure would be advantageous in dealing with the data having a
number of characteristics. Also, a large volume of data can be processed accurately.
Manual processing and analysis can be carried out by using measures of central tendency,
dispersion, correlation regression and other statistical methods if the volume of data is not very
large.
In case a researcher is confronted with a very large volume of data then it is imperative to use
'computer processing'. For this purpose n c ssary statistical packages such as SPSS etc. may be
used. Computer technology can p ove to be a boon because a huge volume of complex data can be
processed speedily with greater accuracy.
Descriptive data provides quantitative information for analytical interpretation for instance: with
respect to the wage distribution of 500 workers in a factory, we can calculate various measures of
central tendency, dispersion, skewness etc. Inferential data relates to statistical inference where
conclusions are drawn on the basis of samples taken randomly from a population, which is
assumed to be normal. Population parameters are estimated on the basis on the basis of sample
statistics.
Depending upon the nature of researcher's problem, relevant sampling methods are used for
obtaining data. However, for the purpose of hypothesis testing, parametric or non-parametric
tests may be used depending upon the fact whether the assumptions in regard to population are
based on 'distribution' or 'distribution-free characteristics'.
Classification is the process of sorting 'similar' things from among a group of objects with
different characteristics. In other words, heterogeneous data is divided into separate homogeneous
classes according to characteristics that exist amongst different individuals or quantities
constituting the data. Thus, fundamentally classification is dependent upon similarities and
resemblances among the items in the data.
The main object of classification is to present vividly, in a simplified and quickly intelligible
form, a mass of complex data. Without condensing detai s in a classified form it is difficult to
compare quickly, interpret thoroughly and analyse properly different sets of quantitative and
qualitative phenomena. The basic requireme ts of good classification are stability, non-
ambiguity, flexibility and comparability.
Depending on the characteristics of the data, they can be broadly categorized into two separate and
distinct groups - descriptive and numerical. Descriptive characteristics are those that can be
described in words and are expressible in qualitative terms. Numerical characteristics are
quantitative in nature. For instance, literacy, sex, caste and religion are descriptive characteristics.
Height, weight, age, income and expenditure are numerically expressible characteristics.
Descriptive or qualitative classification is termed classification according to attributes. Numerical
or quantitative classification of data in certain class intervals is termed as classification in terms of
classes with certain intervals, or classification according to class intervals.
On the other hand, analysing and classifying collected data under several attributes in different
classes is called manifold classification. For example, if each of the two classes, literate and
illiterate, is divided into males and females, then there would be four classes. If classified further
on a regional basis, there would be a number of other classes. Such a process of classification of
data into a number of classes and class s within classes is known as manifold classification.
Phenomena like income, heights and weights are all quantitatively measurable and data on them
can be classified into separate class intervals of uniform length. For instance, the marks obtained
by a group of 50 candidates in a subject at an examination can be classified into the following
classes: 0-10, 10-20, 20-30, 30-40, 40-50, 50-60, 60-70 etc. Each class has a lower and an upper
limit and the number of candidates getting marks between these two limits of the same class
interval is called the frequency of the respective class. To give an example, if 12 candidates get
between 40 and 50 marks, 12 is the frequency of the class 40-50.
Number of Classes
The number of classes into which particular data should be classified depends upon the mass of
data. The larger the mass, the more should be the number of classes. Usually data is classified into
not less than six classes and not more than 20 classes, depending upon the mass and size of the
data and the length of the class intervals. The fundamental object of classifying data is to get the
maximum possible amount of information most precisely. According to Sturges' Rule, the number
of class intervals (n) = 1 + 3.322 log N, where N = total number of observations.
Class Limits
The choice of class limits is determined by the mid-value of a class interval, which should as far as
possible be identical ith the arithmetic average of the items occurring in that class interval.
Tabulation
Tabulation is the process of arranging given quantitative data based on similarities and common
characteristics in certain rows and columns so as to present the data vividly for quick
intelligibility, easy comparability and visual appeal.
A statistical table comprises a title, a head-note, a stub head and stub details, captions and columns
under the captions, field of the table under different column heads, footnotes and, source notes.
Here's a sample:
1. General Purpose Table: This is primarily meant to present the entire original data on a
subject. Such presentation of numerical data in a tabular form is especially useful as a source of
information and reference for constructing d fferent special
purpose tables.
2. Special Purpose Table: As its name implies, this is a statistical table that specially presents
and emphasizes certain phases or significant aspects of the information given in a general purpose
table. Presenting data in a special table not only makes it easy to understand specific data, it also
facilitates easy comparison and clear-cut interpretation.
Types of Tabulation
1. One-way Table (single tabulation): A one-way table gives answers to questions about one
characteristic of the data.
2. Two-way Table (double tabulation): A two-way table gives information about two
interrelated characteristics of a particular type of data.
4. Higher Order Table (Manifold Tabulation): This table gives information under several
main heads and subheads on questions relating to a number of interrelated characteristics.
1. Every statistical table should be given an appropriate title to indicate the nature of the data.
The title should be simple, intelligible and unambig o s and should not be too lengthy or too short.
4. The stub heads and the main heads should be consistent with the nature of the data and be
very clear.
5. The main headings under the caption should be as few as possible in keeping with the
requirements of space and type of data. If the main headings are few, comparison between
different sets of data becomes easy.
6. The entire arrangement of data should be appropriate, compact and self-explanatory so that
it is not necessary to rearrange the data in any manner.
7. Comparisons between different figures such as totals and averages-are easier if they are
arranged vertically and not horizontally.
8. In order to show important parts of the data (under main heads) distinctly, it is necessary to
draw thick double or multiple ruled lines.
9. Depending upon the nature of the data, items in the stub column may be arranged
according to:
10. Figures in the data that are estimates, approximate or revised should be indicated by an
alphabet, asterisk, number or any other symbol. An explanation should be given in the footnote.
11. The different units used in the data should be indicated in the column heads. For example:
'figures in rupees', 'figures in metres', and so on.
12. The source of the data should be indicated under the footnote. It is necessary to mention
the source for further references and other details and also for indicating the reliability of the data.
UNIT --IV
UNIT --V