Research Design

Research Design
Research design
 The roadmap that guides the research
 It is also the basis upon which a valid conclusion on the
of the research will emanate.
 The choice of the research design is influenced by:
– Type and nature of study (quantitative/qualitative)
– Type of data required

• Objective measurements
• Subjective measurements
• Variability in measurement values
– The type of analyses to be performed

Types of research designs
 Falls under two designs:
– Non Experimental design/Correlational designs
• Descriptive research designs or
– Experimental research designed

• True Experimental studies
• Quasi experimental studies
Descriptive designs
 Design of choice in behavioural sciences mainly
because phenomenon under study are not usually
reproducible in controlled settings.
 They examine the following types of relationships:

– correlations among variables
– differences among groups
 The focus of the design is on the present and

therefore collect current status data from a
population about specific or a wide range of issues.
 Descriptive studies are also called observational

studies.
Descriptive designs
 They do not establish causality between variables
(dependent and independent variables).
– Causality is a cause and effect relationship between dependent

and independent variables
 Generally higher external validity due to large samples but

lower internal validity because of inability of researchers
to control confounding variables.
 They are also more prone to design errors due to:

– Poorly designed measurement instruments
– Error emanating from study subjects (e.g., response error)
– Contextual factors that reduce a sound/accurate measurement
Common Descriptive Research designs
 Census survey
 Sample surveys
– Cross-sectional surveys
– Longitudinal surveys
 Case Study
 Observational (a.k.a. qualitative)
 Action research
Census
A survey which obtains responses from the
entire population.
– For example: the Uganda 2014 National and
Household census
Cross-sectional design
 Cross-sectional study design: This involves the study of a sample of the
population of interest at a particular point in time.
 It provides a single point analyses of the characteristics of the

population at that time. For example the 2016 UGHS.
– For example the contraceptive prevalence rate
– HIV/AIDS prevalence in the population
– Prevalence of SGBV
– Maternal Mortality Ratio
– Etc
 This is the most common types of research design.

– it is generally cheaper
– It takes a short time
– It is less complex and easy to conduct
 It is used to provide a snapshot about development indicators.

Longitudinal research design
 This is a series of studies or measurements conducted about the
population more than once usually at fixed intervals.
 It is intended to assess changes in the population of interest over

time.
 Longitudinal designs are of three types:
– Repeated cross-sectional designs (Data are collected from different

samples within the same population ).
– Fixed sample panel designs (Data are collected from the same
individuals -samples).
• At least 3 stages using a fixed time interval (suffers from attrition, sample
fatigue, expensive).
– Event based designs (Data is collected from different samples on

the same event within a cohort (also known as “cohort studies”).
• A cohort are individuals or groups with a common event experience.
Experimental designs
 Mostly conducted in the physical sciences and life
sciences.
 In the human sciences, it is conducted in medical
research where it is also called Randomized Controlled
Trials (RCR).
 The primary purpose is to investigate cause-and-effect
relationships among variables
– Experimental groups vs. control groups
– Each group of participants receives a different treatment
– Always involves manipulation of the independent variable
 Answers the question “What will be?”
 Conducted in controlled conditions/environments

(laboratories).
Experimental designs
 Causality is the situation in which an action brings about a
consequence.
― It is a relationship in which a change should be expected in a
particular behaviour when an intervention is introduced.
― The most powerful design for testing causal hypotheses
because experiments establish association, time order, and
non-spuriousness.
― Experimental research is most appropriate for answering
research questions about the effect of a treatment or some
other variable whose values can be manipulated by the
researcher.
― Greater confidence of validity of causal conclusions than in
other design.
 Experimental research follow a definite, orderly procedure more than

any other type of research.
 Are generally more reliable and with fewer errors.

Main features of experimental research
 Manipulation
– This is an intervention which requires the deliberate
introduction of the experimental agent or independent
variable/s.
– This is administered to the experimental group while

withholding it from the control group.
 Control
– This is the comparison group which is not exposed to the
independent variable/s.
– This group acts as the bases for assessing of the effect of

the independent variable/s on the experimental group.
 Randomization
– This is a procedure which provides every participant
equal chance of belonging to either the experimental or
the comparison/ control group.
Steps to follow in experimental research
 Usual steps in experimental research
– State the research problem
– Determine if experimental methods apply
– Specify the independent variable(s)
– Specify the dependent variable(s)
– State the tentative hypotheses
– Determine measures to be used
– Pause to consider potential success
– Identify intervening (extraneous) variables
– Formal statement of research hypotheses
– Design the experiment
– Final estimate of potential success
– Conduct the study as planned
– Analyze the collected data
– Prepare a research report
Discussions

Research Design

Uploaded by

Copyright:

Available Formats

Research Design

Uploaded by

Document Information

Original Description:

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

Research Design

Uploaded by

Copyright:

Available Formats

Research Design

– Type of data required

– The type of analyses to be performed

– Experimental research designed

 They examine the following types of relationships:

 The focus of the design is on the present and

 Descriptive studies are also called observational

– Causality is a cause and effect relationship between dependent

 Generally higher external validity due to large samples but

 They are also more prone to design errors due to:

 Observational (a.k.a. qualitative)

 It provides a single point analyses of the characteristics of the

 This is the most common types of research design.

– It takes a short time

– It is less complex and easy to conduct

 It is used to provide a snapshot about development indicators.

 It is intended to assess changes in the population of interest over

 Longitudinal designs are of three types:

– Repeated cross-sectional designs (Data are collected from different

– Event based designs (Data is collected from different samples on

 Conducted in controlled conditions/environments

 Experimental research follow a definite, orderly procedure more than

 Are generally more reliable and with fewer errors.

– This is administered to the experimental group while

– This group acts as the bases for assessing of the effect of

You might also like