Research Document IV [U]

MSU 07401 [Research Methodology]

Research Design
2022

Dr. E. P.Mushi
 What is research design?

 A research design is the ‘procedures

for collecting, analyzing,
interpreting and reporting data in
research studies’ (Creswell & Plano
Clark 2007, p.58).

 Research design sets the procedure

on the required data, the methods to
be applied to collect and analyze this
data, and how all of this is going to
 Three research approaches
 Before advancing to explore
concepts, steps and methods in
Research Design, let’s explain the
concept of “Research Approaches”.

 There are 3 research approaches: (a)

qualitative, (b) quantitative, and (c)
mixed method.

 But the three approaches are not as

 Basic Concepts in Research
Design
 Research: Is a process of
developing new knowledge by using
systematic and reliable scientific
methods, through analysis and
synthesis of data; to find truth from
facts.
 In reference to an area of interest,
three basic questions can cover
research process:
 1. What do we want to know?
 2. By what methods we can get the
 Basic Concepts
 Research Design: is the
specification of the research
methodology suitable to obtain the
valid knowledge of the research
objective.
 Research Design can be under any of
three approaches 1. Qualitative Design
2. Quantitative 3. Mixed
 We will discuss the three approaches

and identify strengths of each.

 Research methodology specifies
the tools and techniques in terms of
 Basic Concepts cont.
 Variable: A variable refers to
specific attribute, characteristic, a
thing/matter/substance that a
researcher wants to study.
 Independent variable: occasionally
called ‘input’, ‘manipulated’, cause’,
or influence variable.
 Dependent variable: is the result
of the effect of the independent
variable, occasionally called ‘output’,
‘outcome’,
 Basic Concepts cont.
 Extraneous variables:
 Intervening variable is the extraneous
variable that connects or mediate the
independent variable to the dependent
variable. In research design most
researchers take the intervening
variables as part of the independent
variables.
 Control variable is the extraneous

variables that need to be controlled or

held constant their effect are cancelled
out. These are factors such as age, sex,
IQ, social economic status and
 Basic concepts, cont.
 Data are the value of a variable, obtained
from the measurement of a variable
using an instrument. There are two types
of data –
 1. The qualitative description of a
variable (qualitative data) – Nominal
scale of measurement.
 2. Quantitative value of a variable
(quantitative data)
 2.1 Discreet or Categorical data – by
assigning numerical value to qualitative and
nominal data.
 Basic Concepts, cont.
 Scales of measurement:
 Nominal Scales – description of
characteristics, labels, numerical labels
eg. male = 1, female = 2
 Ordinal Scale – ranks the observation,

from best to worst, first to last

 Interval scale/rating scale – eg likert

scale, equal interval.

 Ratio scale – True Zero Scales,

absolute zero, equal distance among

units.
 Question:

 Write short notes on the four

 Steps in Research Design
 There are many forms of design.
Some focus on the process of data
collection only, while others extend
their boundaries to cover data
analysis (Sarantakos, 2013: 121).

 Most writers see RD as covering all

aspects of research from the selection
of the topic to the publication of the
data (see Flick, 2000b).

 The most common steps in research

The process of Research
1. Select Research Topic and an
Design
approach to use: Identify Variables
From variable(s)

2. Methodological Construction of a topic:

Operationally define the variables, questions,
objectives and hypothesis.
3. Sampling: Design the sources of data. From
where? (documents), from whom? (participants) -
Population and Sample
4.Choose the method of data collection:
Instrument, supervising data collection,
actual data collection. .
5. Data Processing: Design the data analysis methods
that will give the knowledge for answering the
 From Research Questions to
Research Objectives
 Types of Research Questions
 1.Descriptive Questions: Want to

quantitatively or qualitatively describe a

variable.
 E.g. i. What are the characteristics of an

effective management?
 ii. What is the level of employees

satisfaction?
 iii. To what extent do students’ possess

critical thinking?
 2.Causal-comparative questions: want to

know why?, what contributes to the key

variable. In the analysis there is comparing.
 Types of research questions, cont
 3. Relationship questions: what to

know what contributes to the value of a

variable, to know why the value of a
variable is high or low.
 Examples:

 i. What is the degree of association

between employee trust and
performance?
 ii. What is the extent of relationship of

organizational decision making,

administrative style, and employee
performance?
 Research questions,
cont.
 Cause and effect questions: want to
see the effect of innovation on the
desired result, want to see if a new
practice will bring desired results, or
want to see degree of contribution of
thing on another.

 E.g. i. What is the effect of

organizational leadership on
employees performance?
 ii. To what extent does the student

personality type affect the student

 Transform research questions to
become research objectives
Research Research
Questions Objectives
What are the To identify the
characteristics of an characteristics of an
effective management? effective management.
To what extent do To determine the extent
students’ possess critical of students’ possession
thinking? of critical thinking.

What is the difference of To determine the

managerial approaches difference of managerial
among four mining approaches among four
companies in Geita mining companies in
Mwanza? Geita Mwanza.
 Identify variables from Research
Research ObjectiveObjectives
Variable
To identify the 1. Characteristics of
characteristics of an effective management
effective management.

To determine the extent 2. Critical thinking skills

of students’ possession
of critical thinking.

To determine the 1.Managerial

difference of managerial approaches (dep. V)
approaches among four 2.Four mining
mining companies in companies (indep. V)
Geita.
To determine the effect 1.Organizational
of organizational leadership
leadership on employees 2.Employee
 From research objectives to
Research Objective hypothesis
Research Hypothesis
To identify the No hypothesis
characteristics of an
effective
management.
To determine the Not required
extent of students’
possession of critical
thinking.
To determine the There a difference of
difference of managerial
managerial approaches between
approaches among the four mining
four mining companies in Geita.
 Research hypothesis
 It is the expected answer to the
research.
 A research objective requires a
hypothesis if it is inferring the
observed value from the sample to the
expected value or difference or
relationship in the population.
 Hypothesis statement should be
congruent with the objective.
 Research hypothesis should be
transformed into statistical
hypothesis for the purpose of testing.
 Two types of statistical hypothesis:
 H0 - Null hypothesis
 Null vs Alternative
hypothesis
 Research Hypothesis – is the
expected result. It should be
transformed into statistical
hypothesis in form of a Null (Ho) or
Alternative Hypothesis (H1).
Statistical hypothesis can be tested.

 Null hypothesis (Ho) is the

statement of “no difference” or “no
relationship” or “no effect”.


 Transform H1 to H0
Alternative Null Hypothesis
Hypothesis
There is impact of
moral value gained
from student
camping activities.
There is difference of
student reading skills
between two learning
approaches: the
uniform reading
materials and varies
learning materials.
 Description of Population and Sample
Population
 The target population is quantitatively and

qualitatively defined.
Sample
 Clearly describe the sampling method leads to

the unbiased sample strongly represents the

target population.

 The sample size meets recommended guidelines

for minimum sample size, e.g. refer to Krejcie &
Morgan (1970.), or based on degree of standard
errors (α level).
 Table of population and sample should be
provided.

 In case of small population and all subjects are

 Population
 The research title and the scope of research will
indicate the type of population.

 A “population” consists of all the people (called

subjects) the researcher wants to study. The
entire group of people to which the researcher
wishes to generalize the study findings.

 Population also refers to the target value of

statistics, e.g. Population Mean of Satisfaction of
Teachers in International Schools in Tanzania.

 E.g.
Population and Sample
 Generally it is impossible (from both a
strategic and a resource perspective) to
study all the members of a population
for a research project. It just costs too
much and takes too much time. Instead,
a selected few participants (who make
up the sample) are chosen with the
confidence that the sample is the
representative of the population.
 Sampling is the process of selecting a
group of subjects for a study in such a
way that the individuals represent the
larger group from which they were
 Population and Sample
Population with true population value called “Parameter”

Sampling
(Selection of subjects from
the population)

Sample
Compute the sample
value called “Statistic”

Generalize/Infer the
“Statistic” to the
“Parameter”, may end
up with the gap
between “True Value”
and “Sample Value”.
This gap is called
“Sampling Error”
 Desirable Sample
 A random sample is required for
Inferential Statistics
 Not biased.
 Optimal size: preferred big sample; at
least 15 subjects for experiment, at
least 30 for causal-comparative and
correlation.
 Use Krejcie and Morgan table or
Yamane’s formula to estimate the
sample size for survey research.
 SAMPLING METHODS AND
TECHNIQUES
 Sampling methods and
techniques
We use sampling technique to fulfill the population
representation so as to ensure the external
validity ability to generalize the findings from
the sample to the population. The acceptable
techniques are:
Probability Sampling Non Probability
Sampling
 Simple Random  Judgmental or
Sampling, Purposive
 Stratified Random Sampling,
 Systematic
 Convenience
Sampling
 Cluster Sampling
sampling,
 PROBABILISTIC
SAMPLING
Any sampling method where some
elements of population have equal
chance of selection.
The findings can be generalized to the
population of study.
 1. Simple Random Sampling
Technique
 Individual subjects are chosen with
independent and equal chance.
 Prefer in homogeneous population, equal
characteristics of participants is assumed.
 The random methods: Lottery, Random
Number, Computerized.
 In experimentation, 1st step is randomly
select subjects for the study, 2nd step
randomly select each of them into the
experimental group and the control group.
 May not represent the proportional
characteristics of population.
 2. Stratified Random Sampling
Technique
 Divide the population into Strata,
proportionally random the subjects from
each Stratum

A A
B B
D D
C C
 Large Population with known
number of people
Title: Study of best practices for effective quality management in public schools
in Kinondoni District, Dar es Salaam, Tanzania, as perceived by teachers.
Population: population of this study is the teachers in seven types schools in
Kinondoni District, Dar es salaam, Tanzania.
 Public Schools with Kiswahili Language System,
 Public Schools with Bilingual System,
 Public Schools with English Program System,
 Private Schools with Swahili System.
 Private Schools with Bilingual System.
 Private Schools with English Program System
 International schools.
Sample: Stratified Random Sampling Technique will be used for
obtaining teachers for this study. Based on Krejcie & Morgan
(1970) the optimal number of 370 teachers are required from total
10,060 teachers . The stratified sampling ratio, 370 ÷10060
= .036 or .04 will be used to compute the sample size for each
type of school.
School Type Population of Sample of Teachers
Teachers
Public School with 4,000
Kiswahili Language
Public Schools with 960
Bilingual System
Public Schools with 850
English Program
System
Private Schools with 2,400
Swahili System.
Private Schools with 1,600
Bilingual System
Private Schools with 1,200
English Program
System
International schools 650
1. Krejcie & Morgan Table: Required Sample size – 370.
TOTAL 10, 060
2. Stratified Sampling Ratio – 370/10060 = .036.
3. Multiply each stratum by .036 to obtain the sample size
for each stratum.
4. The sample size of 420 is equivalent to 370.
 3. Systematic Sampling
Technique
 A complete list of subjects in a
homogeneous population
 Decide the desired proportion, e.g. 5% or
1 from 20
 Randomly select the first subject in the
list
 Consequently, use the 20th subject for the
sample until the sample size is complete.
 4. Cluster Sampling
Technique
 When the selection of individual
members of the population is
impractical or too expensive, it may
be possible to select groups or
clusters of members for the sample.
 Cluster sampling is useful in
situations where the population
members are naturally grouped in
units that can be used conveniently
as clusters. Then the researcher
selects groups or clusters of
 5. Multi-Stage Sampling
Technique
 The combination of Stratified
Random Sampling, Cluster
Sampling, and Simple Random
Sampling Techniques
 To be applied to the heterogeneous
group, to ensure the representation
of all distinctive characteristics
 The Survey of Teachers’ Attitudes in
Primary Schools and Secondary
Schools in Tanzania
 Stage I: Apply cluster sampling technique, randomly
select two Zones: North, East, West, South.
 Stage II: Based on the selected zones consider the
list of schools classified by school type (public
primary, public secondary, private primary, and
private secondary), apply stratified random
sampling technique proportionally select the
schools for the total, say, 30 schools.
 Stage III: Within each selected school, apply
stratified random sampling technique identify the
required number of teachers in proportion to the
subject teachers in school for the total number of
required teachers, say 25 teachers.
 Stage IV: Apply simple random sampling for
drawing individual teachers in each subject toward
the required number.
 6. Multi-phase sampling
 Part of the information collected from
whole sample & part from subsample.

 In Tb survey MT in all cases – Phase I

 X –Ray chest in MT +ve cases – Phase
II
 Sputum examination in X – Ray +ve
cases - Phase III

 Survey by such procedure is less

 NON PROBABILITY
SAMPLING
Any sampling method where some
elements of population have no
chance of selection (these are
sometimes referred to as 'out of
coverage'/'undercovered'), or where
the probability of selection can't be
accurately determined. It involves the
selection of elements based on
assumptions regarding the population
of interest, which forms the criteria
for selection. Hence, because the
 1. Purposive Sampling
Technique
 Also known as judgmental
sampling.
 It is a nonprobability sample
 It is the sample of convenience.
 Mostly used in situation where the
administrative limitation is prevalent.
 The degree of generalization is limited
only to similar situation under study.
 Judgmental or Purposive
 The sampling
selection cont.
of a purposive sample is often
accomplished by applying expert knowledge of
the population to select in a nonrandom
manner a sample of elements that represents a
cross-section of the population. For example,
one might select a sample of small businesses
in the United Republic of Tanzania that
represent a cross-section of small businesses in
the nation. With expert knowledge of the
population, one would first decide which
characteristics are important to be represented
in the sample.
 Once this is established, a sample of
businesses is identified that meets the various
 2. Snowball sampling
 Selection of individuals who are
recommended by others already
selected.
 Snowball sampling is a special
nonprobability method for developing
a research sample where existing
study subjects recruit future subjects
from among their acquaintances.
 This sampling technique is often used
in hidden populations which are
difficult for researchers to access
(such as drug users or commercial sex
workers), or in cases where a sampling
frame is hard to establish,
 It is assumed that cases are affiliated
 Snowball sampling…….

Cont.
This technique involves beginning with a
small group of people to interview (a small
ball of snow) and then increasing the size of
the group of people to interview by asking
people who have been interviewed to suggest
other people to be interviewed (rolling a snow
ball along the ground makes it pick up more
snow and get bigger).

 For example, At the beginning, in each

monitoring site, a small group of people (5 or
6), who have different types of impairments,
are interviewed. At the end of each interview,
Monitors ask the Interviewee if he or she can
suggest someone else to be interviewed. The
Monitors record this person’s name,
background and contact information on the
Snowball sampling……. Cont.
 3. Quota sampling
 The population is first segmented into
mutually exclusive sub-groups, just as
in stratified sampling.
 Then judgment used to select subjects
or units from each segment based on a
specified proportion.
 For example, an interviewer may be
told to sample 200 females and 300
males between the age of 45 and 60.
 It is this second step which makes the
technique one of non-probability
sampling.
 In quota sampling the selection of the
4. Accidental/convenience
sampling

 Use results that are easy to get

 Accidental/convenience
sampling….. Cont.
 Examples of convenience samples
include mall intercept interviewing,
unsystematically recruiting
individuals to participate in the study
(e.g., what is done for many
psychology studies that use readily
available undergraduates), visiting a
sample of business establishments
that are close to the data collection
organization, seeking the
participation of individuals visiting a
Web site to participate in a survey,
 Accidental/convenience
sampling….. Cont.
 For example, in the case of surveying those
attending the Super Bowl using a
convenience sample, a researcher may
want data collected quickly, using a low-
cost method that does not involve scientific
sampling.

 The researcher sends out several data

collection staff members to interview
people at the stadium on the day of the
game. The interviewers may, for example,
carry clipboards with a questionnaire they
 How to Deal with Small
Population
Generally qualitative research and case study research
has to do with all people in one organization, e.g. in
one or more schools, in one or more classrooms.
Certain quantitative educational research also applies
this framework. The researcher will describe the
population and sample in the way that all people (all
subjects) in the population will be used as the sample
of the study.
An Example:
Title: A relationship between school climate and
students’ satisfaction at International School of
Tanganyika in Dar es salaam
Scope: In this study only A-level students in IST will be
used for the study.
Population: population of this study is all A-level students
Class 1-3 at IST in the year 2020.
 Group Practice 1

Task: From the example above:

 Formulate research questions.
 Transform the questions into research
objectives.
 Identify variables in each objective.
 Provide the operational definition of
the variables,
 Specify the sample or source of data,
and
 Design the research instrument for