Vaal University of Technology

WRITING MY FIRST RESEARCH PROPOSAL

Faculty of Management Sciences

Prof M Dhurup
2019
 For B.Tech, Advanced, Post Graduate Diplomas, Masters
and Doctoral students

FACULTY OF MANAGEMENT SCIENCES

Purpose

This guide is intended for undergraduate/and post graduate students who wish to pursue
an enlightened path to undertake research within the broad sphere of business
management, marketing, human resources, business, accounting, sport and related
disciplines. Consequently, this guide, whilst simplistic in its nature sets in motion a generic
stance to envelope a basic procedure in enhancing students “beginner understanding” for
proposal writing. The central purpose of this guide is aligned to assist students to develop
academic research proposal writing on quantitative research and qualitative research.

I sincerely hope that this guide will assist students to refine their proposal writing ability
and prepare a research proposal according to the Faculty and University guidelines.
Feedback and suggestions in improving this guide is welcomed.

Prof. M. Dhurup
Executive Dean: Faculty of Management Sciences: Vaal University of Technology
Private Bag X021,
Vanderbijlpark
Email: roy@vut.ac.za
Tel: +27 016 950 6886
 WORDS OF WELCOME
Welcome as a research student in the Faculty of Management Sciences at the Vaal University of
Technology. We hope that the time we spend together while learning to understand aspects of
your chosen field of study will be fruitful, rewarding but above all ENJOYABLE.

In order to assist you as a student to prepare a research proposal, we have put together a generic
(common) and simple guide to help you along in your chosen field of study so that your proposals
are of standard for a University of Technology. In this way proposals in our faculty will be
conform to the general guidelines set out by the University.

Students must take responsibility for their own progress. Since research proposals require that
the Internet should be used, you should ensure in time that you gain access to a computer with
Internet facilities in the library. Should you choose to use an Internet café; the costs incurred will
be for your own account. Students are requested to use bibliographical and text references
according to the guidelines laid down by the Institution. A copy of the guidelines is available in
the library and at the local bookshop.

PLEASE NOTE: The Faculty of Management Sciences and the University views PLAGIARISM (i.e.
copying, breach of copyright) in a very serious light. Severe penalties (for example, expulsion from
the University) may be imposed on you in accordance with the University’s assessment regulations
and disciplinary procedures. It is in your best interest to refrain (keep away) from such
malpractices. Please ask your lecturer to elaborate on this issue during lectures.
 CHAPTER 1
Introduction
Learning Outcome 1

1.1 What is research?

Research is a systematic process that entails obtaining scientific information and
knowledge by means of various objective methods and procedures. The term objective
means that that the methods and procedures the researcher uses must be free from
personal feelings or opinions and specific methods must be used in the various stages of
the research process.

1.2. What is research methodology?

Research methodology considers and explains the logic behind research methods,
techniques and procedures used in a study. It has a much wider scope than research
methods.

1.3. What is a research design?

A research design is the set of methods and procedures used in collecting and analyzing
measures of the variables specified in the research problem. The design of a study defines
the study type (exploratory, descriptive, or causal). A research design is a framework that
is created to find answers to research questions. The function of a research design is to
ensure that the evidence obtained enables you to effectively address the research problem
logically and as unambiguously as possible. In social sciences research, obtaining
information relevant to the research problem generally entails specifying the type of
evidence needed to test a theory, to evaluate a program, or to accurately describe and
assess meaning related to an observable phenomenon. Gupta (2011:39) define a research
design as a blueprint that specifies the methods and procedures employed for data
collection and analysis in a study.

1.4 Types of research designs

1.4.1 Exploratory design are often used to identify the real nature of the research
problem/s and perhaps formulating hypothesis for later tests. These designs may make use
of primary data or secondary data (published data). This method is used when the
researcher wishes to identify the problem more closely which becomes a research filter for
more extensive and intensive research activities which could be done later. An exploratory
research design examines comparatively unknown conditions with the aim of gaining new
ideas and to generate understandings about situations that can at a latter stage suggests
the hypotheses or instigate further future research. Salkind (2012:1254) highlights that
generally, exploratory research is more valuable when less about a phenomenon is known
and the investigation is conducted when the researcher is not clear about the problems
that may be encountered.
1.4.2 Descriptive research designs are built on the supposition that the prevailing
relationships amongst the study variables are already known (Feinberg, Kinnear & Taylor
2013:578). Therefore, descriptive research is undertaken on the basis of a well-known
prevailing nature of a research problem, even though this problem necessitate thorough
motivation. A descriptive research study could be longitudinal or cross-sectional (Iacobucci
& Churchill 2010:86). Two commonly utilised research design approaches are briefly
explained.
Longitudinal studies comprise data collection and analysis of the same sample elements
(sample survey) over a period (Babbie & Mouton 2010:92-93). In other words, the process
of measuring the sample with a longitudinal approach is that the procedure is repeated at
various points in time, where the sample elements remains the same throughout (Malhotra
2010:106).
Cross-sectional studies include the gathering of data from the same sample elements only
once (Burns & Bush 2014:103). This research pattern is found in two different designs
namely, multiple cross-sectional design and single cross-sectional design. The data
collection process is not repeated like longitudinal research studies.

1.4.3 Causal designs

Causal designs tries to identify cause-effect relationships and interactions between or
among variables used in a study.

Figure 1 provides a summary of research designs.

Figure 1: Types of research designs

Uses Types
• * Literature examination;
• * Articulate precise problems; • * Experience inspection;
• * Develop hypotheses; • * Focus groups;
Exploratory • * Create the significances of the study; • * Interviews;
research • * Remove unrealistic ideas; • * Projective tests;
• * Explain concepts • * Ethnographies

Descriptive • * Define the characteristics of segments; • * Longitudinal study;

research • * Estimate the percentage of persons • * Panel;
• who act in a specific manner • * Sample survey
• * Make definite predictions •

• Provides signals concerning causal • * Laboratory

Causal • relationships through: • experiments;
research • * Connected disparity; • * Field experiment
• * Time order in which variable occurs; •
• * Exclusion of other descriptions.

Source: (Iacobucci & Churchill

•
2010:60)

5|Page
1.5. What is a research proposal?
A research proposal is a plan or a road map that outlines the framework of how you will be
conducting your research on your chosen topic (title). It outlines the procedure necessary
for obtaining the information in order to structure or solve your research problem/s or
identify research opportunity/s. This plan forms the basis for the reminder of the research.
A research proposal is written before the study commences. It is a document which
basically outlines the following:

1. What the proposed 2. What are we trying to

research is about? find out or achieve?

3. How will I go about 4. What will we learn from

doing it? it and why that is worth
knowing?

5. How much will the

project cost?

1.5.1 Why do I need a research proposal?

6|Page
 ❖ A research proposal is required in order to plan and communicate your ideas on
your chosen title. It gives you an opportunity to think through you research
carefully and clarify and define what you want to research. You need to convince
people like your supervisors, funding agencies, the University that your research is
worth spending time, effort and money on. You therefore need to assure people that
your research will assist in solving the research problem or identify an opportunity.
When you have a clear plan of action regarding your research from the beginning,
the research process will move smoothly until its completion stages. It helps you to
measure your progress.

❖ You need to demonstrate your level of expertise on the area of study that you have
chosen. You need to convince the readers that you have read widely on the topic and
that there is adequate literature on your topic in order to successfully conduct the
research.

1.5.2 So what are my responsibilities?

❖ You must make sure that you have read widely on the topic. Do not jump into
conclusions and make wild assumptions.
❖ Since you have to demonstrate to your readers on the viability and merits of your
research, you need to be professional and adopt a clear, concise and scholarly
writing style.
❖ Use your supervisor/other researchers/lecturers as a sounding board for your
topic.

7|Page
1.6. Overview of the stages in developing a simple research proposal (Quantitative
study)

Identify a broad research area

Example: E - Marketing

Within this broad research area-identify a research topic/title that needs to be researched.
Example: E-Marketing adoption patterns and barriers in the implementation among small and
medium enterprises (SMEs) in Southern Gauteng

Conduct a literature review on the topic by reading widely. This will form part of the
introduction and background to the study

After reading sufficiently on your topic, try and formulate the research problem, research
questions (if required), hypotheses (if required) and objectives for your study

Formulate the design/methodology you will use in the study

Decide on how the data will be collected

Decide how the data will be analysed.

Decide on the ethical considerations that must be taken into account before embarking upon
the study

Decide on the how you will ensure reliability and validity

Prepare the bibliography according to the guidelines

8|Page
1.7 The process of developing a research proposal therefore entails:
❖ Choosing a topic
❖ Narrowing and focusing on your topic
❖ Formulating an appropriate review of literature with a theoretical
substantiation
❖ Formulating the problem statement with sufficient elaboration, including
research questions/hypotheses
❖ Formulating research objectives
❖ Deciding on your research methodology (Paradigms/research design and
methods)
❖ A substantiation of the significance/importance of the study (How it will
contribute to the existing body of research and literature)
❖ Proposing an approach to data analysis
❖ Proposing how reliability, validity and trustworthiness will be established
❖ Propose how ethical issues will be addressed within your study
❖ Proposing how many chapters your research project will entail and suggested
chapter headings
❖ Developing a bibliography
❖ Developing timelines and a budget for your study.

PLEASE NOTE: Your proposal does not permanently set what you would do during the
early stages of its development. It is just a starting point and throughout your research you
would probably discuss with your lecturer/presenter/HoD/Supervisor/Promoter and
various other role-players.

❖ You still have to learn quite a lot more about your object of research. After all,
you are a humble apprentice.
❖ One should NEVER identify oneself with the object and purpose of the study!
Otherwise you are not fair in your objectivity. Also keep in mind that you are an
‘external scientist” trying to make a contribution to a body of knowledge or new
knowledge. You need to discover and create new knowledge.
❖ Thinking and writing are closely associated with creativity and innovation. Just
think about it? Therefore fuzzy writing shows semblance to fuzzy thinking.
❖ When you have difficulty in translating your ideas into proper sentences you
often ramble (write incoherent information) and find difficulty to relate the
material that you have read. Therefore, when you write, ask yourself … I have
just written the information in a clearer and simpler manner, without being
subjective, but yet to the point.
❖ When you write do not preempt your specific epistemological viewpoint but
view the world as social systems with complexities.
❖ If you cannot find another research which relates closely to your topic, then you
are an unannounced genius – but listen to me… its better to assume that you are
NOT a genius BUT try to fill a GAP in research that MAY exist in management
and social sciences.

9|Page
 Chapter 2
Research area and research topic
Learning Outcome 2

2.1 Research area and research topic

 Research areas are usually broad fields of study within a discipline stated in few
words or sometimes just one word.

Examples
Consumer buying behaviour Market orientation
Relationship marketing Service quality
Shopping motives Occupational stress
Burnout Job satisfaction
Workplace accidents Recruitment and selection
Workplace absenteeism Training needs analysis
Reverse logistics Inventory control
Appraisal systems Budgeting
Fan attendance Cash flow management
Motivation Coaching/mentorship
Skills development/learnerships Employee retention
Work ethics Out-shopping
Fan behaviour Role conflict/role ambiguity
New product development Higher Education funding
Sexual harassment E-marketing
Branding Internal marketing

 Topics or titles are usually longer and written in a sentence form. Topics
typically comprise twelve to fifteen words. A topic can also be written with a
short sentence. This can be followed by a colon and another short sentence.

 The title should be concise and descriptive. For example; “An investigation of …”
could be omitted because they are referred to a deadwood in titles. Often titles
are stated in functional relationships because titles clearly indicate the
independent (IV) and dependent variables (DV).
 Examples

❖ Shopping motivation of Generation Ys in the purchase of fashion apparel: A

study among university students
❖ Shopping motivation of Generation Ys: A study of sport apparel consumers in
Southern Gauteng
❖ Training needs analysis among small and medium size enterprises: A study in
Southern Gauteng
❖ Financial needs analysis: A study of micro enterprises in Sebokeng and
Sharpville.
❖ NB: Make sure that your topic has a unit of analysis. Examples of unit of
analysis: Consumers, employees, SMEs, Engineers, Miners…

11 | P a g e
 Chapter 3
Literature review
Learning Outcome 3

3.1 Conducting a literature review

The literature review section presents one of the greatest challenges of the research
proposal to experienced and inexperienced researchers. The researcher should first
consider the purpose of the review of literature before searching for secondary data. A
literature review should serve the following purpose:

❖ To demonstrate your familiarity with and knowledge of the subject, it is

important that you are able to provide an integrated overview of your topic.
❖ To provide an outline of the relevant theories and concepts important within
the research project. Theory is important to allow the researcher to have a
framework within which to explain the findings. You must show awareness of
the most important and relevant theories, models, studies and methodologies.
❖ To assist in focusing on the research question. One of the biggest problems faced
by researchers in that they are unable to complete the research because they
have not sufficiently focused or delimited their research questions/objectives.
❖ To determine the extent of past research on the subject matter.
❖ To develop a hypothesis or a set of hypotheses.
❖ To identify methodologies and methods that past researchers have used. The
researcher may gain valuable insights into how other researchers have
approached similar questions in various contexts.
❖ To ensure that all relevant variables are identified.
❖ To allow a comparison of your findings with the findings of other researchers.

3.2 Sources of literature and reading

The main sources available are as follows:

❖ Books
❖ Peer-reviewed journals
❖ Conference papers
❖ The World Wide Web (The Internet/databases)
❖ Past theses/dissertations
❖ Government publications
❖ Bibliographies - An article which is relevant to your topic may also have a
relevant bibliography. Mark those sources in the bibliography and use the inter-
library loan section to borrow/or try to source the articles from the VUT
databases.
3.3 Mistakes often made in literature review

Many students find writing the literature review a difficult part of the research process and
often a weak section. The common mistakes made in this regard are outlined below:

❖ Making the review simply a list of past studies.

❖ Not relating the literature review to the study (title).
❖ Not taking time to find the best sources.
❖ Not going beyond your chosen discipline to find information.
❖ Relying on secondary rather than primary sources of literature e.g. only using
books rather than the relevant journal articles.
❖ Not considering contrary findings
❖ Raising problematic issues but not addressing them.
❖ Using very old literature (more than ten years old)
❖ Failing to cite influential references
❖ Failing to keep up with recent developments
❖ Cite irrelevant or trivial references

NOTE:
❖ It is important to have sub-heading in this section.
❖ It is important to keep in mind that you are telling a story to an audience. Try to
tell your story in a stimulating and engaging manner. Do not bore your readers.

So… lets’ start …

Introduction (approximately 2/3 paragraphs)

At this early stage, the topic needs to be introduced to the reader. The introduction
presents the subject under discussion and makes it clear to the reader. In addition to the
introduction, the researcher provides background material to strengthen arguments,
arising out of the review of literature and orientate the reader to what is already known
about the problem. It should also serve as a demonstration of the researcher’s command of
the current knowledge in a field. .

13 | P a g e
 How do I write my introduction to the study?

Your introduction...
❖ An introduction presents the topic. An introduction may be written in your
own words with source/s of reference/s and serves to introduce your topic to
the reader.
❖ Make sure you write your introduction in simple language. This section should
contain the rationale for your research. Why is the research needed!
❖ The introduction begins with a few short introductory paragraphs (1 to 2
paragraphs). The primary goal of the introductory paragraphs is to gain the
attention of the readers and to get them interested in the subject. The
introduction should clarify terminologies and definitions (and referenced
correctly).
❖ It should set the stage for the proposal and puts your topic in perspective. The
introduction often contains dramatic and general statements about the need
for the study. The first sentence in an introduction should be in your own
words and usually begins as a way of “setting the scene” for the topic you wish
to discuss. The second sentence then should provide a reference that supports
your opening argument. You cannot make any broad statements without using a
source of reference to substantiate what you are saying.

❖ Your first sentence in the introduction gives you an opening to be able to

describe the current situation in the area which you intend to research.
❖ When you are writing your paragraphs, make sure you have one idea per
paragraph. One paragraph should flow logically into the other paragraph to
show the linkages.

Examples of opening statements in an introduction could include:

❖ “In recent years the situation in/about/at …………” (Brown 2007: 34) or
❖ “South Africa is a developing country with focus on …., therefore a different
approach to …… is needed” (Robin & Jack 2006: 78).
❖ “Decision making is an integral part of the existence of any business which
impacts on the following areas ……” (Jack, Taylor & Brown 2004: 678).
❖ “Banks are currently operation in a highly competitive industry in South Africa.
It is necessary then to examine the following ….” (Madiba 2004: 897).

14 | P a g e
 Theoretical framework

How to write your theoretical framework

❖ In one to two sentences state the purpose of the background to the study
(literature review), i.e. what your literature review is about and what you hope
to achieve with it. Provide a brief overview what you will be doing in the rest of
the literature review. Clarify terminologies and definitions (and reference
correctly).

❖ Next, find supporting evidence from the literature to show that the purpose of
your literature review is a significant issue, i.e. why your literature review is
necessary. It is important because it shows what previous researchers have
discovered. It is usually long and primarily depends upon how much research
has previously been done in the area you are planning to investigate. If you are
planning to explore a relatively new area, the literature review should cite
similar areas of study or studies that lead up to the current research.

❖ Further definitions explanation of terminologies may also form part of your

literature review. Make sure you source and reference these explanations
correctly. Include it if your proposal uses special terms that are unique to your
field of inquiry or that might not be understood by the general reader.
"Operational definitions" (definitions that you have formulated for the study)
should also be included. An example of an operational definition is: "For the
purpose of this research, improvement is operationally defined as post-test
score minus pre-test score".

❖ Do not record your own point of view but rather use secondary/primary
sources of literature (textbooks and journal articles etc.) to emphasise problems
/opportunities/gaps in research. DO NOT use… I, me, we … anywhere in your
proposal.

❖ The arguments must lead to a concise statement of the problem. Requirements

may vary with regard to format and amount of details required. Yet the basic
structure and “major ingredients” or components are fairly standardized.

❖ If there is any theories underlying you area of research, you need to discuss the
theory/ies briefly and reconstruct any debates in literature.

NB: Make sure that you have discussed/defined (addressed) all the variables in your
topic.

15 | P a g e
 Chapter 4
Problem statement
Learning Outcome 4
4.1 Identifying the research problem

This section is often not an easy section to report on. The emphasis here is on the problem
being investigated in the research. After having explored the review of literature the
research problem may become much clearer. The research problem may arise out of a
debate in literature, or a gap in relevant research in a specific area. If it arises out of a
debate in literature briefly introduce the debate. If it arises out of a gap in literature explain
the gap and motivate in terms of the lack of research undertaken to address the gap.
Sometimes, the problem may arise from a request from a particular community; describe
the community and the problems confronting it with supporting sources, for example with
media or government/integrated development plan (IDP) reports. Within the problem
statement, clarify or quantify any concepts which may not be very clear. The general role of
defining the research problem is that the definition should:
❖ Allow you to obtain all the information needed to address the problem and
guide you in proceeding with the research by identifying suitable research
questions.
❖ The research problem should not be too broadly defined; otherwise you will end
up with very general research questions, rather than specific research
questions. Students often bring into the discussions problems that are not
relevant to the study or beyond the scope of the study.
❖ Determine all the variables (IV and DV) and their probable relationships (if
any).

A research problem should satisfy as many as possible the following criteria:

❖ be timely
❖ relate to a practical problem
❖ relate to a wider population
❖ relate to an influential or critical problem
❖ fill a research gap
❖ permit generalization
❖ sharpen the definition of an important concept or relationship
❖ create or improve an instrument
❖ provide possibilities for investigation

4.2 Do I need research question/s?

Sometimes, at the end of the problem statement and linked to the problem, research
questions may be formulated. This is permissible. Make sure that the research questions
are relevant and just not forced out (copied) from previous research (journal articles). The

16 | P a g e
research questions must be later linked or incorporated into your empirical objectives in
order to ensure that they are answered (achieved).

Research questions should be specific, detailed and concrete. Some researchers refer to
specific research questions as critical questions. Specific research questions are directly
answerable because they point candidly at the data needed to answer them.

Here are some examples of research questions

❖ What is relationship between job insecurity and job satisfaction?
❖ What is the relationship between parental break-up and the incidence of youth
suicide?
❖ What is the relationship between e-marketing adoption and business
performance?
❖ What are the barriers to e-marketing adoption?

Note: The first three examples are relationship based while the fourth example is not. In
the first three examples, you may have to undertake correlation/regression analysis to
establish the relationship. In the forth example, there is no relationship but instead the
questionnaire has to be developed in such a way to incorporate the question/s on barriers
to e-marketing adoption. This can be done by ranking the possible barriers (arising out of
literature) and making provision for other barriers that are not listed or through a factor
analysis approach to establish meaningful barriers to e-marketing.

Please note:
❖ A question well asked is a question half answered.
❖ Do not try to tackle multiple research projects in one study. Think about what
will happen when developing questions and how many questions you will have
to ask! Be realistic on what you can accomplish in your study. Ask questions that
directly relate to your study that are not too general or vague. Here is an
example of a vague question:

 What are the impacts of climate change?

 Rather ask a question like: What are the effects of climate change on agriculture?

❖ Try and limit your scope of the question. Use limiters such as
area/region/geography/ gap in literature/policy needs.
❖ A research question should be a product of your own critical thinking after you
have mastered the literature and clarified the relevant theoretical debates.

17 | P a g e
4.3 Do I need a hypothesis?

This question often comes up when writing research proposals. A hypothesis should be
used as and when necessary. A hypothesis is used when it can be deduced from theory or
when it is explained by a theory so that the study, in testing the hypothesis is actually
testing the theory behind the hypothesis. This comes from the classical or traditional view
of hypothesis deduction. However, not all social research should necessarily align itself
with this view. It is not necessary for all research studies to have a hypothesis or
hypotheses. One should note that a well developed research question/s can also perform
these functions as well. It should also be noted that hypothesis testing is usually associated
with probability samples where inferential statistics can be used.

You should use hypothesis in your study when:

 For each specific empirical research question, you can predict (in advance of the
empirical research, before analysis of the data) what you are likely to establish.

 If so, is the basis for that prediction based on a rationale or theory to back such
assertions?

 Do not have a hypothesis in your study just for the sake of having one. Well then,
use hypothesis if it is appropriate for your study.

Here are more issues in the usage of hypothesis:

 A hypothesis is a suggested answer to a problem. It has to be tested empirically

before it can be accepted, validated and incorporated into theory. If a hypothesis is
not supported by empirical evidence, it must be rejected.

 A hypothesis must be conceptually clear. All variables identified must be clearly

described.

 A hypothesis should have empirical referents. This is an essential feature of a

scientific approach to problems. One should not use moral judgments, attitudes
and values in developing a hypothesis. Expression like “good”, “bad”, “ought to”, “I
know or I saw that this is happening” should not be used as they are not scientific.

 A hypothesis must be specific. The range of the problem must be narrow enough to
allow a precise, well-delimited investigation. If the research problem is too wide,
the hypothesis will also be too general and not testable. If you skillfully developed
your specific research questions, you could translate them into testable hypothesis.
(Meet with your supervisor/promoter to obtain more clarity).

18 | P a g e
 A specific hypothesis makes mention not only of the independent and independent
variables/s, but also of the moderator or mediator and all control variables.

Here are some examples of hypothesis:

Ho1 There is no significant relationship between job insecurity and job satisfaction.
Ho2 There is no significant relationship between e-marketing adoption and business
performance (measured by turnover).
Ho3 Store loyal customers are more risk averse than non-loyal customers when buying
electronic products.

NB: The hypothesis in the first two instances is stated in the negative. In the third case
the hypothesis is stated as a statement. If the hypothesis is true then it is accepted. If the
hypothesis is false, then the hypothesis is rejected.

19 | P a g e
 Chapter 5
Research objectives
Learning Outcome 5

OBJECTIVES

PRIMARY OBJECTIVE THEORETICAL EMPIRICAL OBJECTIVES

OBJECTIVES
▪ This objective is
▪ These objectives are based on ▪ These objectives are based on what
formulated by using the
your literature review. you want to achieve in the actual
title as a basis because it is empirical study.
the overall objective for
your study. ▪ Start with words like: to conduct
a literature review; to carry out a ▪ Use your specific research questions
literature review; to conduct a developed earlier to carefully
literature synthesis etc. formulate these objectives.

NB: Sometimes you do not want to ▪ Start these objectives with words
start with the word “to” like: To evaluate;
▪ To ascertain;
▪ To determine;
▪ To identify etc.

PLEASE NOTE:

There are variations in literature on the format in formulating objectives. Some proposals
formulate only primary and secondary objectives (with the latter reflecting objectives
arising out of literature review and incorporating the empirical objectives). From
experience in evaluating students’ research proposals it was found that delineating
objectives into primary, theoretical and empirical objectives allow students who are
engaged in first time empirical research to develop a clear “mind map” on how to set out
their literature sections of the study. Students can also visualize the link between
theoretical objectives; specific research questions, hypotheses (if any) and empirical
objectives. Using this format also assist students to effectively “tie up” their research
questions, hypotheses, objectives and literature review in their concluding chapter. If you
are using primary and secondary objectives, then your objectives outlined in your proposal
must be stated as: primary objective and secondary objectives.

21 | P a g e
 Chapter 6
Research methodology/research approach/research design
Learning Outcome 6
6.1 Introduction

Methodology is a framework/plan for conducting your research project. It guides you in

selecting a plan which is appropriate to collect and analyze data. The research
methodology ensures that the study becomes relevant to the problem and uses
economically sound procedures. Sometimes and often used interchangeably in literature is
the issue of research designs. There are three types of research designs, namely,
exploratory, descriptive and causal research designs. These designs are usually
incorporated in the quantitative or qualitative approaches to the study. However these
designs can be discussed in your methodology chapter (Chapter 3). This only applies to
masters/doctoral students.

Usually the methodology/approach in a research study comprises a review of literature

and an empirical section to the study. The literature review is undertaken to develop a
theoretical background/framework to the study. The literature review must cover all
aspects of the theoretical objectives set out in the theoretical objectives section in order to
ensure a coherent flow of arguments and findings from previous studies. Textbooks,
journal articles, conference papers, government publications, industry publications,
electronic data bases and other Internet sources are used to develop a theoretical
background for the literature review section. Avoid using newspaper articles.

The empirical section of the study usually follows a qualitative or quantitative

approach. However, a third approach referred to as mixed methods which combines
aspects of both qualitative and quantitative perspectives are also used in research.
Qualitative approaches usually entail the following data collection methods: Focus group
interviews, depth interviews, projective techniques, word association tests, cartoon tests,
photo shoots, third person techniques and observation. Quantitative approaches usually
follow the following data collection methods: Surveys (door-to-door/face-to-face/personal
interviews, mail (postal), mall intercept, telephone, self-administered, computer-aided
telephone interviews (CATI) and experimentation.
.

6.2 Does your research focus lend itself to qualitative or quantitative or a bit of
both?

Here are some differences.

Table 6.1 Qualitative and quantitative research designs

Qualitative research Quantitative research

1. Used mainly for exploratory research 1. Used mainly to obtain descriptive and
conclusive data
2. Methods of inquiry are likely to be 2. Methods are formal, carefully planned and
informal and flexible. tightly controlled
3. Based on a small number of respondents 3. Interviewer uses a questionnaire and
who are interviewed individually or in must be followed exactly as instructed in
small groups. every interview.
4. It cannot be known, unless quantitative 4. It is often possible (depending on the
research is also conducted, how far the sampling method) to estimate how much
findings are true of the population from reliability can be given to the project’s
which the respondents are drawn findings.
5. Data collection is usually handled by 5. Data collection is usually handled by
research professionals trained interviewers or by using self-
completion questionnaires.
6. It is impossible to replicate (i.e. to exactly 6. Usually possible to replicate a quantitative
repeat) a qualitative project; every data project if required, because every interview
collection event in a project e.g. each focus in the project follows the same set of
group is different because of the informal procedures.
nature of the research.
7. Findings cannot easily or validly be 7. Findings usually expressed in numbers
expressed in statistical form and can be analysed using statistical
techniques.
8. The analysis and conclusions rely heavily 8. Because statistical procedures are used,
on the perceptions and interpretation skills the analysis is usually less open to dispute.
of the researcher.

6.3 Mixed methods in research

Mixed methods research is:
“An approach to research in the social, behavioral, and health sciences in which the
investigator gathers quantitative (closed ended) and qualitative (open ended) data,
integrates the two, and then draws interpretation based on the combined strengths of both
sets of data to understand research problems” (Creswell, 2015:2).

The three basic mixed methods designs are:

6.3.1 A convergent design, in which the intent of the researcher is to collect both
quantitative and qualitative data, analyse both datasets, and then merge the results of the
two sets of data analysis with the purpose of comparing the results. Below is an illustration
of convergent design.

23 | P a g e
Figure 6.1: A convergent design

Qualitative data Quantitative data

collection and collection and
analysis analysis
Phase 1

Merge

Interpretation

6.3.2 An explanatory sequential design in which the intent is to first use quantitative
methods and then use qualitative methods to help explain the quantitative results in more
depth. Below is an illustration of explanatory sequential design.

Figure 6.2: An explanatory sequential design

Qualitative data Qualitative data

collection and Explained collection and Inferences
by analysis drawn
analysis

6.3.3 An exploratory sequential design, in which the intent is first to explore a problem with
qualitative methods because the questions may not be known, the population may be
understudied or little understood. After this initial exploration, the researcher uses the qualitative
findings to build the second quantitative phase of the project. This phase may require the
designing an instrument to measure variables in the study, or designing a typology that is then
measured using existing instruments. Below is an illustration of exploratory sequential
design.

24 | P a g e
Figure 6.3: An exploratory sequential design

Phase 1 Phase 2

Qualitative data Quantitative phase-

collection and Builds into instrument design or
analysis intervention design

i
s Quantitative phase-
instrument design or Inferences
intervention design drawn

Phase 3

6.3 Some philosophical foundations and research paradigms

The relationship between the nature of reality and the generation of knowledge is highly
contested in academic research (Easterly-Smith, Thorpe & Lowe 2002:27). Accordingly,
Denscombe (2008:271), Denzin (2010:419), Symonds and Gorard (2010:123) characterise
debates on the nature of reality and knowledge as “research paradigm wars.” For instance,
Crotty (1998:66) noted that “different ways of viewing the world shape different ways of
researching the world”. Based on Crotty’s (1998:66) viewpoint, it seems that one of the
fundamental challenges confronted by researchers is the selection and justification of the
research paradigm(s).

Conceptually, a research paradigm refers to a set of philosophical assumptions and

beliefs that directs research execution (Jonker & Pennink 2010:29). A research paradigm
is defined by Guba and Lincoln (1994:200) as a “set of beliefs that define the nature of the
world, the individual’s place in it and the possible relationships to that world and its parts”.
A paradigm is simply a belief system (or theory) that guides the way we do things, or more
formally establishes a set of practices. In empirical studies, research paradigms are
instrumental in guiding the researcher in the selection of the research design, research
methodology and research instruments (Easterly-Smith et al., 2002:27). In short, the
research paradigm(s) guide(s) the researcher by defining the ontological, epistemological
and methodological assumptions of the study (Guba & Lincoln 1994:107). Across
disciplines (and within) there are varying views of what research is and how this relates to
the kind of knowledge being developed. Paradigms guide how we make decisions and
carry out research (Guba 1990). Your own discipline is also be guided by a paradigm and
through the research papers you read in your subject, you will begin to identify, through the
methodology the kind of paradigm that is used. As a researcher, it is important to know
where your discipline belongs, that there are different ways of viewing the world and that
your approach to knowledge is one of many.

25 | P a g e
A research paradigm is “the set of common beliefs and agreements shared between
scientists about how problems should be understood and addressed” (Kuhn, 1962).
According to Guba and Lincoln (1994:108), research paradigms can be characterised
through their:

• ontology – What is reality?

• epistemology – How do you know something?
• methodology – How do you go about finding it out?

The diagram below explains the above terms and the relationship between them:

Figure 6.4 Characteristics of paradigms

Ontology and epistemology are identified as the main sets of philosophical assumptions
that differentiate research paradigms (Saunders, Lewis & Thornhill 2009:119). Ontology
refers to the researcher’s perceptions of what constitutes knowledge and the nature of
reality (Bryman & Bell 2011:20; Saunders et al., 2009:119). Ontology can be defined as the
science or study of being and it deals with the nature of reality. Ontology is a system of
belief that reflects an interpretation of an individual about what constitutes a fact.

Thus, Guba and Lincoln (1994:108) identified the ontological question as “What is the form
and nature of reality and what can be known of it?” Epistemology, also known as the
“theory of knowledge” and is defined as the belief system that informs the generation of
what constitutes acceptable and valid knowledge in a study (Wahyuni 2012:69). According
to Guba and Lincoln (1994:108), the epistemological question is therefore “What is the
nature of the relationship between the knower and what can be known?” Epistemology is
our perceived relationship with the knowledge we are un/dis/covering. Are we part of that
knowledge or are we external to it? Your approach, for example, will be objective if you see
knowledge governed by the laws of nature or subjective if you see knowledge as something
interpreted by individuals. This in turn affects your methodology.

26 | P a g e
Axiology and methodology are the main basic values that guide the research paradigms
(Wahyuni 2012:70). Axiology refers to the influence of the researcher’s values in research
execution whereas methodology captures the research methods that are employed in the
study (Wahyuni 2012:70). By adopting the standpoint of methodological questions, the
researcher sets out to find out the nature of reality by following a set of research guidelines
and procedures (Guba & Lincoln 1994:108). A research paradigm therefore guides the
researcher by defining the ontological, epistemological, axiological and methodological
dimensions of the study (Saunders et al., 2009:119).

The main research paradigms that dominate contemporary literature include positivism,
post-positivism, constructivism and pragmatism (Creswell 2009:6; McMillan & Schumacher
2010:4). Positivists believe that there is a single reality, which can be measured and
known, and therefore they are more likely to use quantitative methods to measure this
reality. The post-positivist paradigm adopts the critical-realist ontology and is grounded
in the belief that the nature of reality is independent of human thoughts and can be
interpreted effectively through objective analysis of research objects (Wahyuni 2012:70).
Epistemologically, the post-positivist paradigm assumes that knowledge is generated
logically, based on a systematic and objective scientific inquiry (Creswell 2009:7). The aim
of the post-positivist paradigm is to generate objective knowledge through the use of a
credible research process that enhances the accuracy, validity, reliability and
generalisability of research findings (Schulze & Kamper 2014:131). Post-positivism
upholds the tenets of phenomenalism and perceives knowledge as objective and accurate
only if it is subjected to objective scrutiny (Bryman & Bell 2011:20). The post-positivist is
supported by the quantitative research methodology that is premised on the view that
knowledge is constructed through objective measurement of relationships among variables
in a given study (Glesne & Peshkin 1992:6). Thus, under the post-positivist paradigm,
knowledge is constructed based on quantified numeric data that are subjected to detailed
statistical analysis (McMillan & Schumacher 2012:12).

From an axiology perspective, the post-positivist paradigm firmly subscribes to the view
that the researcher’s values negatively influence the objectivity of the research outcomes
(Wahyuni 2012:71). Therefore, to enhance the generation of objective truth, the researcher
plays a detached role in a post-positivist paradigm (Schulze & Kamper 2014:131). The
major criticisms levelled against the post-positivist paradigm are its over-dependence on
rationality and its apparent failure to capture knowledge rooted in the subjective realms of
perceptions, beliefs, norms, attitudes, feelings and cultural values (Welman, Kruger &
Mitchell 2011:6). To complement the post-positivist paradigm, the constructivist paradigm
is often used in studies.

From an ontological perspective, the constructivist approach subscribes to the

viewpoint that knowledge is socially constructed and is dependent on the interaction
between the researcher, research objects and the natural environment in which the
research is conducted (Wahyuni 2012:70). Constructivists believe that there is no single
reality or truth, and therefore reality needs to be interpreted, and therefore they are more likely
to use qualitative methods to get those multiple realities. The constructivist paradigm

27 | P a g e
contends that the researcher is an active participant and is immersed in the research
process that naturally unfolds (Schulze & Kamper 2014:132). This paradigm affirms that
individuals in their capacity as social actors are central to the creation of knowledge
through the process of social interaction with the researcher (Bryman & Bell 2011:19;
Saunders et al., 2009:111). The overriding objective of the constructivist paradigm is to
understand behaviour and not to predict it (Harrison & Reilly 2011:8). In line with the
constructivist paradigm, knowledge is generated subjectively by analysing situational and
motivating factors that have an effect on the research objects (Wahyuni 2012:70).

Pragmatists believe that reality is constantly renegotiated, debated, interpreted, and

therefore the best method to use is the one that solves the problem.

The table below gives a more detailed overview of each paradigm and your own research
paradigm could very well sit in between one of the paradigms.

28 | P a g e
Table 6.2 Overview of paradigms

29 | P a g e
In the empirical section of the study the sampling design need to be clarified. The sampling
design components that are common in quantitative research is discussed below:

6.4 Methodology

The guiding principle for writing the methodology section is that it should contain
adequate information for the reader to determine whether it is sound.

Table 6.2: Example of the methodology section

Section Description of what aspects to include

Methodology The methodology is a body of knowledge that utilizes
various methods (study of various methods). Methods are
tools and techniques collecting data. What type of research
paradigm you are going to use! The terms quantitative and
qualitative research is popularly used to describe two
different world views or paradigms. A paradigm is “made
up of the general theoretical assumptions and laws, and
techniques for their application that members of a
particular scientific community adopt. Therefore a
paradigm is a comprehensive belief system, world view or
framework which you would use to guide your research. A
paradigm is a broad view or perspective of something
(Taylor, Kermode & Roberts 2007:5). They are may such
paradigms such as post positivism, critical theory and
interpretivism. Qualitative researchers often use
ethnographic, interpretive, critical or postmodern research
types while quantitative researchers often use positivist,
post positivist or objectivist type of research (Creswell,
1997).

In this section identify the methodology that underpins

your research and give a reason for your approach. You
many need to identify the research paradigm and
epistemologies that underpin your research.
Design procedure This section outlines how you are going to get your data.
Outline whether your research is qualitative or
quantitative or whether you are using triangulation
techniques.

Sub headings: This includes the sampling design process and includes
various sub-headings

▪ The population Define your population for your study.

Describe the location
Inclusion/exclusion criteria (gender/age, ethnic group
etc.)

30 | P a g e
Section Description of what aspects to include
▪ Sampling method, sample frame Describe whether you are using probability/non-
and sample size probability sampling method.
What would constitute your sample frame?
What will be your sample size and how did you arrive at
the sample size?

▪ Measuring instrument/s How you developed your questionnaire?

Did you adopt questionnaires from previous research? If
so you need to state the sources. Are these questionnaires
reliable and valid?
How many sections will your questionnaire comprise?
What will each section measure?
What type of scales will be used in each section?
NB: For qualitative study you will not have a questionnaire
but instead you may use a few guiding questions or open
ended questions on your topic-depending upon the
approach you are using.
▪ Data collection How will data be collected or administered?
Who will collect the data?
When will it be collected?

6.4.1 The target population

The target population refers to a group of people or objects which forms the subject of the
study. The target population can be consumers, employees, engineers, miners, managers,
retails stores (objects). These are often referred to as unit of analysis. Before the study
commences, the target population must be clearly defined. In your proposal, you need to
define the term and then explain clearly who or what will form part of the target
population.

Here is an example:

Topic: Job insecurity and the relationship with job satisfaction and motivation
among employees in steel manufacturing company

The target population for the study could be defined as follows:

Elements: Male and female engineers
Extent: Employed at Accelor Mittal in the Vanderbijlpark plant
Time: 2019

Obviously, you cannot write it like this but rather write it in a complete sentence, for
example: The target population for the study comprised male and female engineers who
are currently employed at Accelor Mittal steel company situated in Vanderbijlpark, in South
Gauteng, South Africa.

31 | P a g e
6.4.2 The sample frame
A sample frame is a list of elements in your target population. It consists of a list or a set of
directions for identifying the target population. Examples of sampling frames are: the
telephone directory, an association’s directory listing of firms in an industry, a mailing list
purchased from a commercial organization, the yellow pages directory, a map, voter’s roll,
database of names of employees obtained from the Human Resources division of a
company or a database of retail stores obtained from a municipal database. NB: A sample
frame is NOT a sampling method. Sometimes it is not possible to obtain an accurate sample
frame. In the absence of a sample frame, the survey location is then used to draw the
sample, for example a mall (for consumer studies).

6.4.3 The sampling method

There are two types of sampling, namely a probability sampling and a non-probability
sampling. Under probability and non-probability there are various techniques to choose
from. Here are some examples:

Figure 6.5 Probability and non-probability sampling methods

Probability sampling Non-probability sampling

Simple random Convenience

sampling sampling

Systematic
Judgment
sampling
sampling

Stratified
Quota
sampling
sampling

Cluster
Snowball
sampling
sampling

NB: Make sure you choose the correct method. For example you cannot choose a probability
sample if you do not have an accurate sample frame. You then have to resort to one of the
non-probability sampling techniques. Various research textbooks provide adequate discussion
on the different sampling s techniques. Choose a technique that best suits your study. You also
need to motivate the reason/s for your choice.

32 | P a g e
6.4.3.1 Probability sampling involves the selection of respondents that are
independent of human judgement producing a known and non-zero probability of any one
particular case in the population of cases being included in the sample.

6.4.3.1.1 Simple random sampling:

To draw a simple random sample it is necessary to have a complete and up-to-date
available sample frame. On this list each population element has to be numbered
sequentially so that they can be identified. The actual drawing of the sample entails the
generation of a pre-determined numbers randomly. One can use a table of random
numbers to generate the sample or a computer generated random numbers. An example of
simple random sampling is the drawing of lotto numbers. Probability can simply expressed
as follows:

Sample size
--------------
Population size

Therefore, if we select a simple random of 1000 from a population of 200 000, the
probability of any element being selected in the sample is 1000 as a proportion of 200 000,
that is 1 in 200 or 0.05.

6.4.3.1.2 Systematic random sampling

The manner in which a systematic sample is drawn is by systematically moving through the
sample frame and selecting every nth element. To do this one must first calculate the
sampling skip interval. Suppose that a project requires that a sample of 100 units is
required from a population of 8000. The sampling interval is calculated as follows:

Skip interval = number of elements in the population

------------------------------------------------
The required sample size

8000
-------
100

= 80.
Every 80thunit will therefore be selected to be sampled. If the random start is at
respondent 26, then the first sampling unit would be (80+26) = 106 and the next 186, 266
and so on.

6.4.3.1.3 Stratified random sampling

In this method of sampling, the population is first divided into a number of homogeneous,
non-overlapping groups which are called strata. Then within each strata a simple random
procedure is used to draw the sample. Stratified random sampling accepts that a
population has a high degree of variability and tries to ensure that the different segments

33 | P a g e
of the population in the sample in the same proportion as they occur in that population
(Webb 2002). The strata can be formed based on natural subgroups, for example the nine
provinces of South Africa. If they adequately represent the sample, then the sample error
will be reduced and the standard deviation will not be large. In stratification, the
population is divided into mutually exclusive strata and one can stratify according to
geographical, demographic, psychographic classifications, or they can be classified
according to income as well.

6.4.3.1.4 Cluster sampling

In cluster sampling, the population is divided into subgroups. Each cluster should contain
sampling units which are alike according to certain characteristics. Then, from each cluster
a random selection is made. An example where cluster sample is useful is if you need to
draw a sample of learners, say in Grade 12, in a certain city. Schools are then clustered. It
would be easier to make a list of all the schools in the city and then use simple random
sampling or systematic sampling to randomly select a few schools. In each of the selected
schools, all the Grade 12 learners can then be selected in the sample.

6.4.3.1.5 Multistage sampling

If the population is widely dispersed with respect to some characteristics, especially
geographical location, then it is wise to use multistage sampling. Multistage sampling is
therefore a form of cluster sampling in which more than one stage is involved in the
selection of the sample.

6.4.3.2 Non probability sampling

With probability sampling, the chance of selection are not known and therefore the ability
to generalize about the population is reduced. The following situations lends itself to use
non-probability sampling:
• Not much time is available and results are need quickly
• The measuring instrument needs to be tested
• Preliminary studies have t be done in the development stage of the survey
• Not much money is available
• The population is difficult to find

6.4.3.2.1 Convenience sampling

The selection of the sample is based on convenience. Such a sample may contain unknown
amounts of error/or bias. The laws of probability is not followed. Convenience sampling
can be carried out quickly and with minimal costs. In many situations it may not result in
representative samples. It is useful in exploratory research. Examples of convenience
sampling are:
• The first 300 customers that enter a departmental store
• People phoning in to give their opinions in response to a TV programme
• The first 200 cars entering a car park.

34 | P a g e
6.4.3.2.2 Judgement sampling
In judgement sampling an attempt is made to ensure that the sample is more
representative than it was in the previous type of sampling (convenience). Researchers
often use their experience or make use of the services of experts in appropriate fields to
choose appropriate sampling units on the basis of their assumed relevance to the research
project. If the population is too large, then judgement sampling may become less reliable.

6.4.3.2.3 Purposive sampling

Sampling units under this system are not chosen with the aim that they should be
representative of the population but the sample units instead should meet some criteria
which are thought appropriate to the successful completion of the research. For example,
the researcher wishes to launch a new type of running shoe. In this case the researcher may
select 250 to 300 athletes who are runners.

6.4.3.2.4 Quota sampling

In quota sampling an attempt is made to mirror in the sample the characteristics of interest
in the same population as they occur in the population. To use this method of sampling, the
researcher must first identify the categories of the people that need to be in the sample and
the required number (quotas) in these categories. If for example, it is decided to extract a
quota sample, from a population to test a new Sunday newspaper, then variables such as
age, sex, social class and income levels might be used to segment the population according
to these variables. The quotas may either be proportional or non-proportional. If the
population, for example, is 1000 and the sample to be selected is 200, and if it is known that
there are 55% females and 45% males in the population. The researcher will then select
55% females (55/100x200) =110 and 45 % males (45/100x 200) =90. This called
proportional sampling. However, if you just choose 200 elements from the population of
1000 regardless of whether they were males and females then this is called non-
proportional sampling.

6.4.3.2.5 Snowball sampling

In snowball sampling, initially a number of respondents are selected at random and then
interviewed and their views sough concerning a research topic. Then they are asked to
nominate other people who are also members of the population who may have the same
interest or characteristics. Thus the researcher first makes contact with a few people who
belong to the population and they are interviewed. For example, a few homeless people are
identified and interviewed. They are then asked to refer the researcher to other homeless
people in the area.

6.4.4 The sample size (quantitative studies)

The sample size refers to the number of elements to be included in the study. Important
factors that are considered in determining the sample size include: the importance of the
decision, the nature of the research, the number of variables, and nature of the analysis,
sample sizes used in similar studies, completion rates and resource constraints (Malhotra,
1999:332). If multivariate statistics are used, the sample size should be large (>100). In
probability samples, the size of the sample can be computed using a sample size formula
provided that certain key information is available. This is not possible in non-probability

35 | P a g e
samples. In non-probability samples, rule of thumb is often used: which is based on past
experience and sample sizes used in past studies (historical evidence approach).

Here is an example of how a sample size was determined:

The historical evidence approach was used to determine the sample size for this research
(Zikmund 1999:320). A comparison of similar past studies served as a guideline for
supporting the chosen sample size as shown in Table 8 below.

Table 8 Sample size determination based on the historical evidence approach

RESEARCHER YEAR SCOPE OF STUDY SAMPLE SIZE

Seyal, A. H., Rahim, M. & 2000 Use of IT among SMEs in Brunei 143
Rahman, M. N. A
Cloete, E., Courtney, S. & Fintz, J. 2002 Small business adoption of E- 195
Commerce in the Western Cape
Upfold, C. T & Sewry, D. A. 2006 Information security in 37
Grahamstown SMEs, South Africa
Beheshti, H. M. & Sangari, E. S. 2007 E-Business adoption in Swedish 500
SMEs
Uzoka, F. M. E; Shemi, A. P. & 2007 Behavioural influences on E- 200
Seleka, G. G. Commerce adoption in Botswana
Source: Seyal et al. (2000:8); Cloete et al. (2002:11); Upfold and Sewry (2006:5); Beheshti
and Sangari (2007:236); Uzoka et al. (2007:6)

Please note that minimum sample size recommendations for quantitative research is based
on having sufficient sample size to reduce the likelihood of convergence problems and to
obtain unbiased estimates or standard errors. Generally sample sizes of 250-400 is
recommended for multivariate analysis (CFA) or a ratio of 5:1 or 10:1 cases to free
parameters, (Anderson & Gerbing, 1984; Pallant, 2016:184). When conducting structural
equations modelling (SEM), Hair, Black, Babin and Anderson (2010:662) recommends that
studies comprising seven or less constructs should have a minimum sample size of 300
participants.

6.4.5 Sample size (qualitative studies)

Qualitative research sampling requires a careful selection of a sample of participants who
can assist to understand the central phenomenon that a researcher is exploring. The
traditional stance on sample size is not to specify a size but to consider size as a function of
when saturation occurs in a study. Saturation is defined as the point in data collection
when the researcher gathers data from several participants and the collection of data from

36 | P a g e
new participants does not substantially add to the codes or themes being developed. At this
point the researcher stops collecting anymore data.

Another method of determining sample size is to examine the number of published

qualitative articles/studies by design (e.g. narrative research, phenomenology, grounded
theory, ethnography, case study research) and use the sample sizes reported in these
studies as a guide.

Further the third method is what some authors such as Creswell (2015: 77) also
recommends:

- for narrative study – one or two participants

- for phenomenology- 3 to 10 participants
- for grounded theory – 20 to 30 participants
- for ethnography - a single culture sharing group.
- for case study research – four to five cases.

6.5 Method of data collection and measuring instrument

The survey method is commonly used whereby information obtained is based on asking a
variety of questions regarding behaviour, intentions, attitudes, awareness and motivations.
You need to substantiate the reasons for using a survey method in your study. The survey
method through the use of a questionnaire is simple to administer, the data obtained is
fairly reliable since responses are limited to the alternatives provided in the questionnaire
and fixed responses reduce variability in the results obtained compared to conducting
interviews when the questions are unstructured (Malhotra, 1999:178). You may want to
state that a structured questionnaire will be used. You may want to describe how the
questionnaire will be broadly structured. For example, a section on demographic profile of
respondents, a section on job insecurity, a section on job satisfaction. If you are planning to
adapt an existing instrument (questionnaire) then state it and provide correct source of
reference/s. You can also name the instrument if it is an established instrument, for
example, the Minnesota job satisfaction scale.

37 | P a g e
 Chapter 7
Data analysis
Learning Outcome 7
7.1 Introduction
In this section of your proposal, describe the statistical analysis that will be undertaken for
the study. Common types of analysis are the following: descriptive analysis in the form of
frequencies, tabulations, pie charts, bar charts which can be used to describe the
composition of the sample and the biographical aspects of respondents. In order to explore
relationships, correlations can be computed between variables (such as job insecurity and
job satisfaction). Regression analysis can be undertaken to examine causal and predictive
relationships among variables. Independent sample‘t’ tests can be used to establish
differences between variables such as gender. Analysis of Variance (ANOVA) can be used to
examine differences among more than two variables. You also need to indicate which
statistical package will be used to analyse the data. For example: The Statistical Package for
Social Sciences (SPSS) version 25.0 for Windows (SPSS: 2018) will be used to analyse the
data.

7.2 Data analysis and some basic statistical concepts

Data itself will not provide you with the answer to your research question/s. The data has
to be interpreted. This is often a stage of panic, especially among non-statisticians. In
quantitative data analysis it is therefore useful to have some idea of what the different tests
that one may encounter. This section presents an overview of some of the commonly used
methods of statistical analysis.

7.2.1 Using descriptive analysis

Descriptive statistics are those statistics which organizes and summarizes numerical data.
Inferential statistics allow inferences to be made to a wider population from your sample
by analyzing the associations or differences between two or more variables. The figure
below provides an overview of the basic descriptive statistics most frequently used.

38 | P a g e
Figure 7.1 Descriptive statistics

DESCRIPTIVE STATISTICS

Measures of Measures of Relational

central tendency dispersion measures

Mean, mode & Standard deviation Rate, ratio &

median percentage

7.2.1.1 Measures of central tendency

❖ Mean: It is the average score of observations of a variable. For example, you may
want to establish the attendance of Kaiser Chiefs soccer fans; rather than taking one
attendance figure (which may be misleading) an average over a period of time, for
example a soccer season may be undertaken.

Example:

5 matches were played by a soccer club over a season. The

attendance was as follows: 15 000, 19 500, 24 800, 7500,
18 900. Therefore the mean = sum of scores/ number of
observations. The mean = 15 000 + 19 500+ 24 800+
7500+ 18 000 (84 500) / 5 = 16 960.

❖ The mode is the most frequently occurring score. For example, you may find out
the most common age for individuals to drop out from competitive sports. In
this case, a more useful score than the mean score would be the mode.

39 | P a g e
 ❖ The median is the point that divides the observations, in order from highest to
the lowest values into equal parts. If the weekly hours spent training for a sport
were measured as:

2 2 4 5 6 10 10 11 15

The median would be 6 as there are 4 values above and 4 values below.

If N is even, the median value may be a decimal. For example, for scores 1, 2, 3, 4, the
median is 2.5.

7.2.1.2 Measures of dispersion

Measures of dispersion indicate the spread of the data around the mean or other measures
of central tendency. The most commonly reported measure is the standard deviation (SD).
The standard deviation measures the extent to which scores deviate from the mean. Thus
two samples may have identical means, but differing SD.

7.2.1.3 Relational measures

Rate: A rate can be useful in comparing different populations. The rate is calculated as
follows:

rate = number of actual occurrences/ number of possible occurrences.

For example, the rate of regular participation in sport in the UK may be 0.15 (that is fifteen
people from every hundred participants) whereas the rate in the USA may be 0.18 (or
eighteen in every hundred participants).

Ratio: You may be interested as to the proportion of different groups undertaking a

particular activity. A ratio describes the numerical relationship of parts of a group to each
other; for example it could describe the numbers of males compared to females in a sports
crowd. It is calculated as follows;

Ratio = number of members of subgroup one/ number of members of subgroup two

Example: If there were 32 000 males and 24 000 females watching a sports event, then the
ratio would be:

32 000/24 000 = 1.3 males for every female.

Percentage: A percentage (%) compares a subgroup to the total group, for example, using
the above example, the calculation will be undertaken as follows:

% of subgroup = (number of members of subgroup/number of members of the entire

group) X 100 % of male fans = 32 000/56 000 x 100 = 57.1 per cent.

40 | P a g e
7.2.2 Inferential statistics
The measures discussed above investigate single variables, such as team scores, hours
spent training in a week and so on. However in research one needs to often examine the
relationship between two or more variables. Inferential statistics assess the association
between independent and dependent variables. These may be bivariate (measuring the
effect of a single independent variable upon a single dependent variable) or multivariate
(involving more than two variables).

Figure 7.2: Two categories of inferential tests

Inferential statistics

Parametric Non parametric

Use interval or ratio data. We These are used with ordinal or
assume that the data is nominal data and the
drawn a normally assumptions parametric tests
distributed population i.e. need not be met.
data is not skewed

Inferential statistics require you to test a hypothesis, or often referred to as the null
hypothesis. A null hypothesis suggests that there is no relationship between your variables.
Thus you are assuming no association, and the role of the test is to contradict this
hypothesis.

7.2.2.1 Interpreting the results

Inferential statistics do not as such tell us whether there is a relationship between two or
more variables. Instead they calculate the likelihood of whether an apparent relationship
or difference between two or more groups is down to chance or not. Inferential statistics
provide a p- value, the less likely the results was due to chance. Here are some of the p-
values.

❖ a p-value of 0.10 indicates that in ninety cases out of a hundred, the result was due
to an actual association rather than chance. For example, p< 0.1 indicates a
likelihood of incorrectly rejecting a null hypothesis one in ten times.
❖ a p – value of 0.05 (p<0.05) is the generally accepted level of significance in social
sciences studies. Thus a p value of 0.05 or less, strongly suggests that you have a
relationship that is not due to chance.

41 | P a g e
7.2.3 Tests of association
Tests of association refer to those tests that measure whether two or more variables are
related, that is, as the independent variable changes, then the dependent variable changes
accordingly. There are a number of tests that you may consider using.

7.2.3.1 Correlations
❖ Correlation investigates the relationship between two variables consisting of
interval or ratio data, for example, the relationship between advertising spent and
sales of a particular sports product. A correlation can indicate:
❖ Whether there is a relationship between the two variables
❖ The direction of the relationship, i.e. whether it is positive or negative
❖ The strength or magnitude of the relationship.

A positive correlation exists where higher scores on one variable correspond with high
scores on another, for example, performance may increase self-efficacy. A negative
correlation is where higher scores on one variable correspond with lower scores on
another, for example, performance in a complex sporting task may decrease with
increasing levels of anxiety. A score of -1.00 represents a negative correlation, that is, as the
score on one variable decreases, then the score on the other variable increases. A score of
0.00 represents variables are not correlated or have no relationships with each other.
Correlation of +1.00 represents a perfect positive correlation, that is, the variables
increases or decreases together.

Whilst correlations are useful to identify relationships, they are unable to determine
causality, that is the extent to which variable X causes variable Y or does Y causes X. For
this reason, regression analysis is undertaken.

The Pearson correlation coefficient is used to analyse two variables collected at the
interval or ratio level of measurement where data is parametric. Thus, a correlation
between the height of a golfer and the average drive length would use a Pearson’s
correlation. The Spearman’s rank order correlation is the non-parametric equivalent of
the Pearson’s correlation, and is used with ordinal, or ranked data.

7.2.3.2 Regression analysis

Regression analysis between two variables effectively calculates a “best fit line”. This line
will subsequently predict the effect of one variable upon the other. Regression analysis is a
powerful procedure to analyze relationship between variables. It can be used to:
❖ Determine whether the independent variable (IV) explain a significant variation in
the dependent variable (DV), i.e. whether a relationship exists.
❖ Determine how of the variation in the dependent variable can be explained by the
independent variable/s i.e. the strength of the relationship
❖ Control other independent variables when evaluating the contributions of a specific
variable or a set of variables.

42 | P a g e
7.2.4 Testing for differences
Tests of differences generally assess whether differences between two samples are likely to
have occurred by chance, or whether they are the result of the effect of a particular
variable.

❖ The independent samples t-test

Using this test the researcher examines whether the mean scores of two different groups
can be considered as being significantly different. For example, we randomly assign
participants to two groups, one a rehearsal and one non-rehearsal group with regard to
golf putting. We then compare their scores to see if there is a significant difference. This
test can be used when:

− the data is interval or ratio in nature.

− the groups are randomly assigned
− the variance, or spread in the two groups is equal.

❖ Paired samples t-test

The paired t-test measures whether the mean of a single group is different when measured
at different times. For example, we wish to establish whether the performance of a group
on putting task after mental rehearsal was significantly different from that before the
rehearsal.

❖ Analysis of variance (ANOVA)

ANOVA is similar in nature to the independent t-tests. However, it allows you to ascertain
mean differences between more than two groups.

❖ Mann-Whitney test
This test is an alternative to the independent t-test, and is used when your data is ordinal
and non-parametric.

❖ Chi squared test

This test is useful for non-parametric nominal data. The chi-squared test compares the
actual or reported frequencies of a given variable with frequencies that would be expected
if the data was to suggest no differences between groups. Let us look at the following
example:

Table 7.1 Example-gender

Gender Prefer contact sport (%) Prefer non-contact sport (%)

Male 79 21
Female 22 78

43 | P a g e
It would seem apparent from the table that there is a significant difference between the
type of sport favoured by males and females. This can be confirmed by undertaking a chi-
squared test.

7.2.5 So which test should I use?

Only a few of the commonly used tests are discussed here. One needs to spend more time
with your statistician and read more about the different types of tests available. Here are
some suggestions:

Figure 7.3 Data types

▪ Consider chi-squared if you are

Nominal data interested in differences in frequency
counts

▪ Correlations for relationships (two

variables)
Interval or ratio ▪ Three or more variables, for
differences between scores, the use t-
test or ANOVA.
▪ To identify differences within one
group –used paired samples t-test.
▪ If comparing two groups, then use
independent samples t-test

▪ If you are examining relationships between

groups, then use Spearman’s rho.
Ordinal ▪ If you are examining differences between
groups, then use Mann-Whitney test.

44 | P a g e
 Chapter 8
Reliability, validity, trustworthiness and ethical issues
Learning Outcome 8

8.1 Reliability
Reliability refers to the extent to which a scale or a questionnaire produces consistent
results if repeated measurements are made (Malhotra, 1999:281). Therefore reliable
instruments provide stable measures at different times under different conditions
(McDaniel & Gates, 1996:350). There are three ways to assess reliability.

8.1.1 Test-retest reliability

Here respondents are administered identical sets of scale items at two different times
under nearly the same conditions with the same respondents. The time interval should be
between 2-4 weeks apart.

8.1.2 Alternative forms

Two equivalent forms of the scale are constructed and administered to the same
respondents at two different times (2-4 weeks apart). Reliability is measured.

8.1.3 Internal consistency

Internal consistency reliability is made up of two types, namely split half and Cronbach
alpha. In split half reliability the scale items are divided into halves. Reliability is then
calculated to examine correlations. High correlations indicate high internal consistency.
The Cronbach alpha technique is the average of all possible split-half coefficients,
resulting from different ways of splitting the scale items. Values range from 0 to 1. A value
of 0.70 and above suggests satisfactory levels of reliability (Malhotra, 1999:282).

NB: Commonly used methods: Cronbach alpha and Split-half methods. Other methods
have major disadvantages and often difficult to apply.

8.2 Validity
Validity is the extent to which a scale measures what it supposed to measure. There are
various types of validity.

Content/face validity
It is a subjective evaluation of a scale’s content to establish whether the scale covers all
aspects of what is being measured. Content/face validity is a subjective assessment made
using one’s own judgement or the judgement of an expert in that particular field. For
example a scale on store image would be inadequate if it omits major dimensions such as
quality, variety, assortment etc. This is a type of judgment is made by the researcher,
experts in the field etc.
Construct validity
Construct validity addresses the question of what construct or characteristics a scale is in
fact measuring. Construct validity requires a sound theory of the nature of the construct
being measured and how it relates to other constructs. There are two common types of
construct validity, namely convergent validity and discriminant validity. Convergent
validity is the extent to which a scale correlates positively with other measures of the same
construct. Discriminant validity is the extent to which a measure does not correlate with
other constructs from which it is supposed to differ (Malhotra, 1999:283). It involves
demonstrating the lack of correlation among different constructs. In convergent and
discriminant validities one can use correlation analysis.

Predictive validity
Predictive validity is the extent to which a future level of a criterion variable can be
predicted by a current measurement on a scale. In other words predictive value concerns
the extent to which the variable that has been measured is able to predict some future
state/value with which it is associated. For example, can a job insecurity scale predict the
intentions to quit a job or can a purchase intention scale predict the actual trial of a product
in the future. Here you can use regression analysis.

Concurrent validity
Concurrent validity is concerned with the relationship between the predictor variable (job
insecurity) and the criterion variable (intentions to quit). For example a long version and a
short version of an instrument can be administered to respondents and the results can be
compared.

NB: Commonly used methods are: Construct, content and predictive validity.

❖ At the proposal stage students are required to state which methods they are
going to use.

❖ In case of qualitative studies you also need to address the question of validity.
For example, if semi-structured interviews are used, what steps will be taken to
ensure that you accurately understand the content of the participants’
responses.

8.3 Trustworthiness in qualitative studies

Guba and Lincoln (2010) propose four criteria for evaluating qualitative findings and
enhancing trustworthiness. The list reflects the assumptions and epistemology underlying
qualitative research. These criteria can be both incorporated into a research design and be
used to assess qualitative findings:

❖ Credibility. This criterion tries to establish the believability or credibility of the

research findings from the perspective of participants of the research study. In
order to establish credibility member checking should be undertaken by
obtaining feedback on the transcripts/contents or the results of the study to find
out whether they are true reflection of the data analysed. The inclusion of

46 | P a g e
 member checking into the findings, that is, gaining feedback on results from the
participants, is one method of increasing credibility. Credibility is analogous to
internal validity, that is, the approximate truth about casual relationships, or
the impact of one variable on another.

❖ Transferability. Refers to the degree that findings can be transferred or

generalized to other settings, contexts, or populations. A qualitative researcher
can enhance transferability by detailing the research methods, contexts, and
assumptions underlying the study. Transferability is analogous to external
validity, that is, the extent to which findings can be generalized.

❖ Dependability. Pertains to the importance of the researcher accounting for or

describing the changing contexts and circumstances that are fundamental to
qualitative research. Dependability may be enhanced by altering the research
design as new findings emerge during data collection. Dependability is
analogous to reliability, that is, the consistency of observing the same finding
under similar circumstances.

❖ Confirmability. Refers to the extent that the research findings can be confirmed
or supported by others. Strategies for enhancing confirmability include
searching for negative cases that run contrary to most findings, and conducting
a data audit to pinpoint potential areas of bias or distortion (falsification).
Confirmability is analogous to objectivity, that is, the extent to which a
researcher is aware of or accounts for individual subjectivity or bias.

8.4 Ethical issues

Ethics refers to moral principles or values governing the conduct or behaviour of an
individual or a group (McDaniel & Gates, 1996:84). Some of the ethical issues that are
relevant for a study are as follows:

❖ Permission must be obtained from the management of the company in writing where
a company is used as a case study. If various companies are used, permission must be
obtained from all the companies.
❖ Participation in the study must be voluntary. Respondents will not be forced to
participate in the study.
❖ Personal data of respondents will be processed fairly and lawfully and used only for
the purpose of the study.
❖ Personal responses from individuals will not be ascribed to any individual. All data
will be computed in aggregate and not be linked to any respondent.
❖ The questionnaire will not contain the names of respondents - anonymity of
respondents will be maintained throughout the study
❖ Personal data shall be adequate, relevant and not excessive in relation to the purpose
of the study for which they are processed.
❖ Professional competence in the data collection and analysis will be maintained.
❖ Independent objectivity in the interpretation of the survey findings will be upheld.

47 | P a g e
❖ Informed consent – do participants freely engage in the research understanding what
the likely consequences are for them?
❖ Risk of doing harm – what harm could accrue to participants

So what should the format of a research proposal look like?

Note that the following is merely a guide! You do not have to slavishly adopt the formats!
Formats may vary depending whether your study is qualitative or quantitative in nature. Use
the following as a guide in writing your research proposal:

For quantitative research studies

1. INTRODUCTION (About 2 - 3 paragraphs)

2. THEORETICAL FRAMEWORK AND LITERATURE REVIEW ( about 2/3 pages)

3. PROBLEM STATEMENT (about half a page to a page)

4. OBJECTIVES OF THE STUDY

4.1. Primary objectives
4.2. Theoretical/secondary objectives
4.3. Empirical objectives

5. RESEARCH DESIGN/RESEARCH METHODOLOGY/RESEARCH APPROACH (choose

one of the headings i.e. research design or research methodology or research approach)
5.1. Literature review
5.2. The empirical research design
5.3. Population
5.4. Sample frame
5.5. Sampling method
5.6. Sample size
5.7. Data collection and measuring instrument

6. STATISTICAL ANALYSIS/DATA ANALYSIS

7. RELIABILITY AND VALIDITY
8. ETHICAL ISSUES
9. BIBLIOGRAPHY

48 | P a g e
 For qualitative research studies

1. INTRODUCTION
2. THEORETICAL FRAMEWORK AND LITERATURE REVIEW
3. PROBLEM STATEMENT
4. PURPOSE OF THE STUDY
5. RESEARCH DESIGN AND METHODOLOGY
6. RESEARCH PARADIGM
Type of design
Role of the researcher
Sample /site and participant selection
Data collection strategies

7. DATA ANALYSIS STRATEGIES

8. METHODS OF ESTABLISHING TRUSTWORTHINESS
- Credibility
- Transferability
- Dependability
- Confirmability

9. ETHICAL ISSUES
10. BIBLIOGRAPHY

The end… thank you

---------------------------------------------------------------------------------------------------------------------
--------
References
ANDERSON, J. C. & GERBING, D. W. 1984. The effect of sampling error on convergence,
improper solutions, and goodness-of-fit indices for maximum likelihood confirmatory
factor analysis. Psychometrika, 49(2), 155-173. https://doi.org/10.1007/BF02294170

BABBIE, E. & MOUTON, J. 2010. The Practice of Social Research. 10th ed. Cape Town: Oxford
University Southern Africa.

BRYMAN, A. & BELL, E. 2011. Business Research Methods. 3rd ed. New York: Oxford
University Press.

BRYMAN, A. 2008. Social Research Methods. 3rd ed. Oxford: Oxford University Press.

BURNS, A.C. & BUSH, R.F. 2014 Marketing Research, 7th ed. Florida: Pearson

49 | P a g e
CRESWELL, J. 1997. Qualitative inquiry and research design: Choosing among five
traditions. Thousand Oaks, CA: Sage Publications.

CRESWELL, J. W. 2009. Research Design, Qualitative, Quantitative and Mixed Methods

Approaches. 3rd ed. Thousand Oaks: Sage Publications.

CRESWELL, J.W. 2015. A concise introduction to mixed methods research. Los Angeles:
Sage Publications.

CROTTY, M. 1998. The Foundations of Social Research: Meaning and Perspective in the
Research Process. London: Sage Publications.

DENSCOMBE, M. 2008. Communities of practice: A research paradigm for the mixed

methods approach. Journal of Mixed Methods Research, 2(3):270-283.

DENZIN, N. K. 2010. Moments, mixed methods and paradigm dialogs. Qualitative Inquiry,
16(6):419-427.

DICK, A. L. 1999. Epistemological positions and library and information science. Library
Quarterly, 69:305-323.

EASTERLY-SMITH, M., THORPE, R. & LOWE, A. 2002 Management research: an.

introduction. 2nd ed. London: Sage Publications.

FEINBERG, F. M., KINNEAR, T. C., & TAYLOR, J. R. 2013. Modern marketing research. 2nd ed.
Mason, OH: South-Western.

GLESNE, C. & PESHKIN, A. 1992. Becoming qualitative researchers: An introduction. White

Plains, NY: Longman

GUBA, G.G. & LINCOLN, Y.S. 1981. Effective evaluation: Improving the usefulness of
evaluafion results through responsive and naturalistic approaches. San Francisco, CA:
Jossey-Bass.

GUBA, E. G. 2010. Criteria for assessing the trustworthiness of naturalistic inquiries.

Educational Communication Journal, 29(12):75-91.

GUBA, E. G. & LINCOLN, Y. S. 1994. Competing Paradigms in Qualitative Research. In

Denzin, N.K & Lincoln, Y.S. eds. The Handbook of Qualitative Research (pp.105-117).
Thousand Oaks: Sage Publications.

GUBA, E. & LINCOLN, Y. 1994. Competing paradigms in qualitative research. In N. Denzin

and Y. Lincoln (eds.), Handbook of Qualitative Research (pp. 105-117). California: Sage
Publications.

GUPTA, S. L. 2011. Marketing Research. New Delhi: Excel Books.

50 | P a g e
HARRISON, R.L. & REILLY, T.M. 2011. Mixed methods designs in marketing research.
Qualitative Market Research: An International Journal, 14 (1): 7-26.

IACOBUCCI, D. & CHURCHILL, G.A. 2010. Marketing Research: Methodological Foundations.

South-Western: Cengage Learning.

JONKER, J. & PENNINK, B. 2010. A Concise Guide for Master and PhD Students in
Management Science. Verlag Berlin Heidelberg: Springer.

MALHOTRA, N. K. 1999. Marketing Research: an applied orientation. London: Prentice-Hall

International.

MALHOTRA, K. N. 2010. Marketing Research: An Applied Orientation. 6th ed. Upper Saddle
River: Pearson Education International.

McDANIEL (Jr), C. & GATES, R. 1996. Marketing Research Essentials. Texas: Wiley.

McMILLAN, J. H. & SCHUMACHER, S. 2010. Research in Education: Evidence Based Inquiry.

7th ed. Boston: Pearson Education International.

SALKIND, N, J. 2012. Exploring Research. 8th ed. New Jersey: Pearson Education.

SAUNDERS, M., LEWIS, P. & THORNHILL, A. 2009. Research Methods for Business Student.
5th ed. Essex: Pearson Education Limited.
SCHULZE, S. & KAMPER, G. 2014. The use of mixed methods as reflected in two eminent
South African educational research journals. Journal for New Generation Sciences,
10(1):130-147.

SYMONDS, J.E. & GORARD, S. 2010. Deaths of mixed methods? Or rebirth of research as
craft. Evaluation and Research in Education, 23(2):121-136.

TAYLOR, B., KERMODE, S. & ROBERTS, K. (2007). Research in Nursing and Health Care:
Evidence for Practice. (3rd ed.) Sydney: Thomson.

WAHYUNI, D. 2012. The research design maze: Understanding paradigms, cases, methods
and methodologies. Journal of Applied Management Accounting Research, 10(1):69-80.

WEBB, J.R. 2002. Understanding and Designing Marketing Research. 2nd ed. Glasgow,
Scotland: Thomson.

WELMAN, C., KRUGER, F. & MITCHELL, B. 2011. Research Methodology. 3rd ed. Cape Town:
Oxford University Press.

51 | P a g e