Sunteți pe pagina 1din 56

UNIVERSITY OF KWAZULU-NATAL

(PMB CAMPUS)
FACULTY OF MANAGEMENT STUDIES
SCHOOL OF MANAGEMENT, IT AND GOVERNANCE
DISCIPLINE OF HUMAN RESOURCE MANAGEMENT

HONOURS: RESEARCH METHODOLOGY

2016
COURSE NOTES

CHAPTER 1 - RESEARCH IN BUSINESS


1.1 WHY STUDY RESEARCH METHODOLOGY?
There is a need for training in the scientific method and its application to decision-making.
There is a need for more scientific decision-making.
Firstly, the manager needs more and better information. Secondly, there is an availability of
improved techniques and tools to meet this need.
Due the great complexity in organisations today, there is increasing risks in business decisionmaking. Hence there are more variables that need to be considered. This explains the
increasing demand for sound information.
1.2 WHAT IS RESEARCH?
Research is a systematic and organised effort to investigate a specific problem that needs a
solution. In doing research a series of steps are designed and followed. The first step is to know
where the problem areas are in the organization, and to identify as clearly and specifically as
possible, the problems that need to be studied and rectified. Once the problem is clearly
defined, steps are taken to gather information, analyse the data and delineate the factors
associated with the problem. By taking the necessary corrective action, the problem could be
solved. This entire process by which problems are attempted to be solved, is called research.
Thus, research involves a series of well-thought-out and carefully executed activities that will
enable one to know how organisational problems can be solved or at least minimised. Research
thus encompasses the processes of inquiry, investigation, examination and experimentation.
These processes have to be carried out systematically, diligently, critically, objectively and
logically. The expected end results would be to discover new facts that would help to deal with
the problem situation.
1.3 THE PURPOSE OF THE STUDY

Reporting - simply summarising or giving an account of some data, for example, statistics.

Description - Tries to discover answers to questions of who, what, when, where and how.
Defines subjects and concepts.

Explanation - Gives reasons of what has been described via observation. Theories and
hypotheses are used to account for influences that caused the phenomenon to arise.

Prediction - Once you can explain why a phenomenon occurred, it is useful to predict when
it will occur again. Prediction is important in planning, development of better theory of the
phenomenon, forecasting.
2

Control - Once we can explain and predict, it is useful to control the phenomenon.
Control is thus the logical outcome of prediction.

1.4 TYPES OF RESEARCH


Applied Research
When research is done with the intention of applying the results of its findings to solving
specific problems currently being experienced in the organisation.
Practical problem-solving emphasis.
Pure/Basic/Fundamental Research
When research is done chiefly to improve our understanding of certain problems that commonly
occur in organisational settings and how to solve them.
Theoretical problem-solving.
No direct impact on performance, action or policy decisions.
1.5 WHAT IS GOOD RESEARCH?
1) The purpose of the research, or the problem involved should be clearly defined and
delineated.
2) The research procedures used should be described in sufficient detail to enable another
researcher to repeat the research.
3) The procedural design of the research should be carefully planned to yield results that are as
objective as possible.
4) The researcher should report clearly and honestly any flaws in the procedural design and
estimate their effect upon the findings.
5) Analysis of the data should be sufficiently adequate to reveal its significance and
appropriate methods of analysis should be used.
6) Conclusions should be confined to those justified by the data of the research and limited to
those for which the data provides an adequate basis.
7) Greater confidence is established if the researcher is experienced, has a good reputation and
is a person of integrity and conforms to the hallmarks of scientific research.

1.6 CONCLUSION
Due to the complexity and dynamic nature of organisations and the environment, the manager
today has to know more. Research is one way to obtain pertinent information and knowledge.
The manager can either engage in his own research or get the research services. Hence, in both
cases the study of research methodology is important. In the case of the latter, the manager
needs a knowledge of research methodology to be able to judge research quality. Good research
must:a) Be purposeful.
b) Have clearly defined goals.
c) Have carefully planned procedures.
d) Be objective.
e) Have good analysis, reporting, conclusions and recommendations.
The manager-researcher relationship is important as both have an obligation to make the project
meaningful.

CHAPTER 2 - SCIENTIFIC THINKING


2.1 KEY CONCEPTS
1) OPERATIONAL DEFINITIONS

Operational definitions state a concept in terms of specific testing criteria.


Empirically define, that is, be able to count, measure or gather information in some way.
The purpose is to provide an understanding and measurement of the concept.

2) VARIABLES
A variable is anything that can take on differing or varying values. The values can differ at
various times for the same object or person, or the values can differ at the same time for
different objects or persons.
The dependent variable is the variable of primary interest to the researcher. The researcher's
goal is to explain or predict the variability in the dependent variable.
The independent variable is the one that influences the dependent variable in either a positive
or negative way. That is, when the independent variable is present, the dependent variable is
also present, and with each unit of increase in the independent variable, there is an increase or
decrease in the dependent variable also. In other
words, the variance in the dependent
variable is accounted for by the independent variable.
The moderating variable is one that has a strong contingent effect on the independent variabledependent variable relationship. That is, the presence of a third variable (the moderating
variable) modifies the originally expected relationship between the independent and dependent
variables.
The moderating variable is the explanatory variable.
It is the second independent variable and has a significant contributory effect on the original IVDV relationship.
It depends on the focus.
The extraneous variables are variables that affect a given relationship.
The intervening variables are a mechanism through which the IV and MV may affect
the
DV. It is a factor that theoretically affects the observed phenomenon but cannot be seen,
measured or manipulated. Its effects are inferred from the effects of the IV and the MV on the
DV.

2.2 THE ROLE OF OBJECTIVES

Guides the direction of the study.


Identifies facts that are relevant.
Limits to key points and interesting information.
Provides a framework for discussion of literature chapters and for organising the
conclusions and recommendations that result.

2.3 THEORY VS MODELS

Theoretical (academic) as opposed to practical.


When explanation is not sufficiently attuned to specific empirical conditions.
Ideal is to combine fact and theory.
Theory is a set of systematically interrelated concepts, definitions and propositions that are
advanced to explain and predict phenomena (facts).
Theory narrows down the range of facts we need to study. Theory suggests ways in which a
study should be done to yield the greatest meaning. Theory outlines the objectives of the
study.
Theorie's role is explanation.
Model's role is representation.

CHAPTER 3 - THE RESEARCH PROCESS


3.1 BACKGROUND
The research task is a sequential process involving clearly defined steps. The idea is to develop
a project and keep it orderly as it unfolds. But one can miss steps, do them concurrently and
simultaneously.
3.2 FORMULATING THE RESEARCH PROBLEM
I) Is it researchable?
Research - by asking questions - answered by data collection methods. Questions must be
prioritised in terms of management decisions and policies. There must be procedures and
techniques suitable to answer research questions.
Ill-defined or ill-structured problems are complex problems that are bound by constrains eg.
analysis, too many interrelated facets for measurement, suitable method of data collection not
available.
AVOID - Needs thorough exploratory study.
II) The Question Hierarchy
Approach the research process by stating the basic problem that prompts the research. Develop
other questions by progressively breaking down the original question into more specific ones,
that is, the hierarchy of questions whereby questions may be defined in accordance with
management questions, research questions, investigative questions and measurement questions.
The management question: Decision that the manager must make.
The questions is broad - broken down to more specific subquestions - does not specify what
research is to be done.
The research question: The research translates the manager's question into a research question, that is, a factorientated, information-gathering question.
There are different ways to address a problem - the expertise of the researcher will determine
the strategy to be adopted - A wrong research for the right problem can be dangerous.
Eg. Reluctance to multiskilling may not be job rotation but group dependency or
shiftwork.
Managers must express a genuine need for specific information on which to base a decision.
7

A research question is the hypothesis or question that best states the objective of the
research.
The investigative question: From general questions to a more specific level.
General questions are well-defined characteristics of the problem and clearly stated research
questions. These lead to essential sub-questions that will guide the planning stages of the
research process.
Investigative questions are those that the researcher must answer to sufficiently respond to
the general research question.
The measurement question: Questionnaire formulated - previously validated ones - questions asked to respondents.

3.3 EXPLORATION
Search for published data.
Researchers speak to well-informed people on the topic - interview well known people in the
field.
Read specific books and periodicals.
Revise the problem and determine what is needed to secure answers to the proposed
questions.
Some problems may be resolved only by exploration and may then terminate the problem.

3.4 QUESTION REVISION: FINE-TUNING


After exploration, the researcher obtains a clearer/crystallised picture of the problem. Either the
question has been answered and the project is terminated or a different or modified question has
appeared. The refined question has greater focus and clarity in terms of:

The variables to be used in the study - operational definitions of key variables.


Review of investigative questions into specific questions.
Be sure that hypotheses fulfil the objective of the study.
Determine evidence needed to be collected to answer hypotheses and objectives.
Set the limitations of the study by clearly stating what is not part of the problem. This will
set the boundaries of analysis for the researcher.

3.5 DESIGNING THE STUDY


The design of the study prescribes how objectives will be fulfilled and questions answered.
There are numerous procedures and data collection methods available which makes it difficult
for the researcher to choose from. The design of the study determines the nature of the study,
for example, exploratory, descriptive or hypothesis-testing, the type of study, the duration of the
study, data collection methods, the field work, the type of questions, how to establish reliability
and validity, sample size, who are the administrators and do they need training?
The researcher must use the most appropriate techniques given the nature of the study and not
choose that which is most convenient or that which he/she most comfortable with -Rather, one
that offers the greatest insight.
NB. Consider several competing designs first!
3.6 SAMPLING
Identify the target population and select the sample.
Determine: Sample group ie. who?
Sample size ie. N.
What is the focus ie. to observe, to question.
A sample is part of the whole population carefully selected to represent the population.

3.7 RESOURCE ALLOCATION AND BUDGETS


Data collection is the most costly.
Administrators - paid, trained, transported.
Divide Budget into: Project planning.
Data gathering.
Analysis, interpretation and reporting.
Plan use of funds so that the study does not fall away.
The researcher must be able to justify costs of the project and identify source and method of
funding, for example, rule-of-thumb (% of annual sales revenue), departmental budgeting,
task budgeting.

3.8 VALUING RESEARCH INFORMATION


The research should produce added revenues or reduce expenses. This is easier said than done
because it is difficult to measure. Evaluation methods include Ex Post facto Evaluation (after
the research process, an objective estimation of the contribution of each project to corporate
profitability, cost-benefit analysis - useful for future research), Prior Evaluation, Option
Analysis and Decision Theory.

3.9 THE RESEARCH PROPOSAL


Activity that develops concurrently with the project planning phase of the study.
The choices the researcher has made in the preliminary stages.

Written proposal - Purpose:Parties understand the project's purpose and proposed methods of investigation.
Time, budgets and responsibilities are clearly outlined.
Research methodology ie. sample, sample size, specifications of field work, proposed
analysis techniques, data capturing and processing.

Proposal contents:Problem statement:What has been done?


What is lacking?
What will be done by the researcher to contribute to what is lacking?
Gives problem and reflects nature of research.
What will be done, that is, the methodology.

3.10 PILOT TESTING


Smaller version of data collection process - same procedures and protocols but with smaller
samples.
Colleagues, respondents, or potential respondents.

3.11 DATA COLLECTION


The method selected determines how data will be collected, for example, questionnaires,
standardised tests, observational forms, records, interviews.
Objectivity versus subjectivity.
Raw data - processed
Captured, tabulated, analysed.
Computers make it possible to edit, code and enter data.
10

3.12 ANALYSIS AND INTERPRETATION


After collecting the data, we need to analyse it. Data analysis usually involves reducing data
collected to a manageable size, developing summaries, looking for patterns or categories,
and applying statistical techniques.
Precoding and scaled responses are useful.
Data must be analysed in terms of finding solutions to problems and fulfilling objectives and
hypotheses.

3.13 REPORTING THE RESULTS


Prepare a report and reflect the findings and recommendations for the purpose of making the
decision. The style and organisation of the report depends on the target audience, the occasion,
and the purpose of the research. Solving the problem in the manager's mind - hence accurately
assess the manager's need throughout the research process and incorporate this understanding in
the final report. The research project must contain:
a) Introduction
b) Literature Survey/Review
c) Research Methodology
d) Research Findings - Reporting of Results
e) Discussion of Results
f) Conclusions and Recommendations

11

CHAPTER 4 - THE RESEARCH PROPOSAL


4.1 PURPOSE OF THE PROPOSAL
The purpose of the proposal is to: Present problem and its importance.
Discuss research efforts of others who have worked on related problems.
Suggest data necessary for solving problems and how the data will be gathered, processed
and interpreted.
A well planned and well documented proposal indicates:

What will be done?


Why it will be done?
How it will be done?
Where it will be done?
To whom it will be done?
What is the benefit of doing it?

SUPERVISOR: Can ensure that the project is: Carefully planned.


Properly written.
Properly organised.
Can ensure that you are adequately defining the problem you are addressing.
Can ensure that you reflect the correct objectives from which carefully worded research
questions would emerge.
Can ensure that you have set the logical steps to accomplish the stated goals.
THE PROPOSAL ENABLES YOU AND YOUR SUPERVISOR TO SPOT FLAWS OR
ERRORS TIMEOUSLY AND TAKE CORRECTIVE ACTION!
THE PROPOSAL SERVES AS A GUIDE!

12

4.2 STRUCTURING THE RESEARCH PROPOSAL


1) Topic
2) Motivation
a)
b)
c)
d)

What has been done relating to the key variables so far. (Literature Review)
What is missing?
What you are researching and how study is going to contribute to what is missing?
Importance/Benefits of your study.

3) Objectives of the study. (Minimum 3)


4) Hypotheses.
5) Research Methodology/Research Design.
a)
b)
c)
d)

Sample - who?
Sample - Size.
Data collection Method.
Questionnaire:-

Established? or to be formulated.
What variables to be covered?
What scaling?
Use of biographical data.

e) Administration - who, when and where?


f) Capturing data.
g) Processing.
h) Analysis.
6) Bibliography (Optional)
Budget.
Resources Required.

NB. NEATLY PRESENTED.

13

CHAPTER 5 - MEASUREMENT
5.1 PROBLEM DEFINITION
A problem may be defined as any situation where a gap exists between the actual and the
desired ideal state. A problem definition or problem statement is a clear precise, and succinct
statement of the question or issue that is to be investigated with the goal of finding an answer or
solution. Problem definitions could pertain to: existing business problems where a manager is looking for a solution.
situations that may not be posing any problems at present but where the manager feels that
things may be improved.
areas where some conceptual clarity is needed for better theory building.
situations in which a researcher is trying to answer a research question empirically.

Examples of well-defined problems


1. To what extent has the new advertising campaign been successful in creating the highquality, customer-centered corporate image that it was intended to produce?
2. What is the influence of information on price and quality on consumers' evaluation of
competing brands?
5.2 MEASUREMENT
5.2.1 SCALES AND MEASUREMENT
After having identified the variables of interest, how are they measured?
To do this, a knowledge of scales is important.
A scale is a tool or mechanism by which individuals are distinguished on the variables of
interest to our study, in some form or the other. There are four basic types of scales:NOMINAL SCALE
A nominal scale allows the researcher to assign subjects to certain categories or groups. For
example, with respect to the variable of gender, respondents can be grouped in two categories,
namely males and females. These two groups can be assigned code numbers 1 and 2. These
numbers serve as simple and convenient category labels with no intrinsic value. The
information that can be generated from nominal scaling is to calculate the percentage (or
frequency) of males and females in our sample of respondents. Thus, the nominal scale gives
some basic, categorical, gross information.

14

ORDINAL SCALE
An ordinal scale not only categorises the variables in such a way as to denote qualitative
differences among the various categories, it also rank-orders the categories in some meaningful
way. With any variable for which the categories are to be ordered according to some
preference, the ordinal scale would be used. The preferences would be ranked (for example,
from best to worst, first to last) and numbered 1, 2, and so on. The ordinal scale provides more
information than the nominal scale. However, in ordinal scaling, even though it is known there
are differences in the ranking of objects, persons, or events investigated, the magnitude of these
difference is unknown. This deficiency is overcome by interval scaling.
EXAMPLE
Rank the following five characteristics in a job in terms of how important they are for you. You
should rank the most important item as 1, the second most important as 2, and so on, until you
have given each of the five items a rank of 1, 2, 3, 4, or 5.
Job Characteristic

Ranking of Importance

The opportunity provided by the job to:1.


2.
3.
4.
5.

Interact with others.


Use a number of different skills.
Complete a whole task from the beginning to the end.
Serve others.
Work independently.

15

______
______
______
______
______

INTERVAL SCALE
An interval scale enables certain arithmetical operations to be performed on the data collected
from the respondents. Whereas the nominal scale only allows to qualitatively distinguish
groups by categorising them, the ordinal scale allows to rank-order the preferences, and the
interval scale allows to compute the means and the standard deviations of the responses on the
variables. In other words, the interval scale not only groups individuals according to certain
categories and taps the order of these groups, but it also measures the magnitude of the
differences in the preferences among the individuals.

EXAMPLE
Indicate the extent to which you agree with the following statements as they relate to your job,
by marking a cross (X) against the appropriate number, using the scale below.
1 - Strongly disagree
2 - Disagree
3 - Neither agree nor disagree
4 - Agree
5 - Strongly agree
_______________________________________________________________
1
2
3
4
5
____________________________________________________________________________
The following opportunities offered by the job
are very important to me:
a. Interacting with others.
1
2
3
4
5
b. Using a number of different skills.
1
2
3
4
5
c. Completing a task from beginning to end.
1
2
3
4
5
d. Serving others.
1
2
3
4
5
e. Working independently.
1
2
3
4
5
____________________________________________________________________________

16

RATIO SCALE
The ratio scale overcomes the deficiency of the arbitrary origin point given to the interval scale,
in that it has an absolute (in contrast to an arbitrary) zero point which is a meaningful
measurement point. Thus, the ratio scale not only measures the magnitude of the differences
between points on the scale but also taps the proportions in the differences, and is thus the most
powerful of the four scales. It is useful for multiplication and division. For example, 250 kgs is
twice as heavy as 125 kgs and 250 divided by 125 is 2.
In summary: The nominal scale separates individuals or items into categories eg. seniors and juniors.
The ordinal scale helps to rank order the separated items.
The interval scale helps to determine the extent to which the items are different from each
other. In other words, is the difference between item 1 and 2 the same as the difference
between 4 and 5?
The ratio scale helps to determine the proportion of differences among items since it has an
absolute zero origin, and is the most powerful of the four scales.
5.3 SOURCES OF MEASUREMENT DIFFERENCES
Error sources include:a) The Respondent
Temporary factors, for example, fatigue, anxiety, boredom, hunger, impatience, lack of
interest, mood swings - affect the ability to respond accurately and completely.
b) Situational Factors
Factors that put a strain on the interview eg. presence of a third person, fear of breach of
confidentiality.
c)

Measurement Strategy
Rewording, rephrasing, reordering questions.
Voice, tone, prompting with smiles/nods.
Problems in mechanical processing and failure to record full replies (illiterate).
Incorrect coding, careless tabulations, incorrect statistical calculation.

d)

The Instrument
A defective instrument can cause distortion.
Confusing and ambiguous questions.
Content items - instrument that does not explore the potentially important issues.
17

5.4 CHARACTERISTICS OF GOOD MEASUREMENT


EVALUATING MEASURES
Two main criteria for testing the goodness of measures are validity and reliability. Validity tests
how well an instrument developed measures the particular concept it is supposed to measure.
Reliability tests how accurately and consistently a measuring instrument measures whatever
concept it is measuring. In other words, validity is concerned with whether the right concept is
being measured, and reliability is concerned with stability and consistency in measurement.

18

CHAPTER 6 - SCALING DESIGN


6.1 RATING SCALES AND ATTITUDE SCALES COMMONLY USED
A) LIKERT SCALE
The Likert Scale utilises the anchor of strongly disagree, disagree, neither disagree nor agree,
agree, strongly agree. The respondents indicate the extent to which they agree or disagree to a
variety of statements (as shown below).
EXAMPLE
Indicate the extent to which you agree with the following statements as they relate to your job,
by marking a cross (X) against the appropriate, using the scale below.
1 - Strongly disagree
2 - Disagree
3 - Neither agree nor disagree
4 - Agree
5 - Strongly agree
___________________________________________________________________________
1
2
3
4
5
___________________________________________________________________________
The following opportunities offered by the job
are very important to me:
a. Interacting with others.
1
2
3
4
5
b. Using a number of different skills.
1
2
3
4
5
c. Completing a task from beginning to end.
1
2
3
4
5
d. Serving others.
1
2
3
4
5
e. Working independently.
1
2
3
4
5
____________________________________________________________________________

19

B) SEMANTIC DIFFERENTIAL
In this scale a number of bipolar attributes are indicated at the extremes of, usually, a sevenpoint scale and respondents indicate their attitudes toward a particular individual, object or
event on these attributes. The bipolar adjectives could represent such words as good --- bad;
strong --- weak.
EXAMPLE
------------------------------------Think of the person with whom you have had difficulty working. It does not have to be a
person you really dislike, just someone you must work with and with whom working is difficult.
Describe the person on each of the scales below by checking the number between each pair of
adjectives that best describes that person. Then transfer the numbers to the spaces at the right
and total them.
--------

Pleasant

.
7

.
6

.
5

.
4

.
3

.
. Unpleasant
2 1

Friendly

.
7

.
6

.
5

.
4

.
3

.
2

. Unfriendly
1

--------

Accepting

.
7

.
6

.
5

.
4

.
3

.
2

. Rejecting
1

--------

Relaxed

.
7

.
6

.
5

. .
4 3

.
2

. Tense
1

--------

Interesting

.
7

.
6

.
5

. .
4 3

.
2

. Boring
1

--------

Nasty

.
7

.
6

.
5

. .
4 3

.
2

. Nice
1

--------

__________________________________________________________________________

20

C) RANK ORDERING

EXAMPLE
Assume you are to judge a set of statements that will be used in developing a differential scale
for an employee attitude survey. Score each of the statements listed below from 1 to 10, with 10
being the most positive favourable statement and 1 the least favourable.
SCORE
______ The pay in this company is terrible.
______ I think the company goes outside to fill good jobs instead of promoting people
who are here.
______ The company is sincere in wanting to know what its clients think about it.
______ I think this company treats its employees better than any other company does.
______ I believe accidents will happen no matter what you do about them.
______ One can get ahead in this company if one tries.
______ The workers put as much over on the company as the company puts over on them.
______ My boss gives all the breaks to lodge and church friends.
______ Soldiering on the job is on the increase.
______ In the long run this company will "put it over" on you.

21

CHAPTER 7 - SAMPLING DESIGN


7.1 CONCEPTS IN SAMPLING

A population refers to the entire group of people, events or things of interest that the researcher
wishes to investigate. If the researcher is interested in investigating the saving habits of typists
in the plastics industry in South Africa, then all typists in the plastics industry throughout the
country will form the population.
An element is a single member of the population.
Generalisability refers to the scope of applicability of the research findings in one
organisational setting to other settings. For wider generalisability, the research sampling design
has to be logically developed and a number of other meticulous details in the data-collection
methods need to be followed.
A sample is a subset of the population. It comprises some members selected from the
population. If 200 typists are drawn from a population of 1000 typists, these 200 members form
the sample for the study. By studying these 200 members, the researcher would draw
conclusions about the entire population of the 1000 typists considered in the study. A sample is
thus, a subgroup or subset of the population. By studying the sample, the researcher would be
able to draw conclusions that would be generalisable to the population of interest.
A subject is a single member of the sample, just as an element is a single member of the
population.
Sampling is the process of selecting a sufficient number of elements from the population so that
by studying the sample, and understanding the properties of the characteristics of the sample
subjects (statistics), the properties or characteristics may be generalised to the population
elements (parameter).

7.2 TYPES OF SAMPLING


There are two major types of sampling designs: probability and non-probability sampling. In
probability sampling, the elements in the population have some known chance or probability of
being selected as sample subjects. In non-probability sampling, the elements do not have a
known or predetermined chance of being selected as subjects. Probability sampling designs are
used when the representativeness of the sample is of importance for purposes of wider
generalisability. When time or other factors rather than generalisability become critical, nonprobability sampling is generally used.

22

There are two types of probability sampling designs: the unrestricted or simple random
sampling and the restricted or complex probability sampling plans.
In the simple random sampling design, every element in the population has a known and
equal chance of being selected as a subject.
The complex probability plan consists of five different sampling designs: Cluster sampling - least expensive and least dependable.
Stratified random sampling - most efficient in the sense that for the same number of sample
subjects it offers more precise and detailed information.
Systematic sampling design - has the built-in hazard of possible systematic bias.
Area sampling - a form of cluster sampling.
Double sampling - takes place when a subgroup of a sample is used a second time to obtain
more information.
There are two main types of nonprobability sampling designs: convenience sampling and
purposive sampling.
Convenience sampling - least reliable in terms of generalisability but suitable when quick and
timely information is needed.
Purposive sampling - two categories: judgement and quota sampling.
Judgement sampling, though restricted in generalisability, may sometimes be the best
sampling design choice, especially when there is a limited population that can supply the
information needed.
Quota sampling, though not very reliable for generalisability, is often used for cost and time
considerations.
Snowball sampling

23

In summary:Simple random sampling is best fitted when generalisability of the findings to the whole
populations is the objective of the study.
Stratified random sampling is the most efficient and is a good choice when differentiated
information is needed regarding various strata within the population known to differ in their
parameters.
If the frame is large and a listing of the elements is conveniently available in one place (as in the
telephone directory, company payroll) then a systematic sampling procedure will offer the
advantages of ease and quickness in developing the sample.
Cluster sampling would be most useful when a heterogeneous group is to be studied at one
time. The unit costs of such a study are much lower compared to the other probability sampling
designs of simple or stratified random sampling or systematic sampling.
Area sampling design is constituted of geographical clusters eg. retail stores location plans,
advertising focused specifically on local populations, use an area sampling design to gather
information on the interests, attitudes, predispositions and behaviours of the local area people.
Area sampling is thus best suited when the goal of the research is confined to a particular
locality or area.
Double sampling is resorted to when more complete or comprehensive information needs to be
obtained from a group from which some information has already been collected. A subset of
the original sample is used for the purpose. This results in minimal additional expenditure for
further data gathering.
Convenience sampling is not generalisable at all and is not appropriate for 'scientific' research.
It may however, be used at times to get some 'quick' information and data obtained might offer
a 'feel' on some variable or variables of interest.
Judgement sampling is used where gathering 'specialised informed inputs' on the topic area
researched is vital and using any of the probability sampling design would not offer
opportunities to gain the specialised information. This method calls for special efforts to locate
and gain access to the individuals who do have the requisite information.
Quota sampling ensures certain groups are adequately represented in the study through the
assignment of a quota. Generally, the quota fixed for each subgroup is based on the total
numbers of each group in the population. However, since this is a nonprobability sampling
design, the results are not generalisable to the population. In a workplace and society that is
increasingly heterogeneous because of changing demographics, quota sampling can be expected
to be used more frequently in the future.

24

EXERCISE
1. What kinds of sampling designs would be used for the following:a) A study to get a quick idea of the medical acceptability of a new aspirin substitute which
cannot be dispensed over the counter without a prescription.
b) A study involving a sample of 325 students in a university where 2 000 students are enrolled.
c) An investigation of the career salience of professionals in the fields of medicine,
engineering, business and law.
d) The generalisability of the attitudes of managers from a sample of 184 to the total population
of 350 managers in the entire company.

25

CHAPTER 8 - SOME STRATEGIES OF INFORMATION SEARCH


Dialog offers bibliographies, corporate directories, technical reports, journal abstracts, and
complete texts of several journals and newspaper items.
ERIC gives access to the Psychological Abstracts and the Social Science Citation Index, in
addition to covering information from several federal agencies.
Sabinet
Ebscohost
HSRC Search - complete literature search on topic based on key words given. Fee.

ABSTRACTS:Personnel Management Abstracts is an index of articles that deal with the management of
people and organisational behaviour.
Psychological Abstracts summarises the literature in psychology, covering several hundred
journal reports, and other scientific documents.
Internet - This is an international network that enables one to search library catalogs,
information databases, new reports, government information and reference works.

JOURNALS:OB/PERSONNEL/HR MANAGEMENT
Academy of Management Journal
Academy of Management Review
Journal of Applied Behavioural Science
Journal of Applied Business Research
Journal of Applied Psychology
Journal of Social Psychology
Journal of Business Research
Journal of Communication
Journal of Management
Organisational Behaviour and Human Decision Processes
Harvard Business Review
Organisational Dynamics
Personnel Psychology
Psychological Bulletin
Industrial Relations
Training and Development Journal
MARKETING
26

Journal of Academy of Marketing Science


Journal of Advertising
Journal of Advertising Research
Journal of Consumer Research
Journal of Marketing
Journal of Marketing Research
Journal of Retailing
Marketing Science
Psychology and Marketing

JOURNALS EXIST IN THE FOLLOWING AREAS OF STUDY ALSO:STRATEGIC MANAGEMENT POLICY


PRODUCTION OPERATIONS MANAGEMENT
MANAGEMENT SCIENCE/OPERATIONS RESEARCH
MANAGEMENT INFORMATION SYSTEMS
RESEARCH AND DEVELOPMENT
ACCOUNTING
FINANCE

NB. Different libraries carry a different combination of journals. Therefore, make use of the
Inter-library loans section to order you articles.

27

CHAPTER 9 - SURVEY METHODS


9.1 PERSONAL INTERVIEWS
a) DESCRIPTION
People selected to be part of the sample are interviewed in person by a trained interviewer.

b) ADVANTAGES
Good co-operation from respondents.
Interviewer can answer questions about survey, probe for answers, use follow-up questions
and gather information by observation.
Special visual aids and scoring devices can be used.
Illiterate and semi-literate respondents can be used.
Interviewer can pre-screen respondents to ensure he/she fits the population profile.
CAPI - Computer-assisted personal interviewing:- Responses can be entered into a portable
microcomputer to reduce cost and error.

c) DISADVANTAGES

High cost.
Need for highly trained interviewers.
Longer duration needed in the field to collect data.
May be wide geographic dispersion.
Follow-up is labour intensive.
Not all respondents are available or accessible.
Some respondents are unwilling to talk to strangers in their homes.
Some neighbourhoods are difficult to visit.
Questions may be altered or respondents coached by interviewers.

28

9.2 TELEPHONIC INTERVIEWS


a) DESCRIPTION
People selected to be part of the sample are interviewed on the telephone by a trained
interviewer.

b) ADVANTAGES

Lower costs than personal interviews.


Expanded geographic coverage without dramatic increase in costs.
Uses fewer, more highly skilled interviewers.
Reduced interviewer bias.
Faster completion time.
Better access to hard-to-reach respondents through repeated callbacks.
CATI - Computer-assisted telephone interviewing:- Responses can be entered directly into a
computer file to reduce cost and error.

c) DISADVANTAGES

Response rate is lower than for personal interviews.


Higher costs of interviewing geographically dispersed sample.
Interview length must be limited.
Many phone numbers are unlisted or not working thereby making directory listings
unreliable.
Some target groups are not available by phone.
Illustrations and visual aids cannot be used.
Responses can be less complete.

29

9.3 MAIL SURVEY


a) DESCRIPTION
Questionnaires are mailed to the sample to be self-administered. A stamped return envelope is
generally included. Incentives may be used to increase response rate.
b) ADVANTAGES

Often lowest cost option.


Expanded geographic coverage without increase in costs.
Requires minimal staff.
Perceived as more anonymous.
Allows the respondents time to think about the questions.
Allows contact with otherwise inaccessible respondents.

c) DISADVANTAGES

Low response rate.


No interviewer intervention available (for probing or explanation)
Clarification cannot be sought.
Respondents cannot be observed at all.
Cannot be long/complex.
Accurate mailing lists needed.
Often respondents returning survey represent extremes of the population (skewed responses
and distributions).

TO SUM: COMPARISON OF THE THREE TECHNIQUES


Personal interviews are most expensive and time-consuming.
Telephone interviews are moderate in cost and the quickest option.
Mail Survey is the least expensive and require a longer data collection period.
If you have hard-to-reach, inaccessible subjects, then the best method is telephone
interviews.
If you require extensive questioning and probing, then the best method is personal
interviews.
Never use the mail interview method if you require quick data collection.
NB. One can use the mixed mode of two or more alternatives.
30

CHAPTER 10 - SURVEY INSTRUMENTS


10. QUESTIONNAIRE CONSTRUCTION
10.1 DEFINITION OF QUESTIONNAIRE
A questionnaire is a preformulated, written set of questions to which respondents record their
answers, usually within rather closely defined alternatives.

10.2 WHEN TO USE THE QUESTIONNAIRE


When the researcher knows precisely what information is needed.
When large numbers of people are to be reached in different geographical regions.
When groups of people can be assembled in conference rooms to whom questionnaires can
be administered and collected immediately.

10.3 ADVANTAGES OF QUESTIONNAIRES

Helps researcher to obtain data fairly easily.


Information from questionnaires is easily coded.
Benefits the scientific community if the measures are well validated and are reliable.
Often is a catharsis for respondents.

10.4 GUIDELINES FOR QUESTIONNAIRE DESIGN


Good questionnaire design principles focus on three areas:Word content:
Principles of measurement:
Getup:

wording of the questions.


how the variables are categorised, scaled and ordered when
general appearance of questionnaire.

31

PRINCIPLES OF WORDING

Appropriateness of the content of the questions.


How the questions are worded and the level of sophistication of language used.
Type of questions.
Form of questions.
The sequencing of questions.
Biographical data.

1. CONTENT AND PURPOSE OF THE QUESTION


The kind of questions to be asked will depend on the nature of the variable being investigated,
for example, subjective feelings or objective facts. For subjective feelings, tap on dimensions
and elements of the concept. For objective facts, ask single, direct questions that are ordinally
scaled.
2. THE LANGUAGE AND WORDING OF THE QUESTIONNAIRE
Language in the questionnaire should approximate the level of understanding of the
respondents. The choice of words would depend on:

educational level of the respondents sampled.


the usage of terms and idioms in the culture.
frames of reference of the respondents.

If some questions are not understood or are interpreted differently by the respondents, the
researcher will be getting the wrong answers to the questions and the responses will be biased.
3. TYPE OF QUESTIONS
Open-ended
Type of questions
Closed-ended question
Open-ended questions allow people to answer questions in any way they choose, for example,
State 5 things that are challenging in a job.
Closed-ended questions ask the respondent to make choices amongst a set of alternatives, for
example, given 15 challenging aspects in a job, each person is asked to rank order the first five
in order of importance.
Closed questions: advantages
enables quick decisions
enables easy coding for later analysis.
32

Closed questions: disadvantages


confining - eliminate problem by providing space at the end of the questionnaire to allow
respondent to comment of any aspect left out.

4. FORM OF THE QUESTION


Positively worded questions
Form of questions
Negatively worded questions
Instead of phrasing all questions positively, it is advisable to include some negatively worded
questions as well so that it minimizes the tendency in respondents to mechanically circle the
points toward one end of the scale. This helps to ensure the respondents are aware. However,
avoid double-negatives and excessive use of not and only, for example, say 'coming to work is
not great fun' instead of 'Not coming to work is greater fun than coming to work'.
Do not include:Double-barrelled questions:
Two questions in one, for example, . Do you think there is a good market for no-name brands
and would it reap high profits?
Ambiguous questions:
To what extent would you say you are happy?
? happy at home, at work, or in general.
Recall-dependent questions:
Asking respondent to recall can be dangerous and result in incorrect answers. For example,
How long he worked in a specific department? Rather get information from personnel records.
Leading questions:
Questions should not be phrased such that they lead the respondent to give the responses that
the researcher would like to elicit or may come across as wanting to elicit.
For example, "Don't you think that as a result of escalating costs of living, employees
Rather ask:- "To what extent do you think that employees should be given pay raises?'
Strongly agree (5) to strongly Disagree (1).
Loaded questions:
Questions should not be emotionally charged.
For example, "To what extent do you think that management would be vindictive if the
'strike' and 'vindictive' are emotionally charged.
If the intention is to determine whether employees are in favour of a strike or the extent to
which they fear strong reactions, then questions should be phrased such.
33

Social desirability:
Questions should not be phrased such that they elicit socially desirable responses.
For example, "Do you think that older people should be laid off?".
Rather ask:- "There are advantages and disadvantages to retaining senior citizens in the
Length of questions:
Ask short questions. Questions should never exceed more than 20 words.
5. SEQUENCE OF QUESTIONS
The sequence of questions in the questionnaire should be such that respondents are led from
questions that are of a general nature to those that are more specific and from questions that are
relatively easy to answer to those that are progressively more difficult. This is called the 'funnel
approach'. This approach facilitates the easy and smooth progress of the respondent through the
items in the questionnaire.
Sequence of questions - positive and negative.
For example, 'I am allowed to interact with colleagues at work.'
'I am not allowed to interact with colleagues at work.'
This is not necessary and insults the respondent.
6. BIOGRAPHICAL/PERSONAL DATA
For example, eg, marital status, income.
Avoid asking the name but if it is necessary, have a separate list of names with each person
have a code or number.
At the beginning or end?
Beginning - sense of commitedness/psychologically identify with questionnaire.
End - convinced of the genuineness of the inquiry.
Better to provide a range of response options.
For example, Age - Under 20, 21 - 30, 31 - 40, 41 - 50, 51 - 60, Over 60.
PRINCIPLES OF MEASUREMENT
Principles of measurement are followed to ensure that the data collected are appropriate to test
the objectives/hypothesis. The principles encompass: the scales and scaling techniques used in measuring concepts.
the assessment of reliability and validity of measures used.

34

GENERAL APPEARANCE OF THE QUESTIONNAIRE


Must be attractive and neat, with: appropriate introduction.
clear instructions.
well array-set of alternatives with easy response alternatives.
Appropriate introduction:
establish rapport by stating purpose and aim of study.
assure confidentiality.
thank respondents.
Clear instructions:
Organise questions logically and neatly.
Open-ended question at end:
To give respondent the opportunity to express feelings about any aspect not covered in the
questionnaire.

35

CHAPTER 11 - OBSERVATION
11.1 BACKGROUND
The collection of data by means of observation is often haphazard. However, in research,
observation can be a scientific inquiry when:a)
b)
c)
d)

it is specifically designated to answer a research question.


it is systematically planned and executed.
it uses proper controls.
it provides a reliable and valid account of what happened.

Besides collecting data visually, observation involves listening, reading, smelling and touching.
Observation includes monitoring behavioural and non-behavioural activities and conditions and
is thus good when studying animal behaviour and children.

11.2 THE NATURE OF OBSERVATION


a) DESCRIPTION
Gathering data without asking questions of the respondents by observing people in their natural
work environment or in the laboratory setting and recording their behaviours. The research can
play one of two roles while gathering field observation data: non-participant observer and
participant observer.
Non-participant observer - The researcher can collect the data in the role of a pure researcher
without trying to become an integral part of the organisation system.
Participant observer - The researcher enters the organisation or the research setting actually
becoming a part of the work team.

b) ADVANTAGES
This is the only method available to gather certain types of information, for example,
children, animals, illiterate people. Secures information that cannot be derived from
experiments or surveys.
Can collect the original data at the time they occur and need not depend upon reports and
thus, overcomes the problems of filtering information, incomplete reporting and forgetting.
Can secure information that most participants would ignore because it is so common and
expected or because it is considered irrelevant.
It also can capture the whole event as it occurs in its natural environment.
Subjects seem to accept an observational intrusion better than questioning.
36

c) DISADVANTAGES
The observer normally must be at the scene of the event when it takes place and it is often
impossible to predict where and when the event will occur.
Observation is a slow and expensive process that requires either human observers or costly
surveillance equipment.
The most reliable results are restricted to information that can be learned by overt action or
surface indicators. To go below the surface, the observer must make inferences.
The research environment is more likely suited to subjective assessment and recording of
data than controls and quantification of events.
Observation is limited as a way to learn of the past.

NB. Observation, especially participant observation, contributes to a reactivity response where


subjects alter their behaviour in response to the researcher. Thus, some innovative
observational procedures that are non-reactive have been identified. These unobtrusive
measures encourage creative and imaginative forms of indirect observation. They include
measures involving indirect observation based on physical traces that include erosion (measures
of wear), for example, replacement of floor tiles in a shopping mall, wear and tear on library
books and accretion (measures of deposit), for example, service of car radios indicate
advertising listnership, family trash examination indicates usage of brands of products.

37

CHAPTER 12 - QUALITATIVE ANALYSIS


12.1 INTRODUCTION
Research in which the results cannot be statistically analysed.
Positivist approach versus phenomenological approach.
Qualitative research as an interactive process.
Qualitative versus quantitative data.
Grounded theory approach.
Secondary data collection.
Some qualitative methods:
experience survey - questioning people who have encountered a similar situation or who are
knowledgeable about the situation being faced.
case study is an attempt to apply what is learned from a detailed, in-depth examination of
other situations to the situation at hand.
focus groups consists of 8 to 12 persons who are led through an unstructured discussion of a
topic by a moderator.
brainstorming/synetics.
Projective techniques.
Action Research
Triangulation

38

12.2 POSITIVIST AND PHENOMENOLOGICAL APPROACHES


The Positivist approach to research - scientific research.
Distinguishing features:-

deductive
explains causal relationships between variables.
normally uses quantitative data.
employs controls to allow the testing of hypotheses.
highly structured methodology used to facilitate replication.

Easterby-Smith et al. (1991: 23) list eight features of positivism:

Independence

Value-freedom

Causality

Hypothetico-deductive

Operationalisation

Reductionism

Generalisation

Cross-sectional analysis

The Phenomenological approach

The Phenomenological approach is based on the way people experience social phenomena
in the world in which they live.

Characteristics:-

focus on the meanings that research subjects attach to social phenomena;


what is happening and why it is happening.
Context in which events take place.

39

POSITIVISM

ADVANTAGES

Economical collection of large amount of data.


Clear theoretical focus for the research at the outset.
Greater opportunity for researcher to retain control of research process.
Easily comparable data.

DISADVANTAGES

Inflexible - direction often cannot be changed once data collection has started.
Weak at understanding social processes.
Often does not discover the meanings people attach to social phenomena.
PHENOMENOLOGY

ADVANTAGES

Facilitates understanding of how and why.


Enables researcher to be alive to changes which occur during the research process.
Good at understanding social processes.

DISADVANTAGES

Data collection can be time consuming.


Data analysis is difficult.
Researcher has to live with the uncertainty that clear patterns may not emerge.
Generally perceived as less credible by 'non-researchers'.

12.3 QUALITATIVE RESEARCH AS AN INTERACTIVE PROCESS


There are a number of reasons for describing qualitative analysis as an 'interactive process'.
Analysis needs to occur during the collection of data as well as after it. This helps to shape the
direction of data collection, especially where you are following a grounded theory approach.
The interactive nature of data collection and analysis allows you to recognise important themes,
patterns and relationships as you collect data. As a result you will be able to re-categorise your
existing data to see if these themes, patterns and relationships are present in the cases where you
have already collected data. In addition, you will be able to adjust your future data collection
approach to see if they exist in cases where you intend to conduct your research.

40

12.4

QUALITATIVE VERSUS QUANTITATIVE DATA

Quantitative Data
Based on meanings derived from numbers.
Collection results in numerical and standardised data.
Analysis conducted through the use of diagrams and statistics.

Qualitative Data
Based on meanings expressed through words.
Collection results in non-standardised data requiring classification into categories.
Analysis conducted through the use of conceptualisation.

12.5 GROUNDED THEORY APPROACH


Reasons for adopting a grounded theory approach to your research project and the analysis of
the data which are produced.
Commencing an exploratory project seeking to generate a direction for further work.
The scope of your research may be constrained by adopting restrictive theoretical
propositions which do not reflect you participants' views and experience (Bryman 1988).
Finally, the theory's generalisability may also be tested in other contexts (Glaser and Strauss
1967; Strauss and Corbin 1990).
However, you should not draw the conclusion that you may use a grounded approach as a
means of avoiding a proper level of preparation before commencing your research project.
Grounded theorists do not jump into a subject area without a competent level of knowledge
about that area. Their research will commence with a clearly defined purpose, even though this
may be altered by the nature of the data which they collect. You should seek to compare your
grounded explanations to existing theory once these have emerged. The use of a grounded
theory approach will also involve you in a lengthy period of data collection and concurrent
analysis in order to analyse a theme adequately or derive a well-grounded theory. Strauss and
Corbin (1990) suggest that this process is likely to take months to complete.
The procedures related to a grounded approach are therefore designed to develop an explanation
or a theory. They are not designed to test existing theory (Strauss and Corbin 1990). In
comparison to the strategy of testing existing theory, the advocates of grounded theory lay down
fairly precise procedures to be adopted in relation to each of the stages of the qualitative analysis
process. Without paying particular attention to the nature of the procedures outlined for
grounded theory, you may not produce a research report which is sufficiently rigorous to
substantiate the analysis or theory which you are seeking to advance. In the grounded theory
approach of Strauss and Corbin (1990), the disaggregation of data into units is called open
coding; the process of recognising relationships between categories is referred to as axial
coding; and the integration of categories to produce a theory is labelled selective coding.

41

12.6

SECONDARY DATA

When considering how to answer their research question(s) or meet their objectives, few
researchers consider the possibility of re-analysing data that have already been collected for
some other purpose. Such data are known as secondary data. Most researchers automatically
think in terms of collecting new (primary) data specifically for that purpose. Yet, despite this,
secondary data can provide a useful source from which to answer, or begin to answer, your
research question(s).
Secondary data include both raw data and published summaries.
Most organisations collect and store a variety of data to support their operations, for
example, payroll details, copies of letters, minutes of meetings and accounts of sales of
goods and services.
Daily newspapers contain a wealth of data, including reports about takeover bids, and
companies' share prices.
Government departments undertake surveys and publish official statistics covering social,
demographic and economic topics.
Consumer research organisations collect data which are used subsequently by different
clients.
Trade organisations survey their members on topics such as sales which are subsequently
aggregated and published.
Some of these data, in particular documents such as company minutes, are only available
from the organisations which produce them and so access will need to be negotiated.
Others, including government surveys such as the population census are widely available in
published form and on CD-ROM in university libraries. An increasing variety have been
deposited in and are available from data archives. In addition a rapidly increasing number of
on-line computer databases can be accessed via the Internet.
Three main subgroups of secondary data are documentary data, survey-based data and those
compiled from multiple sources (Saunders, Lewis and Thornhill 1997: 158 - 159).
ADVANTAGES
May have fewer resource requirements.
Unobtrusive.
Longitudinal studies may be feasible.
Can provide comparative and contextual data.
Can result in unforeseen discoveries.

42

DISADVANTAGES
May have been collected for a purpose which does not match your need.
Access may be difficult or costly.
Aggregations and definitions may be unsuitable.
Initial purpose may affect how data is presented.

12.7

QUALITATIVE RESEARCH METHODS

12.7.1 Experience Survey


Simply refers to questioning people who have encountered a similar situation or who are
knowledgeable about the situation being faced.
12.7.2 Case Study
An attempt to apply what is learned from a detailed, in-depth examination of other situations
to the situation at hand.
When conducting case studies: The case should be defined or demarcated, that is, its boundaries should be determined.
During the course of the study the boundaries may be adjusted, if found necessary.
Whichever technique is used for purposes of data collection, the concern is not merely with
a description of what is being observed, but to search, in an inductive fashion for recurring
patterns and consistent regularities.
In discerning these patterns triangulation is often used, for example, three different
approaches being adopted such as recordings of conversations, semi-structured interviews
and newspaper reports.
12.7.3 Focus Groups/Brainstorming

Useful, especially in the exploratory phase.


They generate ideas on key variables to be included in the study.
Relatively inexpensive - provides rather dependable information in a short space of time.
Consists of 8 to 10 members randomly chosen with a moderator leading the discussion on a
particular topic.
Focus groups and brainstorming are useful during questionnaire construction as well as in
order to ensure validity of the questions.
Brainstorming is particularly useful because of the multiplicity of ideas.
43

12.7.4 Projective Techniques


Certain thoughts and ideas that cannot be easily verbalised or that remain on unconscious
levels in the respondent's mind that usually be brought to the surface through motivational
research.
Trained professions - apply different probing techniques.
Familiar techniques include: Word associations, for example, Work
Sentence completion, for example, Work is .
Thematic Apperception Test - story
Inkblot tests - coloured inkblots

12.8 ACTION RESEARCH


Performed with a view of finding a solution to a particular problem is a specific, applied
setting.
Action research is not aimed at testing or development of a theory.
Action research uses a design which may continually be changed and adapted in reaction to
information and results obtained during the course of the research project.
Action research places a high premium on involving all participants in each of the above
phases.
External validity does not enjoy a high priority in action research. Internal validity is high.

12.9 TRIAGULATION

Three data collection methods.


Used to verify information collection from various data collection sources.
Increases validity of results.
Identify three appropriate methods of data collection, explain the methods and describe how
you would use it to collect the data and from who the data will be collected.

44

CHAPTER 13 - DATA PREPARATION AND PRELIMINARY


ANALYSIS
13.1 GETTING DATA READY FOR ANALYSIS

After the questionnaire data have been obtained, the data need to be edited. The blank
responses, if any, have to be handled in some way, the data then have to be coded, categorised
and keyed in, and decisions need to be made as to how they will be computer analysed.

EDITING DATA
Especially when the data come from questionnaires containing open-ended questions, the data
have often to be edited. In other words, information that may have been written by the
respondent in a hurry must now be clearly deciphered so that all of it can be coded
systematically.

HANDLING BLANK RESPONSES


Not all respondents answer every item in the questionnaire. Questions may have been left blank
because the respondent did not understand the question, was not willing to answer, or was
simply indifferent to responding to the entire questionnaire. It the latter is the case, the
respondent may have left many of the items blank. If more than 25 % of the questionnaire has
been left blank, it may be a good idea to throw out the questionnaire and not include it in the
data set for analysis. If however, only two or three items are left blank in a questionnaire with
about 30 or more items, a decision needs to be made as to how these blank responses will be
handled. There are several ways of handling blank responses. A common way of dealing with
this problem is wither to give the midpoint in the scale as the value or to ignore the particular
item during the analysis.

45

CODING
The next step is to code the responses. the easiest way to illustrate a coding scheme is through
an example. The responses for the demographic variables can be coded from 1 to 5 for the
variable age and 1 or 2 for sex, depending on which box in the columns was checked by the
respondent, as follows:-

AGE (in years)


Under 25

25 - 34

35 - 44

45 - 54

55 and over

SEX
Male

Female

CATEGORISATION
At this point it is useful to categorise the variables such that the several items measuring the
concept are all grouped together for feeding into the computer later. Responses to some of the
negatively worded questions have to be reversed so that all answers are in the same direction.

KEYING DATA
The raw data will have to be manually keyed into the computer. Mistakes can occur at this
stage as well. To ensure the correct figures are entered into the right columns, initially every
tenth record or so can be spot-checked for accuracy. if mistakes are detected, it would then
become necessary for one person to read each of the figures on the code sheet as another checks
for the accuracy of the keyed-in data. Because the results obtained from data analyses can only
be as good as the raw data entered, care should be taken when the data are collected.

46

13.2 DATA ANALYSIS: PROGRAMMING


Results may be generated using one of the following programmes:*
*
*
*

SPSSX
SPSS
SYSTAT
SYMSTAT

PS. Validity and Reliability

13.3 DATA ANALYSIS AND INTERPRETATION


We may distinguish between two types of statistics:Descriptive statistics - statistics that describe the phenomena of interest.
Inferential statistics - statistical results that let us draw inferences from a sample to the
Both descriptive and inferential statistics can be obtained through computer programs that are
already designed to produce results for various types of data.

DESCRIPTIVE STATISTICS
FREQUENCIES
Frequencies refer to the number of times various subcategories of a certain phenomenon occur,
from which the percentage and the cumulative percentage of the occurrence of the subcategories
can be easily obtained. This information can also be presented in the form of a histogram or a
bar chart. Frequencies are generally obtained for nominal variables such as gender, the
departmental unit in which one works and type of organisation.
MEASURE OF CENTRAL TENDENCY AND DISPERSION
There are three measure of central tendency: mean, median and mode. The measures of
dispersion include the range, variance and standard deviation.
The mean is the average value.
The median is the central item in a group of observations when they are arranged in ascending
or descending order.
The mode is the score value occurring most frequently.
47

The range refers to the extreme values in a set of observations. It is the highest minus the
lowest score value.
The variance is calculated by subtracting each of the observations in the data set from the
mean, taking the square of this difference and dividing the total of these by the number of
observations. For example, in set 2, 8, 12 and 10:Mean (8) = 8
N
= 4
Variance =

(2 - 8)2 + (8 - 8)2 + (12 - 8)2 + (10 - 8)2


4
36 + 0 + 16 + 4
4

14

The standard deviation is calculated simply as the square root of the variance.

TABULAR AND GRAPHICAL PRESENTATION OF DATA


(see Transparencies of Bar, line and pie charts, histograms and frequency polygons)

48

CHAPTER 14
HYPOTHESIS TESTING AND INFERENTIAL STATISTICS
14.1 INTRODUCTION TO HYPOTHESIS TESTING
An hypothesis is a logically conjectured relationship between two or more variables
expressed in the form of testable statements. The null hypothesis is a proposition that states
a definitive, exact relationship between two variables. The null hypothesis is expressed as no
relationship between two variables or no significant difference between two groups. It is
represented by Ho.
The alternate hypothesis is the opposite of the null hypothesis and is thus, a statement
expressing a relationship between two variables or indicating differences between groups. It
is represented by H1.
Ho : UM = UF
or

Ho : UM - UF = 0

H1 : UM UF

If p < 0,05 for a directional or one-tailed test, reject the null hypothesis and accept the
alternate hypothesis.
If p < 0,025 for a non-directional or two-tailed test, reject the null hypothesis and accept the
alternate hypothesis.

SHOW ON NORMAL DISTRIBUTION CURVE (0,025 AND 0,5).


For nominal and ordinal data, use non-parametric statistics.
For interval and ratio data, use parametric statistics.

14.2 TYPE I AND TYPE II ERRORS


Type I error occurs when a true null hypothesis is rejected, that is, we reject the null
hypothesis when it is in fact true.
Type II error occurs when one fails to reject a false null hypothesis, that is, accept the null
hypothesis when it is in fact false.
49

14.3 INFERENTIAL STATISTICS


14.3.1 CORRELATION
Is one variable related to another?
Want to see the nature, direction and significance of the relationship between two
variables.
Pearson correlation matrix will indicate the direction, strength and significance of the
bivariate relationship among the variables in the study.
Used when we have two interval or ratio variables.
Is correlation significant or not? This is determined by looking at the level of significance.
For example, at the 5 % level of significance, we can say that 95 times out of 100, we can
be sure that there is a significant correlation between the two variables and there is only a
5 % chance that the relationship does not exist.
Range of correlation: -1,0 to +1,0
The DECISION RULE will be to accept the alternate hypothesis and reject the null
hypothesis is p < 0,05 and reject the alternate hypothesis and accept the null hypothesis if
p > 0,05.
For example, if r = -0,1735 and p = 0,083.
This means that p > 0,05. Hence we accept the null hypothesis and reject the alternate
hypothesis.
By using Pearson r we would not know which variable causes which, but we would know
that the two variables are associated with each other.
Hence, Pearson r gives: the strength of the relationship
the direction of the relationship

50

14.3.2 CHI-SQUARE TEST


Is there a relationship between two nominal variables or are they independent of each
other?
For example, Is viewing TV advertisements of a product (YES/NO) related to the buying
of that product by individuals (BUY/DO NOT BUY)?
Hence, we use chi-square which is a nonparametric test.
Chi-square works on the principle of expected and actual frequencies and degrees of
freedom and the x2 statistic is obtained by:2

x2 =

(Oi - E)
________ ;
Ei

where:x2 is the chi-square statistic


Oi is the observed frequency in the ith cell
Ei is the expected frequency of the ith cell
Computers give the x2 statistic for any given set of data, with the level of significance of
the x2 statistic indicated on the output.
The DECISION RULE is to retain the null hypothesis and reject the alternate hypothesis
if p > 0,05 and accept the alternate hypothesis and reject the null hypothesis if p < 0,05.
An alternative test: Fisher Exact Probability Test

51

14.3.3 THE T-TEST


We may be interested in knowing whether two groups are different from each other on a
particular variable of interest.
For example, would men and women indicate their preference for the introduction of
flexitime at the workplace to the same extent or would they differ in their need for it?
To find answers, a t-test is used to see if there are any significant differences in the means
for two groups in the variable of interest.
This means that a nominal variable that is split into two subgroups is tested to see if
there is a significant mean difference between them on a dependent variable which is
measured on an interval or a ratio scale.
The t-test takes into consideration the means and standard deviations of the two groups on
the variable and examines if the numerical differences in the means is significantly
different from 0 as postulated in our null hypothesis.
Hence, the t-test may be used with two independent samples or two different groups on a
variable.
After calculating the t value, we can refer to the t tables to see if the t value is significant
or not. The computer does all this work and gives both the t values and the associated
probability.
The DECISION RULE would be to accept the alternate hypothesis if the t value is
significant at the p < 0,05 level. If the results are significant at the 0,05 level when the
hypothesis is stated directionally, we would conclude that there is a significant mean
difference between the two groups as hypothesized.

52

14.3.4 ANALYSIS OF VARIANCE (ANOVA)


The t-test tells us whether or not there is a significant mean difference in a dependent
variable between two groups.
ANOVA indicates whether or not there is a significant mean difference in a dependent
variable between two or more groups.
Are the means of the various groups significantly different from each other or not?
This is the F-statistic.
ANOVA looks at total variation = between-group variation + within-group variation.
DF - degrees of freedom.
ANOVA results only show differences in the means but there is no way of knowing from
the ANOVA results alone where the differences lie.
However, tests that show this include:

Scheffe's Test
Duncan Multiple Range Test
Turkey's Test
Student-Newman-Keul's Test

The DECISION RULE applied is that, if the F statistic obtained is significant at the 0,05
level or better (as can be seen from the F tables), accept the alternate hypothesis that there
are significant mean differences among the groups. If the F statistic is not significant,
retain the null and reject the alternate hypothesis.

53

14.3.5 MULTIPLE REGRESSION


Correlation gives the relationship between one dependent and one independent
variable.
What happens when we have one dependent variable and several independent
variables (eg. four)?
The regressed effect of each is called the multiple r or multiple correlation.
R2 is the amount of variance explained in the dependent variable by the other (4)
independent variables.
More than one predictor are jointly regressed against the criterion variable to give us the
multiple regression analysis.
The DECISION RULE : If R2 = 0,63 and the F-value = 25,56 and p < 0,005 then we can say that 63 % of the
variance has been significant explained by the set of predictors.

14.4 CRONBACH'S ALPHA


We can check the reliability of measures using Cronbach's Coefficient Alpha.
Reliabilities less than 0,60 are generally considered to be poor, those in the 0,7 range are
considered to be acceptable, and those over 0,8 are good.
The closer the reliability coefficient gets to 1,0 the better.

54

ANNEXURE 1
POPULATION TO SAMPLE SIZE TABLE

N
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
110
120
130
140
150
160
170
180
190
200
210

S
10
14
19
24
28
32
36
40
44
48
52
56
59
63
66
70
73
76
80
86
92
97
103
108
113
118
123
127
132
136

N
220
230
240
250
260
270
280
290
300
320
340
360
380
400
420
440
460
480
500
550
600
650
700
750
800
850
900
950
1000
1100

S
140
144
148
152
155
159
162
165
169
175
181
186
191
196
201
205
210
214
217
226
234
242
248
254
260
265
269
274
278
285

N
1200
1300
1400
1500
1600
1700
1800
1900
2000
2200
2400
2600
2800
3000
3500
4000
4500
5000
6000
7000
8000
9000
10 000
15 000
20 000
30 000
40 000
50 000
75 000
1 000 000

S
291
297
302
306
310
313
317
320
322
327
331
335
338
341
346
351
354
357
361
364
367
368
370
375
377
379
380
381
382
384

N is population size
S is sample size
Sekaran, U. (1992). Research Methods for Business: A Skill-Building Approach. 2nd
Edition. New York: John Wiley. 253.

55

ANNEXURE 2
GUIDELINES FOR CONSTRUCTION OF A QUESTIONNAIRE

56

S-ar putea să vă placă și