Table Of Contents
1:0 Introduction
1.1.
What Research is
1.2 Social
research a science:
1.3 Characteristics
of scientific social research
1.4 Stages
taken in scientific social research:
1.5 The
goals of scientific social research:
1.6 Basic
concepts in research:
1.7 Possible
outcomes of hypotheses
1.8 The
characteristics of a usable hypothesis
2.0 Data Collection Methods
2.1
Introduction
2.2.
Non-survey research
Focus group discussions
Key Informant Interviews
Observations
i.
Non-participant observation
ii.
Observation with some participation
iii.
Participant observation
A
Case Study
2.3 Quantitative
research (survey research)
2.31 Personal
interviews:
Formal
Interviews
Informal
Interviews
What to consider for a successful
interview
Accessibility
Cognition
Motivation
What is expected of the interviewer:
Locating sample members.
Obtaining
the interview:
The
interviews:
Writing down (recording) responses:
Advantages of interview as a method of
data collection
Problems met in using interview as a
method of data collection
Response errors
How to reduce response errors
Self administered questionnaires:
Mail Questionnaire
Home/office
Advantages of using mail Questionnaire:
Its Limitations
3.0
Sampling
3.1Some Concepts:
Stages taken in sampling
Why sampling is
necessary
Sampling error
Non-sampling error
Sampling techniques
3.2 Random sampling techniques
·
Simple Random Sampling
·
Systematic Sampling
·
Stratified Sampling
·
Cluster sampling
Area Sampling
3.3 Non-random/probability sampling
·
Convenience Sampling
·
Purposive sampling
·
Network sampling
·
Quota Sampling
·
Dimensional sampling
4.0 Questionnaire design
Set of Principles for Proper questionnaire
Design:
The design of non-survey research
instructions:
5.0 Data Processing
o Editing
o Coding
o Tabulation
And Analysis Of Data
6 Measurement
Levels Of Measurement
·
Nominal Measurement
·
Ordinal Measurement
·
Interval Measurement
·
Ratio Measurement
7 Evaluation Of Research
8. The Use Of Secondary Data
9. Dissertation Writing
·
Formulation Of A Research Problem
·
Formulation of a Research Design (overall
approach to data collections)
·
Method of data Collection (Personal
Interviews, mail questionnaire)
·
Formulation/Design of Research Instruments
·
Deciding on the sample size
·
Data Collection
·
Analysis and Interpretation of Data
Collected
·
Preparation and presentation of Research
Report
1:0 Introduction
1.2.
What Research is:
It is taking another step forward is
searching for the unknown. Someone else has searched and explained some
phenomenon, and you are adding on what is already known.
Since
time immemorial, Scientists have been searching for different kinds of medicine
for different types of illness; for a better drug, more search has to be
carried out.
Since
society is dynamic, a consistent and persistent search for explanations as to
why phenomena happen the way they do is not only necessary but a must.
Research
therefore, is a consistent and organised process of collecting and analysing
information for a specifically defined purpose.
For
an academic(ian), a community and a business entrepreneur; the purposes could
be to know more, to solve a problem and to plan ahead; respectively.
A
good research is systematic, well planned, organised and has specific goals. It
involves observing, theorising, experimenting to test a theory or set of
theories, drawing conclusions and reporting results (Kakooza:2000).
1.2
Social
Research A Science
Social Research is a science since it
involves gathering evidence and applying reason. It finds out why, when, who
and what; about things in a systematic way. It uses theory[1]
and facts[2]
to understand our environment.
1.3
Characteristics Of Scientific Social
Research
i)
Its
results are based on observations that can be verified and are on reason. The
results are not based on speculation. Scientific Social Research is therefore empirical.
ii)
It makes an attempt to analyse and explain
observations in abstract terms in order to understand causal relations existing
between them; that is Scientific Social Research is theoretical.
iii)
What is discovered is added on earlier
discoveries and knowledge. Data keeps adding to the pool of information. It is
therefore cumulative.
iv)
Research comes up with conclusions that
are independent of personal biases of the researcher, and therefore scientific
social research is objective.
1.4
Stages Taken In Scientific Social Research
i)
Identifying the problem:
The researcher begins by identifying the prevailing problem. Taking an example
of the prevailing poverty in Uganda, or abuse of young girls. One must first
identify that there is a problem in order to do research.
ii)
Review of related literature:
This is a necessary and important step. After the problem has been identified,
the researcher must review the literature that is related to the subject under
study. This helps the researcher to know what has already been studied and
identify the gaps in which he/she can do the research without duplication of
findings.
iii)
Research Design formulation:
After identifying the prevailing problem, the methods of approach are then
designed. This is determined by a multiple of factors of which there is the
population structure, the population number, agency of the need for results,
availability of funds etc. These could either be qualitative or quantitative
design.
A qualitative design:
This design promotes greater understanding of not just the way things are, but
also why they are the way they are. Through intensive and extensive
observation, interviews and discussions, the qualitative researcher seeks to
derive and describe findings that promote greater understanding of how and
why people behave the way they do. It
explains and gains insight and understanding of phenomena through intensive
collection of narrative data (Martin E. Amin, 2005).
In this approach, data collected is
subjective and the main measurement tool for collecting data is the
investigator him/herself (Martin E. Amin, 2005), therefore a need for an
intensive training and practice in the methods to be used if one is to conduct
an effective research. Its data is basically descriptive in nature.
A quantitative design:
On the other hand however, this design deals with numbers. Quantitative
research involves the collection of numerical data in order to explain, predict
and control phenomena of interest, data analysis being more statistical (Martin
E. Amin, 2005). It involves collecting data in order to test hypotheses or
answer questions concerning the current status of the subject of the
study.
iv) Data Collection:
This is always determined by the design and therefore the factors that
determine the design. The methods of data collection include: -- Personal
interviews
- Mail
Questionnaires
- Observations
etc.
v) Processing of the Data:
This involves checking and removing all mistakes (editing), categorising
(coding) and analysing the data. Editing should be done within the same day of
collection.
vi) Interpreting the results and writing the report.
After the data has been analysed, it is interpreted and a report is made. This
is the final stage. In case the researcher is sent by some agency, this is what
he/she is expected to hand in.
The Goals/purpose Of Scientific Social
Research; It must specifically be noted that
society is dynamic; Leaders, Policy makers and other concerned persons cannot
rely on previous knowledge but must keep doing research. As an example, low
Life Expectancy in Uganda for years was a result of deaths due to malaria
fever; but currently, deaths due to HIV/AIDS could have overtaken deaths due to
malaria fever and therefore a need to find to find out whether HIV/AIDS is the
number one current killer disease.
-
The general goals therefore are:-
i)
to measure and describe a particular
phenomena
ii)
to predict phenomena
iii)
to control the magnitude or occurrence of
phenomena on the basis of the knowledge of the conditions under which the
situation occurs.
Scientific Social Research follows two
criteria in measuring and describing phenomena.
(i)
All aspects, areas or issues that require
the attention of the researcher is identified and a place for them is found on
the research agenda i.e Exhaustiveness. If for example; of the five
courses offered at Uganda Christian University (2001),
(LLB, BBA, BSWSA, M-COM and EDUC),if a questionnaire
is designed which does not consider M.Com, then the research is not exhaustive.
Even if all courses are identified but particular aspects that make up the
course are left ie some subjects that make up the course, then still the
investigation is not exhaustive.
(ii)
Each particular element of interest must
not be ambiguously place in more than one location ie one cannot be both guilty
and innocent, Black and coloured etc.
When events are measured, the instruments
used must be both valid and reliable. Valid in that the instrument used must be
with the capacity of measuring what it intends to measure i.e Thermometer for
temperature levels. The instrument must also be reliable by showing consistence
in giving the same information always ie the same weight in kilograms for the
same object whichever number of times its measured. In this case of
reliability, let us take an illustration for clarity; suppose a non-resident
student comes late to a lecture and the lecturer ask the student that, “where do
you come from?” The student may interprete the question either to mean where he resides or the home
area. To avoid this ambiguity, one may ask, “where do you reside?”
1.5
BASIC CONCEPTS IN RESEARCH:
i)
A variable:
This is a measurable event with the capacity to change ie level of income, age,
level of education etc.
ii)
A constant:
A measurable event, which does not change. eg sex like male of female.
iii)
A proposition:
This is a statement of relationship between variables. One which discusses a
single variable is univariate, two a bivariate and a multivariate for three or
more variables. The example of each respectively could be; He is a drug addict,
He is a drug addict and he stay in a slum, He is poor, stays in a slum a
drug addict.
iv)
A
Hypothesis: This is a tentative statement formulated for purposes of
empirical testing. In a hypothesis, there are always two variables, the cause
and the outcome (effect). The cause is
always independent of other factors, therefore an independent variable but the
outcome is always dependent on the cause, therefore a dependent variable. If we
take an example of level of Income and standards of living, one can confidently
say that the level of income directly affects the standards of living of an
individual.
The dependent variable can also be called
the predicted variable while the cause, the predictor variable.
1.6
POSSIBLE OUTCOMES OF HYPOTHESES
The essence of formulating a hypothesis is
to verify or reject the ideas in the mind of the researcher. A hypothesis can
therefore be either rejected or approved. Hypotheses are always derived from:
i)
the thoughts a researcher has as a result
of observation or experiences
ii)
the already formulated theories
iii)
the already available data etc
1.7
THE CHARACTERISTICS OF A USABLE HYPOTHESIS
i)
It must be conceptually clear
ii)
It must be measurable
iii)
The operations and predictions indicated
by the hypothesis must be spelt out clearly
iv)
It must be related to the available
methods of research
v)
Must be related to a body of theories.
CHAPTER TWO
2.0 DATA
COLLECTION METHODS
2.2
INTRODUCTION
There are basically two forms of data
collection methods; these are Non-Survey
Research and Survey Research.
2.3.
NON-SURVEY RESEARCH
These are several methods within this form
of research of which are:-
(i)
Focus group discussions
(ii)
Key Informant Interviews
(iii)
Observations
2.2.1
FOCUS GROUP DISCUSSIONS
This is a type of Qualitative Research in
which the researcher brings together small groups of people, 6–12 in number,
with similar characteristics ie Homogeneous, to discuss a particular topic of
research. The Researcher plays the role of a modulator, and with an open
instrument, the researcher engages the group in a discussion about the subject
under investigation. The group is homogeneous in terms of the relevant
variables for the study. If the study is on “Child Abuse”, the groups could be
of 18-28 years, 29-40 years, 41-60 years etc.
It essentially relies on convenient
sampling in which the researcher brings together groups of people that are
easily accessible and have the information. Purposive sampling can also be used
by deliberately choosing the people to get information from.
These Focus Groups Discussions are
basically used for:
a)
securing background information
b)
getting feed back from project
beneficiaries
c)
Interpreting available quantitative data
d)
Project monitoring and evaluation
e)
Assessing responses to recommend for
innovations, policies etc
Advantages of using Focus Group Discussions:-
a)
It enables rapid generation of information
b)
It reduces individual inhibitions and
hindrances
c)
It helps respondents to raise issues and
concerns that the investigation may not have considered ie they come up with
fresh ideas and open fresh issues
d)
They allow an interaction between the
respondents and the investigator, creating a more in depth understanding of
peoples’ understanding and lives.
e)
It flexibility allows the researcher to
use the responses to frame relevant and necessary questions.
Disadvantages of using Focus Group Discussions:-
f)
Empirical generalisations cannot be from
the data
g)
Liable to interviewer biases since these
are no structural questions
h)
May lead to fear of giving personal sensitive information
i)
Despite the presence of the moderator,
there are people who always dominate and those who cannot express themselves
in-group discussions.
Note: The role of moderator:
-
to control those who may dominate the
discussion
-
to encourage those who may not feel free
to talk
-
to probe in case need arises
2.2.2
KEY INFORMAT INTERVIEWS (K.I.I)
These are informal interviews directed to
the knowledgeable people about the problem. These respondents may not
necessarily be under the problem, but only with a sufficient knowledge ie
i)
A District Local Council Chairman, in case
of the prevailing poverty in an area.
ii)
A Police Office, about the rates of
committing a particular offence in an area.
iii)
District Medical Office, about the
progress in fighting HIV/AIDS.
These informants should be selected
carefully to reflect diverse views and concerns. They should be selected from
different social-economic groups and the ideal method should be according to
the nature of the study. An interview guide with issues to be covered is used
and it is carried out in an informal atmosphere. Probes to elicit more
information are made, and detailed noted by the interviewer are made. If
possible a tape-recorder should be used. This kind of data supplements other
forms of data.
It
is most appropriate when:
A general descriptive information is
sufficient for decision making ie-assessing performance of NGOs.
i)
It is necessary to know why a particular
group of people behave the way they do i.e. why
child abuse is on the increase?
ii)
More light/interpretation is needed on the
available quantitative data.
iii)
The primary purpose of the study is to
generate suggestions and recommendations.
iv)
There is a need for proper questionnaire
design, hypothesis and propositions for further testing and refinement.
I)
Advantages of using Key Informant
Interviews:
i)
Since it is from knowledgeable persons, it
reveals in-depth, inside information, to the extent of providing confidential
information, which may not be the case in a formal setting.
ii)
It is cheap to conduct this kind of
interview
iii)
It reveals new ideas, relevant to the
study, which may not have been anticipated, in the planning process.
iv)
Easy to locate potential respondents
willing to give the information.
II)
Disadvantages of using Key Informant
Interviews:
The
information cannot be generalised and therefore less reliable (ie incase it is
biased).
2.2.3
A CASE STUDY
This studies in details, relatively few
persons who volunteer to give all the required information. It could even be
one person, hence the term “Case Study”. Such an individual could give in
detail the culture of a particular group ie Ankole, explaining why, how and
when it began to be what it is.
2.2.4
OBSERVATION
Observation is
“ a purposive or intentional examination of something, particularly for
purposes of data gathering”. (Chaplain 1968). It is a care full watching and
noting phenomena as they occur in their natural setting. In case of any
suspension from the observed, they may change behaviour temporarily. This calls for high skills and patience in
order for them to go back to the normal way of life.
There are basically three forms of
observation:
·
Non-participant observation
·
Observation with some participation
·
Participant observation
I) Non-Participant observation: In this form of observation, there is careful
watching and noting of events as they occur in their natural setting, without
the Researcher’s participation. This form of observation has some problems:
a) It is easy to identify the researcher since he
is a stranger to the observed.
b) Some
salient aspects may not be observed by the researcher.
II Observation with some participation:
This is where in addition to observation; the researcher takes part in some
activities. The problems involved include:-
a) The researcher may not observe some
other aspects in which he does not
take part.
b) The researcher may also fail to get
real meaning of the practices leading to misinterpreting the information due to
his own perceptions.
II) Participant Observation: In addition to observing the subjects, the
researcher shares in the life and activities of those under investigation. This
requires living in the community for a considerable period of time. This help
to eliminate suspension and the subjects won’t continuously change behaviour.
It also gives adequate time to study the events and practices under
investigation and helps the observer to can understand and properly interpret the
practices.
This form of observation has some problems
of which there is:-
a)
Possibility of failing to play a dual role
of a participant and an observer, one of the two may be compromised.
b)
Once the observer is identified, he may fail
to observe each and every aspect of the practice.
ADVANTAGES OF OBSERVATION AS A METHOD OF
DATA COLLECTION:
(i) Able
to see and observe what exactly takes place in its natural form without any
distortion.
(ii)
Data collected is up to date and there is no memory failure.
DISADVANTAGES OF USING OBSERVATION AS A
METHOD OF DATA COLLECTION:
1. In
case of any suspension, the subjects have the potential to change their
behaviour and act otherwise.
2. There
us time constraint, in that some activities take place once for a period of
time, so the researcher has to wait until that activity is performed.
3. Some
events are exclusive to none members and in such cases it is impossible to do
observation.
4. In
case of observer bias, the observer may select on the activities that are more
interesting, and the results may not be complete.
5. Without
enough skills, the results may not be exhaustive and may not describe in
totality the events in the field.
6. There may be unfair and poor
representativeness of things to observe. The observer may fail to select a
representative sample of aspects to observe and results cannot be generalised
to the whole population.
2.30 QUANTITATIVE
RESEARCH (SURVEY RESEARCH)
Unlike the
non-survey types of research, these are based on numbers, and the conclusions
are based on the frequencies of each
category. These include:
2.31 PERSONAL
INTERVIEWS:
In this method, data is collected from the
respondent by asking questions in a face to face, one to one situation. The
interviewer directs questions to the interviewee (respondent) from an interview
guide, which is well structured, and the respondents are filled in by the
interviewer. It is always “one to one” since most people become more free to
give personal information, especially in case of sensitive issues in the
absence of others. This requirement is in exception of special cases like when
an interpreter is needed.
There are two forms of interviews:
I) Formal Interviews:
This is where a set of well framed
questions are asked by the interviewer to respondent and the responses are
recorded in a standard form. Standard, well-structured questions in the same
form, order and manner are directed to all the respondents. It is the Interview
schedule that is used.
II) Informal
Interviews: The interview guide is not structured such that the interviewer
can even change the order and wording
and explain the meaning. Depending on how questions are answered,
supplementary questions can be given. Probing therefore is allowed, especially
when the responses given are not sufficient. This is common when the
interviewer is the researcher.
WHAT TO CONSIDER FOR A SUCCESSFUL
INTERVIEW:
1) Accessibility
- The researcher must seek for information from those who have it. If the
subject of study was “Effects of HIV/AIDS on households”, for sufficient
information, one needs to interview those households which have had a victim of
HIV/AIDS. However, withholding the information, memory failure and
misunderstanding of the questions may lead to inaccessibility.
2) Cognition
– It is a requirement that the respondent knows the role he is called on to do.
There is therefore a need for an understanding on the part of the
interviewee/respondent on what is required of him ie to give the required
information. In some cases, especially where the respondent may want to lead
the conversation which may distort the need and the purpose of the research. It
is necessary therefore that the respondent’s role and purpose of the study be
well explained to the respondent.
3) Motivation
– There is a need to convince the respondent to :
i)
accept to co-operate
ii)
give accurate information
iii)
start and finish the interview
The respondent should never be
embarrassed. Assure the respondents that all responses are correct and that
there is total confidentiality. Symbols/numbers may be substituted for names to
ensure no suspension.
WHAT IS EXPECTED OF THE INTERVIEWER:
1) LOCATING
SAMPLE MEMBERS: This is after the researcher has identified the method of
sampling and is in the field trying to formulate a sampling frame. The sampling
method used determines how easy the task of locating the sample members will
be. If for example simple Random Sampling was used as compared to cluster
sampling where elements will be concentrated in some particular portions of the
whole area.
When
locating a Sample, the researcher meets problems some of which are:
i)
Unfavourable weather conditions
ii)
Hostile attitudes towards strangers
iii)
Failure to locate some addresses
iv)
Mobility of some potential respondents
v)
Unfavourable terraine
vi)
Failure to find potential respondents at
their places, therefore a need for call-backs.
2) OBTAINING
THE INTERVIEW:
To
win the motivation of the potential respondent, the interviewer begins by
introducing himself and the organisation he/she represents to the respondent.
The Interviewer then precisely explains the purpose of the study. The explanation
should include the expected outcomes and how the results will be of benefit.
The researcher should not tell lies about the outcomes of the research. It
should be well explained how the potential respondent was selected and that the
responses are both confidential and anonymous, which cannot be used against the
respondent. It should be made clear to the respondent that the interview is not
testing knowledge and no superior behaviours should be displayed by the
interviewer. A friendly relationship called rapport is thereby created.
3) THE
INTERVIEWS:
This
is the asking of questions and waiting for the responses. The types of
questions expected include; factual questions, opinion questions and knowledge
questions. The questions like: Have you ever been abused?, what opinion do you
have on child abuse? Do you know any effect of child abuse? For factual opinion
and knowledge questions respectively. Of these, it is the opinion questions
which are always neither easy to ask not to answer and the interviewer must
avoid showing his side. The types of responses expected include: Adequate
responses, inadequate responses and non-responses. Of these responses, it is
the inadequate type of responses where keen interest should be put. An
inadequate response could be:-
i)
Partial i.e. incomplete
ii)
Inaccurate i.e. with distortions
iii)
Irrelevant ie that which does not answer
the given question
iv)
Verbalised response ie the respondent
cannot answer all the questions asked.
4) WRITING
DOWN (RECORDING) RESPONSES:
As
the interview goes on, the responses are recorded. This requires careful
listening, and observing any movements/reactions and at times a need for probing. Verbal and non-verbal responses
should be taken care of in order to do proper analysis. Incase of rude, harsh
and embarrassing responses, the interviewer should accommodate it to sustain
the interview.
ADVANTAGES OF INTERVIEW AS A METHOD OF
DATA COLLECTION
1.
It can be supplemental by observation
2.
Can Work for both the literate and the
illiterate
3.
People naturally prefer talking to writing
4.
The interview maintains the order of
questions
5.
Probing is possible, and therefore some flexibility
6.
It is the sampled person who give the
responses
7.
The completion rate is high.
8.
In case of any need for clarifications,
the researcher is available to the respondent.
PROBLEMS MET IN USING INTERVIEW AS A
METHOD OF DATA COLLECTION
1. In
case of bad weather, poor terrained etc, transport and communication are a problem
on the part of the interviewer.
2. In
accessibility caused by the mobility of the potential respondents at the time
they are needed.
3. It
is a very expensive method in terms of training, paying and accommodating the
interviewers.
4. Accommodation
may not be available
5. Some
societies show hostility to strangers
6. In
case of political problems like insurgency
7. Suspension
from respondents for confidentiality and anonymity.
8. Illiteracy
levels of some people.
9. A
problem of language barrier between the respondent and the interviewer.
10. Distortions
and at times filling of the questionnaires by the Interviewer (In case of
tiredness, bad weather etc)
11. Interviews’
opinion may influence the kind of answer received.
RESPONSE ERROS:
When the interviewer records anything that
differs from the real happening, then there is a response error. Response error
therefore is the difference between the true values of the respondent and the
values recorded by the interviewer when interviewing. Each and every question
has the right response; in case a wrong response is given by the respondent, or
recorded by the interviewer, then there is a response error. Therefore,
response errors can either be due to the respondent or the interviewer.
The Respondent as a source of Response
error:
i)
Lack of knowledge: The respondent may
avoid showing ignorance concerning a particular issue and gives a wrong answer.
ii)
Memory failure - The respondent may have
forgotten the right answer
iii)
Misunderstanding the question - The respondent may misunderstand the question
and give an answer which is wrong
iv)
Deliberate lying - The respondent may
understand the question and has the collect answer but decides to tell a lie.
This could be due to:
a)
The question content: e.g. questions
involving social gain, personal circumstances etc
b)
The presence of a “Third Party” which may
jeopardise the conversation
c)
The place of the interview e.g. office,
along the road etc.
d)
Time of the interview - ie busy hours,
after work when one is tired etc
e)
The sponsorship of the inquiry ie
personal, government, institution etc.
The interviewer as a source of response
error:
The response given by the respondent may
be right, but the interviewer either because of mishearing, carelessness or
deliberately records what defers from the right answer.
This could be due to:
a) Interviewer
Opinion: The opinion of the interviewer may be evident which may influence the respondents to give
particular responses.
b)
Personal characteristics: This is what the Interviewer is in relation to
the respondent. It includes age, sex, level of education etc. In case of
sensitive questions a lady of 16 years interviewing an old man of 40 years may
create errors in response.
c)
Interviewer expectations: The interviewer
may expect a particular answer from the respondent which he may not get leading
to recording the expected (wrong) answer e.g. age, level of income etc.
i)
Probing in which meaning is distorted
ii)
Asking ambiguous, vague or leading questions
iii)
Careless recording of the responses
iv)
Wrong translation of the responses
v)
Deliberate cheating or conscious
distortions
vi)
Asking long questions, which the
respondent may forget some parts
vii)
Distorting the order of questions.
HOW TO REDUCE RESPONSE ERROS:
a)
Controlling the interview
b)
Careful coding
c)
Good selection, training and supervision
of the interviewers
d)
Recording verbatim
e)
Use of tape recorders
f)
Matching interviewers with respondents in
terms of age, sex, social status etc.
g)
Motivation of respondents from start to
end
h)
Building a warm rapport and confidence in
the respondent
i)
Concealing of opinions, on the side of the
interviewer
j)
Appropriate questionnaire design i.e.
question order, precision, content etc.
2.32
SELF ADMINISTERED QUESTIONNARES:
In this method of data collection, the
respondent gets and fills in a formerly well structured questionnaire. The
respondent must be literate, i.e. knowing how to read and write.
There are basically two types: -
i)
Mail Questionnaire
ii)
Home/office
I)
The Mail Questionnaire: The questionnaires
are sent to the postal address and the respondent fills and sends them back to
the researcher. This necessitates sending a stamped self-addressed envelope in
order not to burden the respondent. It is popular and most frequent method used
in developed countries.
Advantages of using mail Questionnaire:
i)
costs a stamp and an envelope and
therefore very cheap.
ii)
Reaches isolated and remote areas.
iii)
Eliminates the possibility of non-contact
i.e. not finding the respondent
iv)
Gives the respondent enough time to
reflect, concentrate and at times consult
v)
Can be used to screen for subsequent
researches
vi)
Eliminates response errors due to the
interviewer
vii)
Ensures anonymity and confidentiality
Its Limitations
i)
It works only with the literates
ii)
Works only with topics and questionnaires
which are clear and easy to understand
iii)
There is no probing
iv)
It is not flexible
v)
Works only when there are postal services.
II)
HOME/OFFICE
DELIVERY QUESTIONNAIRES
-
The questionnaires are either sent or
personally delivered to and later picked from the premises of the respondent.
The respondent is given time to fill in the questionnaires. This is commonly
used when:
i)
The postal system is not efficient
ii)
There is need for more explanations and clarification
iii)
It minimises non-response
iv)
It saves time.
Some of its disadvantages are:
i)
It is expensive
ii)
It is tiring
iii)
There may be non-contact
CHAPTER THREE
3.0
SAMPLING
In an investigation, one may decide to do
research on the entire population or
part of the population. The method of
selecting that part of the entire population for purposes of research is known
as sampling.
3.1 Some
Concepts:
a)
Sample: This is the part of the entire population that is
selected for study purposes. It can also be described as a subset of
measurements selected from the population of interest (Mendenhall). A sample
from a population is any subset of that population (Madsen W. R).
b)
Population: This refers to a complete set of elements having some
common characteristics. It is a set representing all measurements of interest
to the sample collector (Mendenhall). A population is a collection of objects
having a well defined set of characteristics (Madsen W. R.)
c) Element:
This is an individual from whom information is collected.
d) Sampling
frame: This is a list from which a sample is drawn i.e. a list of blocks in
a town etc.
e) Attribute: This is a characteristic that describes an
element i.e. rich or poor, illiterate or literate etc
f)
Parameter: This is a summary description of a given variable in a
population. Any numerical value describing a characteristic of a population is
called a parameter. (Walpole R.E.)
g) Statistic:
This is a summary description of a given variable in a sample. Any numerical
value describing a characteristic of a sample is called a stististic. (Walpole
R.E.)
h)
Sampling Error:
The difference between the parameter and
Statistic,
i.e. the difference between the description
of a given variable in a population and in a sample from that population.
STAGES
TAKEN IN SAMPLING
1. Defining
the study population: This shows exactly what the population is in terms of
parameters.
2. Specifying
the sampling frame: In this, a list of all the elements (or group of
elements in form of blocks, maps etc) from which the sample is to be got is
made.
3. Identifying
(specifying) the sampling unit/unit of analysis: This is the smallest unit
to be used in analysis i.e. household, person, school etc.
4. Identifying the method of sampling:
This is the method that is most suitable and chosen in that study for sampling.
It could be a Random or Non-Random method of sampling.
5. Specifying the size of the sample:
The Sample size should be proportional to the population size.
6. Designing a sampling plan: Steps
to be taken at each stage of sampling are described in the plan.
7. Sample Selection: this is the
act of selecting the sample from the population.
WHY SAMPLING IS NENCESSARY
It is necessary because it is:
a) Cheaper: it saves time and
money.
b) Higher level of accuracy. It
permits a higher level of accuracy because of better manpower, and more care
taken in analysis.
c) Quick results: There are some
problems, which require immediate solutions, e.g. an epidemic. In such cases, a
sample can be used.
d) In case of tests which affect the
sampled elements e.g. the melting temperature of a fuse, the effectiveness of a
new drug etc
e)
Higher level of adaptability in samples compared to censuses.
f) Populations
require much management as compared to samples.
SAMPLING ERROR
This is the difference between the true
values of the population and the estimated values of the sample. In other
words, the difference between the parameter and the statistic.
If there is a population of 100 people
whose mean income is Shs. 48,000; and a Random Sample of 10 people is chosen
whose mean income is Shs. 51,000.
Y
= Shs . 48,000
y
= Shs. 51,000
s.e =Y-y
= Shs. 51,000 – 48,000
Shs. 3,000
The smaller the sampling error The more
accurate and exact are the sample
estimates. In the given illustration, if another sample is selected and its
mean income is Shs. 50,000=; the sampling error will be Shs. 2,000= which will
be smaller than the previous one; and the sample estimates will be closer to
the true population values and the better for the researcher.
Sampling error is a function of two
factors; i.e. the structure of the population and
sample size.
a) Population structure: The
population can either be homogeneous or heterogeneous in behaviour. The more
heterogeneous the population, the bigger the sampling error and the more
homogeneous the population, the less the sampling error.
NOTE: If the
population is completely homogeneous, there is no sampling error and if it is
completely heterogeneous, there is no possibility of sampling.
b)
The size of the sample selected: For
heterogeneous populations, the bigger the sample, the smaller the sampling
error. This is so because as the sample becomes bigger, it tends to a
population.
NON-SAMPLING ERROR:
These are the errors outside sampling.
They are due to improper selection of the sample, faulty methods of collecting
and analysing data etc.
SAMPLING TECHNIQUES
These are basically two broad
techniques/methods in sampling:-
i)
Probability (Random) sampling
ii)
Non-probability (non-random) sampling
A technique is under probability sampling
if each element in the population has a chance of being selected into the
sample; otherwise it is non-random sampling.
RANDOM SAMPLING TECHNIQUES:
1.
Simple Random Sampling: this is the
easiest and simplest of all. Each and every element in the study population has
an equal chance of being selected into the sample. A simple Random
Sample of n objects chosen from a population is on chosen in such a way that
all samples of size n are equally likely to be chosen (Madsen W.R). This is applicable when:
a)
The population is relatively homogeneous
b)
The study population is small
such that the sampling frame can easily be specified.
It can be by use of Lottery Method or by
Random Numbers:
In this method, discs/folded papers
written on the names or numbers of the elements to be selected, put in a bowl,
mixed by shaking and picked at random. This can be done with replacement in
which the picked disc is dropped back and the same process continues until the
whole sample is selected. This implies that one element can be selected more
than once. In this case, the probability of choosing an element remains
constant i.e. p= 1/N;
Where
N= population Number.
P=
probability of choosing an element
It can also be with replacement in which
the selected disc is not replaced. The process of shaking and selecting another
disc continues without the previously selected disc until the whole sample is
selected. The probability of choosing each element there varies.
I.e.
p=1/N, 1/N-1, 1/N-2 -------- 1/{N-
(n-1)}
Where P- probability of choosing an
element
N-
Population Number
n-
Sample Number
RANDOM NUMBERS
This uses a table of Random Numbers- The
elements are numbered from 1 – N. The table does not follow any pattern and the
sample is selected in any convenient way i.e. horizontally vertically,
diagonally etc.
ie
109
210 111 413 517 567
316 718 617 273 161 174
100 910 818 512 414 876
327 254 154 146 271 186
511 423 616 641 172 275
156 578 963 985 243 432
694 904 806 372 571 576
_ _ _ _ - -
If a sample of 20 elements is to be
selected from the above table, the first 5 elements could be:
i)
Horizontally – from the left top corner:
109, 210, 111, 413, 517, 567
ii)
Vertically – from the fifth column:
517, 161, 414, 271, 172, 243, 571
2. Systematic
Sampling:
This is always done from a population
which is relatively small and homogeneous. A certain order of selecting
elements is followed. In this technique, a sampling fraction (k) is selected
first. This divides population into the number of samples possible ie k=N/n.
where k = the sampling fraction
N = Population Number
n = Sample Number
The first
selection is done using simple Random Sampling from the first k elements. If W
is the first element selected; the W, W+R, W+2R, W+3R, -------- is the sample.
This is done by selecting every element after the previous one until the whole
sample is selected.
e.g. if the population size is 900; and
the sample size is 30.
i.e N=900, n=30, then
R=N/n = 900/30 = 30.
By simple Random sampling, if W=15, then
the whole sample selected will be W, W+R, W+2R, W+3R, -------
= 14, 45, 75, 105, 135, 165, 195, --------
3. Stratified
Sampling:
In this method, the population is first
divided into groups called strata. It works best in populations, which are:
i)
Large; therefore without a specified
sampling frame.
ii)
Heterogeneous in nature
Each stratum should be as homogeneous as
possible and as different from any other as possible. If for example the study
is on “The Availability of textbooks in the Library.” The Researcher should
stratify according to the Departments.
I.e.
Business Studies
Education
Studies
Social Work Studies
community leadership and Developmentetc.
After stratifying the population, a sub-sample
is selected from each stratum by simple Random sampling or systematic sampling.
These sub-samples can be selected proportionately or disproportionately.
An example of
a proportionate stratified sample:
Uganda Christian University -
1800 students
Law Studies - 300 “
Social Science Studies - 500
“
Business Studies - 400
“
Mass Communication Studies -
100 “
Education Studies - 500
“
If a sample of 180 students is needed from
the total population, the sampling fraction is
S=n/N = 180/1800 = 1/10.
Law Studies - 1/10x300 = 30
Social Studies - 1/10x500 =
50
Business Studies - 1/10 x 400 =
40
Mass Communication – 1/10x 100 = 10
Education
- 1/10 x 500 = 50
---------
Sample Number 180
n
is the sample size
N
is the population size
S
is the sampling fraction
The
sub-sample is got by multiplying the sampling fraction s by the stratum size
(xi).
i.e.
Education Department
S
x Xi = 1/10 x 500 = 50.
For a disproportionate stratified sample,
the sub samples are not proportional to the Stratum
Advantages of using stratified sampling:
i)
It is sometimes the most representative
way of sampling i.e. it ensures representativeness of the sample. This is so
because it helps to capture the view of the minority.
ii)
Heterogeneous populations with their
varied behaviour and opinions bring in sampling error, which is always reduced
by stratification.
iii)
It makes the administration of the whole
exercise easier.
iv)
Different methods of data collect, if
appropriate can be applied
v)
Different sampling frames can be used to
select the sample.
4. CLUSTER
SAMPLING:
This is most common in populations, which
are completely heterogeneous. The population is divided into groups, which are
themselves heterogeneous. A sample of some of the groups (Clusters) is got and
all the elements in the clusters sampled are taken for the study. A part from
this single stage cluster sampling, a multi-stage cluster sampling can be done
where sub-clusters are sampled from the sampled clusters.
This method is most applicable in large
populations, without a specific sampling frame and completely homogeneous.
Reasons for using cluster sampling:
i)
-It is cheap; this is because it saves on
time and money.
ii)
Selection of sample very simple
iii)
It facilitates supervision
iv)
If there is no sampling frame
Differences between stratified sampling
and cluster sampling:
i)
The sample selected from all the strata,
but a sample of clusters is selected.
ii)
The final sample is selected using simple
random sampling from each stratum but all elements in the sampled clusters are
considered.
iii)
The size of the sample is known and fixed
in stratified sampling, but is unknown and varies in cluster sampling.
iv)
To minimise the sampling error, each
stratum should be as homogeneous as possible and each cluster as heterogeneous
as possible.
4.
Area Sampling:
In this method of sampling, areas are
marked; sampled and the elements within the areas are also sampled to make the
final sample. Maps are commonly used and if not available physical features,
such as road, rivers etc are used. These are very commonly used in less
developed countries.
NON-RANDOM/PROBABILITY SAMPLING:
There are several methods under
Non-probability sampling but most of them are used for quick results, and
always by administrators and other persons of some authority.
1. Convenience Sampling: This normally
selects those elements that are easiest to find i.e. those you know by name,
those on the front bench, those you come close to in a gathering etc.
2. Purposive sampling: The Researcher uses
his own judgement about which respondents to choose. He picks only those
elements that best meet the purposed for the study. The Researcher uses his
skills and prior knowledge to choose the respondents.
3. Network sampling: This works best with
very sensitive issues i.e. drug abuse and traffic; sex selling etc. One element
is identified; confidence is built in that respondent who is later given a task
of identifying another. One identified later identifies another until the
sample is complete.
4. Others include:
i)
Quota Sampling
ii)
Dimensional sampling etc.
CHAPTER FOUR
QUESTIONNAIRE DESIGN:
This is commonly called questionnaire
construction. It involves the framing of questions in a format and wording
which will be taken to the respondent. In order to create a favourable atmosphere
(rapport) between the respondent and the researcher, designing a questionnaire
requires beginning with simple and clear questions, to move sensitive and
complex questions.
Set of Principles for Proper questionnaire
Design:
1. Precision: The questions set
must be short and precise but sufficient. They should be set in such a way that
they are answered easily and quickly. Long questions would confuse the
respondent who may probably forget the
first part of the question before the last word is read.
2. Questionnaire
Relevance: The questionnaire should not only be relevant to the objectives
and variables of the study but also to the respondent. This can be achieved by
a multiple of questionnaires or multiple wording which can be assisted by word
like “if No skip To” according to the categories available.
3. Question Order: Questions should
be set in an orderly manner to fit together and form a conversation. There
should be a logical progression for the respondent to be drawn into the study
by raising his interest. This call for beginning with simple and clear
questions to sensitive and complex questions; from more general to more
specific questions. Begin with filter questions, which are followed by
contingency questions.
e.g.
1- Are you in School?
a) Yes b) No “If No, jump to on 4”
2-
In which level are you?
3-
Who pays your School fees?
a) Parent b) Government c) other
4-
How do you always spend your day?
4. Question content – This is
concerned with the wording of the questions. The questions should never offend
the respondent. They should be phrased in a simple and straight forward manner
and should avoid phrases (words) which have several meanings. Generally avoid
a)
Ambiguous questions i.e. questions with
several meanings e.g. where do you come from?
b)
Double-barrelled questions i.e. two or more questions in one e.g. Do you wake
up early on Sunday and go to Church?
c)
Leading questions e.g. You attend Church
Services, don’t you.
d)
Presumed questions e.g. How often do you
attend church services? before you find out whether the respondent attends the
services at all.
e)
Abstract questions e.g. Are you happy with
your Church services?
5. Close ended and open ended questions -
it is normal practice to begin with close ended questions e.g. questions which
require answers like
a)
YES B) NO
a)
Married b) unmarried c) Widowed/separated
Open-ended questions are at times opinion
questions, which bring out salient information. The researcher is at liberty to
even use only close-ended questions or open ended questions.
Advantages of close ended questions:
- Responses are standard and can be easily
compared.
- They are easy to analyse
- The respondent becomes clear of what
the researcher wants
- Answers are
relatively more complete.
Disadvantages of close ended questions:
-
it is easy for the respondent to guess the
answer
-
If one’s category is not available, it
frustrates the respondent
Advantages of open ended questions:
-
Can be used when all answer categories are
not known
-
It allows the respondent to answer
adequately
Disadvantages of open ended questions:
-
They lead to countless responses which are
hard to code
-
Some of the responses may be irrelevant
-
They often require superior interviewing
skills
-
They are often too general to explore
specific aspects
THE DESIGN OF NON-SURVEY RESEARCH
INSTRUCTIONS:
Non-survey Research
is:
i)
not based on probability sampling
techniques
ii)
findings cannot be generalised
iii)
Data tends to be qualitative in nature
Non-survey Research Include:
·
Key Informant interviews
·
Focus Group discussions etc.
Just like the case of survey research,
Non-Survey research is based on the Research problem, the objectives and the
Research Questions. The research instrument is therefore essentially trying to
get the information about the issues determined earlier (The principles used in
designing survey instruments also apply to non-survey instruments though to a less extent).
Other guidelines include:
1. They are generally less structured.
Questions are not pre-coded and are therefore open ended.
2. They tend to have more elaborated
instructions i.e. they introduce and explain the purpose of the study more than
in the survey method.
3. They identify the parameters for data
collection. And in this case, four should be considered:
·
The setting (place, environment)
·
The participants/Actors
·
The events/phenomena
·
The process (i.e. order)
4. The nature of questions: These are
simply guidelines to the interviewer about the kind of issues to be covered.
They tend to be flexible and the interviewer is under no obligation to ask
every question. Each key question is followed by probes, which assist the
researcher to make in-depth investigations.
5. Recording: In this case, unlike in
survey research, the answers are made separate sheet from the research
instrument. The following should be taken care of;
i) The
Descriptive notes - The physical state which the respondent is in.
ii)
The reflective notes - These are personal thoughts.
iii) Information about time, date, and the
number of participants in the interview.
5.
The structure of the instrument:
i)
The Heading
ii)
Instructions to the interviewer
iii)
Introductions
iv)
Key questions to be covered
v)
The probes to follow the key questions
vi)
Space for recording interviewer’s comments
vii)
The space in which the interviewer records
the reflective notes.
CHAPTER FIVE
5.0
DATA PROCESSING
5.1
Data Processing
This involves three stages:-
i)
Editing
ii)
Coding
iii)
Tabulation and analysis of data
When data is collected in a raw form, it
is compressed by putting it into tables, calculated values or graphs. This is
what is known as Data Processing. When the processed data is used to interpret
the findings, the process is known as Data Analysis.
I.
EDITING
This is the checking of the filled
questionnaires to ensure consistency, completeness, and accuracy. It could be
done in the field or back at home but at least within the same day of data
collection. The Researcher checks that all the relevant questions have to be
asked and the corresponding answers recorded. All questions must be answered
and the spaces provided filled in. Otherwise it should be indicated that not
applicable. The researcher therefore goes through the filled instrument to fill
up the gaps, correct the errors as one recalls all the responses and the
reactions of the respondent during the interview.
II.
CODING:
This is a process of giving a category to
every answer so as to bring out their essential patterns.
This has two parts:
a) Coding the frame: This comprised of all
possible answer categories in which the data to be collected may be classified.
For example, at Uganda Christian University, 2001, the Departments can be
categorised as Law, Social Sciences, Business studies, Education and Theology.
The answers are categorised into mutually exclusive1, exhaustive2
and representative categories. Each category of answers is referred to as a
Code. All the codes of the whole questionnaire are what is referred to as a
coding frame.
Foot Note
1. Mutually
exclusive implies that no answer fits in more than one category.
2.
Exhaustive answer category implies that
all the possible categories are provided for in the coding frame.
b)
Coding Answers: This is normally done by
the interviewer by ticking the answer relevant to the question. This is
normally done for close-ended questions. Even for open ended questions, after
all answers have been collected, they have to be coded.
III.
TABULATION:
This is the last stage of data processing, which is always done after editing
and coding. The data is put in tables according to their codes and frequencies.
This is when the data is quantitative in nature, which requires statistical
analysis.
There are two major forms of tables:
a) Univariate Tables which involve the
frequencies of one variable e.g.
Number
of students per subject
SUBJECT No.
R.Methods 400
Economics 500
History 60
Geography 40
------------
Total: 1000
=======
AGE GROUP No.
16-20
200
21-25
600
26-30
200
----------
Total: 1000
======
COURSE No.
Law 300
Education 400
Social
Sciences 300
--------
Total: 1000
====
b)
The Bivariate Tables:
They involve the frequencies of more than
one variable e.g. course by Age.
The final stage is analysis which involves
the interpretation of the research findings. If the data is qualitative i.e.
incase of focus group findings; analysis is done directly from the quotations
given by the respondents and therefore no need for tabulations.
6.0 MEASUREMENT
This is the process of determining the
value or level of a particular unit of analysis.
Importance of Measurement:
a) It is the only way through which
variables are especially defined; implying that a precise description of a
variable can be reached at only through measurement.
b) It enables a researcher to clarify his
theoretical thinking and to suggest new variables.
c) A good theory can be greatly enhanced
through measurement, such that if measurement is done correctly, it can improve
the qualities of research i.e. increasing reliability and validity of research.
d) It is a requirement in everyday life
e.g. measurement of distance, weight etc.
Functions of Measurement:
a)
To describe phenomena empirically
b)
It makes statistical manipulation and
treatment of data possible.
c)
It helps in testing of theories and
hypothesis.
d)
It enables a researcher differentiate
between the objects of study according to the level of properties they posses.
LEVELS OF MEASUREMENT:
There are basically four levels of
measurement.
1. Nominal
measurement: This is the simplest and lowest form of measurement, also
known as the classification measurement because it involves classfying
concepts. In this level of measurement, numbers or lebels are assigned to
identify objects according to their characteristics. This is a qualitative
measurement which classifies elements according to similar characteristics
e.g. Car Numbers.
House Numbers etc
It is therefore for identification
purposes.
It must have:
i)
At least two categories
ii)
The categories must be mutually exclusive
iii)
The categories must be exhaustive.
2. Ordinal
Measurement: This implies following order:
i.e. the ranking of objects according to
their characteristics. This could be either in the ascending or descending
order. In addition to having all the characteristics of norminal scale, it
shows orderliness i.e. it clearly shows pisitions of the objects - Second,
third etc.
3. Interval
Measurement: In addition to all the characteristics of ordinal measurement,
it gives the differences in values e.g. The First student had 10 marks higher
than the second.
Position 1st 2nd 3rd 4th
Total
Marks 340 330 300 287
Differences - 10 30 17
In this form of measurement, O is
arbitrary e.g. ooC= 320F.
4.
Ratio Measurement: In addition to all the characteristics of
interval scale, it has an absolute zero ie a point of beginning. Observations
can be compared meaningfully through ratio measurement e.g.weight 1Kg (A) 10 Kg
(B)
Such that B is ten times heavier than A.
6.1 Evaluation Of Research
Evaluation allows the researcher to decide
for himself what conclusions to draw from evidence/data found.
Questions to think about in evaluation:
i)
Does the data agree or support the
researchers’ conclusions in respect to the population studied?
ii)
If the conclusions are sound, do they make
a generalisation beyond the sampled population and the setting in which the
research was made?
If the date support the conclusions made
in the specific instance studied, then there is internal validity. At times, a
careful look at the conclusions made by the researcher from the evidence can
sometimes reveal errors of reasoning that undermine the conclusions that seem
to follow from the study. Suppose one concludes that poverty is the major cause
of child abuse”, would the data be in agreement with the conclusion?
The second consideration is whether the
conclusion can be generalised to the population, other settings inclusive. This
involves external validity and it is meaningful if the internal validity is
already established. Therefore a study has external validity if its results can
be generalised to other situations in which the same variables operate. Suppose it was true that “poverty is the major cause
of child abuse” in Mukono District
Is
it true in all other places?
Is
it true always?
Chapter
Six
7. The Use Of Secondary Data:
This is the use of the already collected
data that was not specifically gathered for the research question at hand. This
data could be government or non-government/private statistical studies already
undertaken or unpublished observations of a knowledge observer.
7.1 Advantages of using Secondary Data:
1. Cheap, using already collected data
requires less time and money. Therefore the whole research is cheap in terms of
time and money spent.
2. Quick Results: If the researching body
needs quick results, its better to use already collected data than going to the
field for primary data.
3. It acts as a starting point for
additional research from recommendations, policies etc.
4.
It provides the means for increasing the
efficiency of the research costs by targeting real gaps and oversights in
knowledge.
5.
The existing data and the new data can be
compared for purposes of examining differences on trends, and therefore
secondary data acts as a comparative tool in research.
6.
It provides a basis for determining
whether or not new information is representative of a population as in the case
of sampling e.g. comparison of demographic characteristics of a sample to those
of a large population may reveal how representative the sample is of the large
population.
7.2 Disadvantages of using secondary data:
1. There is a possibility of using
outdated data therefore not timely for the research purpose.
2. The unit of aggregation may not be
appropriate for a particular purpose.
3. Categories and measurements used in the
data may not be appropriate for the research purpose at hand.
7.3 How To Evaluate Secondary Data
Secondary Data needs to be evaluated since
it is not necessarily reliable and valid.
Data is always evaluated according to how recent and credible it is.
Data can be evaluated basing on the
following questions:
a) What was the purpose of the study? The
objectives of the study always determine the sampling and data collection
methods, the degree of precision and the types of categories used. This may
limit the usefulness of the data for some other particular purpose.
b)
Who had the responsibility of collecting
the data: The qualifications technical competence, resources available and
potential biases determine the quality of the data collected.
c)
The Research Design i.e. what methodology
was used in obtaining the data? There is need of getting information about:
i)
Size and nature of sample
ii)
Response rates and missing data
iii)
Experimental procedures
iv)
Validation efforts
v)
Questionnaires and coding forms
vi)
Interview guides
vii)
Methods of analysis
This allows for a proper critique of the
data collection procedure. Knowing the sample design is very helpful since it
deals with the question of generalisation and representativeness of the
results.
d)
When was data collection done? Is the data
still current or has been overtaken by some events and is obsolete? Society is
dynamic, like what used to be a taboo in the 20th Century, may be
acceptable today: What used to be a living wage by 1999, may be a “pea-nut” by 2001.
e)
What type of data was actually collected?
How were units and concepts defined? How direct were measures used? How
complete was the information? Apparent inconsistencies across studies often
have more to do with the operational definition of terms than actual
differences in the underlying phenomena.
f)
In comparison to other sources, how
consistent is the information obtained? In case of differences and
disagreements between different sources, there is need of identifying the
reasons why such differences exist and then determine which source is more
credible.
Chapter Seven
7.0
Dissertation Writing
Introduction:
This is not an easy task as some would
want to think. It tests ones ability to manage a research project, and write it
up in a clear and orderly fashion. A high level of originality and initiative
are expected.
7.1 Stages
Of Social Research
1. Formulation
of a Research Problem
2. Formulation
of a Research Design (overall approach to data collections)
3. Method
of data Collection (Personal Interviews, mail questionnaire)
4. Formulation/Design
of Research Instruments
5.
Deciding on the sample size
6.
Data Collection
7.
Analysis and Interpretation of Data
Collected
8.
Preparation and presentation of Research
Report
1.0
Formulation of a Research Problem: This
constitutes writing a Research Topic. Specific situations lead to the
formulation of a particular problem. These include: The Paradigm, values of the
Researcher, Methodology and the time factor.
a) Paradigm:
This is the perspective or frame of reference with which the Researcher views
the Social World - “Child Abuse in Mukono”
b) Values: What one desires influences the
selection of a particular topic – “Women emancipation”.
c) Methodology:
Is the method of research qualitative or quantitative or both. The method you
choose, will influence how you approach the topic you have chosen, ie it
influences on the sampling method, the data collection method, the method of
analysis etc
d) Time Factor: Is it for a long or short
time period? The time schedule in which each part of the research work is to be
completed is very important. The following should be included in your time
planning: -When to have your proposal
approved
-When
to have done the pre tests
-When
to have trained the research assistants
-When
to have collected the data
-When
to have finished data editing
-When
to have completed the analysis
-When
to have completed the research work ready for dissemination.
Be realistic about the time allocated to
each item; one would rather finish an exercise much before than being caught by
time. In case of need, make some readjustments as you go along.
After
a topic has been selected, there is a need for evaluation. The problem area
selected has to be of significant importance to the research world.
What to consider when evaluating a
Research Problem:
a) The
research on that problem is expected to give a solution, which has a
significant contribution to the body of organisation knowledge.
b)
It should be able to fill some gaps in the
present knowledge, solve some inconsistencies in previous research and make a
general improvement in earlier studies.
c)
The solution to the problem should open up
new problems for further research. A good study while arriving at a solution
leads to other problems that need investigation.
d)
The problem must be one that can be
empirically studied i.e. its hypothesis
can be proved right or wrong.
e)
It must be sustainable for the
investigator in terms of:
i)
genuine interest and enthusiasm
ii)
having knowledge and some experience
iii)
being feasible to the investigator i.e.
availability of funds, and ability to complete in the available time.
iv)
Being investigated and completed in the
available time.
1.1
Developing a problem statement (statement
of the problem).
It is advisable to begin with a broader
problem area and narrow it down to a specific research problem. A properly
defined research problem should (i) be stated in the simplest form
ii) Identify the variables being investigated
iii) Indicate the relationship between the
variables being investigated
iv) Identify the target population
1.2
Feasibility of a Research Project:
-
Some students select the topics which have
already been studied, while others select very new topics. The better
alternative is the in between the two extremes. This is because there is no
rationale of researching on what has been researched on and it becomes very
difficult for a student who chooses on very new topics: for it may even be
impossible to identify the relevant literature.
8.2
RESEARCH DESIGN:
This is the general outline of the procedures involved in conducting a
research. It is a set of operating guidelines within which research is carried
out, consistent with the scientific methods available. It can be in form of a
census where a complete enumeration of the whole population is done at a point
in time. It can also be in form of a case study where relatively few persons,
(even one person) are studied in detail. It can also take the form of surveys,
which we are more concerned with where samples using sampling techniques are
selected to represent the whole population.
The design involves the literature review,
the sampling techniques (methods); data collection methods, the identification
of the Independent, dependent and the intermediate variables; formulation of
the hypothesis, the methods of data analysis and the research report writing.
d) Variables:
These are the independent, dependent and the intermediate variables. These are
used in formulating hypotheses where it can clearly be shown that the
independent variable determines the dependent. In tabulation, the independent
is always placed on top while dependent on the left hand i.e. When drawing
graphs, the Independent is always placed on the horizontal axis, while the
dependent is on the vertical axis.
For a variable to be independent,
dependent or intermediate, it dependents on the context of the research. It is
advisable to explain why one considers a variable to be whither independent,
dependent or intermediate.
The Independent variable is always the
“cause” of the event while the dependent is the result or “Effect”.
“Causes of high infant mortality rates in
Uganda”
-Malnutrition, disease & wars could be
some of the causes and therefore the independent variable.
-“Poverty
levels” could be one of the intermediate variables.
Note: Social Sciences, the demographic and
social economic factors are always the Independent variables so long as they
explain the results of the study. They include sex, tribe, religion, level of
income level of education, marital status etc.
e) Hypotheses: A hypothesis is a predicted relationship
between the Independent and the dependent variables, which can be empirically
tested with significant tests.
e.g. The better the Income, the higher the
standard of living.
Income level is the “cause” and the
Independent variable while the standard of living is the “effect” and the
dependent variable.
8.3 Formulation Of Research
Instruments: These are the guidelines used by the researcher to collect
information from the sampled population. They could be formal or informal. The
commonly used are the formal interview schedules called the questionnaire:
There
are five principles, which must be considered for proper researcher instrument
formulation.
a) The
Question order: The questions should be set in an orderly manner to fit
together and form a conversation. Begin with simple clear questions to more
sensitive and complex questions. This helps to remove tension and build rapport
with the respondent.
b)
Questionnaire
relevance: The questions should be relevant to the
study and to the respondent. Make sure that no respondent. Make sure that no
respondent is asked a question that does not apply to him/her. This
necessitates the application of more than one questionnaire or the use of
contingency questions with instructions like “If NO SKIP TO”.
c)
Precision:
The questionnaire should be limited in length and scope. The questions should
be short, easy to answer but sufficient
d)
Question
Content: The wording of the questions must never
offend the respondent. Avoid phrases or words, which have more than one meaning
and the questions, must be phrased in a simple straight forward manner.
i)
Avoid two barrelled questions i.e. two
questions in one e.g. Do you wake up early and attend the chapel services?
ii)
Avoid abstract questions: Are you happy
with this University.
iii)
Avoid leading questions: You don’t smoke,
do you?
iv)
Avoid vague/ambiguous questions: Where do
you come from? This could be residence, place of birth or nationality.
v)
Avoid presumed questions: How often do you
go to Church? Before you find out whether he does.
e)
Open
and close ended questions: The researcher should know how he wants
to categorise his responses and therefore the need for using either open or
close ended questions.
8.4 Writing A Research Proposal: Like any
work or construction, planning is a necessity. As you plan, it is possible to
find out that the study is not possible e.g. in terms of cost, time etc. In is
therefore better to alter things at the planning stage than when you are in the
middle of your study. Luke 14:28.
A
proposal is a detailed description of the proposed study designed to
investigate a given problem. It is a master plan for your study and is
critically important because it will communicate your preliminary ideas on what
you are trying to do, when you plan to do it, how you plan to do it and how you
plan to analyse your data.[3]
It brings about careful thinking about many aspects of the proposed study. In
the process, one may discover matters he otherwise would have ignored.
Evaluation of the proposed study becomes easy and certain weakness may be
discovered and improvement made at the very beginning. It is generally a guide
for conducting the study in terms of orderliness, time schedules etc. It
ensures that the study addresses the problem that has been framed, otherwise
the target may be missed.
Components
of a Research Proposal:
i)Title
page; With the Researcher’s name
ii)
Table of contents
iii)Introduction:
- Background
to the study
- Statement
of the problem
- Justification/significance
of the study
- Objectives
- Hypotheses
iv) Review of relevant literature
v) Methodology
- Subject
- Data Collection instrument
- Design
- Procedure
vi)Time Schedule
vii)The Budget
viii)Bibliography
i) The Title Page: This give the Topic of
study, the name of the researcher and who the researcher is.
ii) Table of contents: This outlines what
is contained in the proposal and on which page it is located.
iii)Introduction
a) Background:
In the background, all information required for an understanding of the problem
is given. This includes what is already known and written on the subject of
study. It helps the researcher to identify the gaps in what is already known
and to identify what is not yet known on the subject of study. It is a stepping
stone in framing the problem statement.
b)
Statement of the problem:
This must be stated as clearly as possible since the rest of the proposal
depends on the problem statement. It should be guided by a series of questions
or statements that will be answered by the proposed research.
c)
Justification of the study:
It must be justified by discussing its significance e.g.
- It
could be filling the gaps in knowledge
- it
could be useful for planning etc.
d)
Objectives – The overall objective that
you want to achieve must be stated; then state specific objectives if
necessary.
e)
Hypothesis:
Each hypothesis should clearly state the expected relationship between the
variables in the study.
iv) Literature review:
This is the reading and writing on what has been written down on the subject
under study. It should be more detailed, coherent and consistent to bring out
the real situation. As the background to the study, it helps the researcher to
clearly identify the gaps in what is known, identify what is unknown and in
framing the hypotheses. Use as many different types of sources as possible.
These may include journals, the internet, textbooks, government publications –
that is; all that you deem relevant and of use. Whenever you read something or
take notes on a source, always take down full bibliographic details to avoid
time wasting and loss of some when you are compiling references and full
bibliography.[4]
v) Methodology:
This helps to identify and work on the following so that a proper research
procedure is followed.
a) Subjects: These are the potential
respondents. Description of the subjects clearly defines the population from
which the sample is selected. Description indicates the size and the major
characteristics of the population. Such should be answered when describing the
subjects:
- Where
are my subjects to come from?
-
What are they like?
-
Are they scattered or close to one
another?
-
Are they homogeneous or heterogeneous?
b)
Design
- This described the basic structure of the study e.g. census, sample survey or
case study. It goes further to explain whether the research will be qualitative
or quantitative or both. It also explains the sampling methods, data collection
methods, method of data processing and analysis and presentation of results.
All this is determined by a number of
factors among which is the population structure, urgency of need of results,
availability of funds etc.
c)
Data
gathering Instrument: This involves the designing of the questionnaire.
Its validity and reliability must be tested before using it in the actual
study. The instrument designed may be structured or not structured. This among
others may be determined by whether the approach is qualitative or
quantitative.
d)
Procedure:
This describes all the steps that will be followed in
conducting the study, in the order in which they will occur.
e)
Data
analysis: Here you specify the method to be
applied in analysing the data.
f)
Anticipated
problems during research. These problems could either be
technical, physical, social or financial. If possible explain how you plan to
overcome them.
vi)Time
Schedule for the different stages of the study mean that the researcher
keeps to the deadlines set for the particular research project.
vii)Budget: It could be in such a form:
Subsistence -30,000/=
Travel -20,000/=
Accommodation -40,000/=
Stationary -100,000/=
Typing
and Secretarial work - 40,000/=
Research
Assistants - 100,000/=
-------------------
Total: 330,000/=
===========
viii) Bibliography
Bibliography
1)
Bailey K. D. (1982). Methods of Social Research 2nd
Edition. The Free Press, New York.
2)
Moser C.A. and Kalton G. (1971) Survey
Methods in Social Investigation. Heinemann, London.
3)
Selltiz c.
et el (1978). Research Methods in Social Relations, 3rd
Edition.
4)
Margaret Riel . Social Science Research
Methods (A handbook of Africa and Revised Edition).
5)
Stuart Melville and Wayne Goddard.
Research Methodology; An Introduction for Science and Engineering Students.
6)
Tereza kakooza. An Introduction to
Research Methodology.
7)
Uganda Christian University (2004) –
Guidelines for Dissertation writing
8)
Jack R. frankel and Norman E. Wallen. How
to Design and Evaluate Research in Education, Third Edition.
9)
Mark L. Berenson and David m. Levine.
Basic Business Statistics: Concepts and Applications.
10)
Hannah W. Kinoti. A handbook of Social
Research Methods.
11)
Richard W. Madsen/Melvin L. Moeschberger.
Introductory Statistics for Business and Economics.
12)
Ronalr E. Walpole. Introduction to
Statistics,3rd Edition.
13)
Mendenhall. Introduction to probability
and Statistics.
13) Martin E. Amin. Social Science
Research; Conception, Methodology and
Analysis
[1]
A collection of facts showing
relationships between them
[2]
An empirically verifiable
observation
[3]
Uganda Christian University – Guidelines for dissertation writing
[4]
Uganda Christian University – Guide lines for Dissertation writing
