UNIT ONE

THE NATURE OF RESEARCH

Definition of research

There are several and different problems that exist and society looks forward for their
solutions. In its basic sense, research means to search and search again and again for
solutions to problems confronting our society. Research is the systematic process of
collecting and analyzing information (data) in order to increase our understanding of the
phenomenon about which we are concerned or interested. It is a study designed to fill a gap
in human knowledge (It is a search for an answer to an unresolved and perplexing question;
an issue that needs in investigation)

How we get to know

There is more than one way to learn about human and animal behaviour. Everyday all of us
use several methods to learn about behaviour. Most of what we know about the world is not
based on only doing a study. One might learn most of what he/she knows using an
alternative to social research. It is based on what your parents and other people such as
friends, teachers have told you. You also have knowledge based on your personal
experiences, the books and magazines, TV shows you have watched. These ways of
knowing are also our sources of knowledge. These ways of knowing include authority,
tradition, common sense, personal experience, media myths, logic and method of science.

Authority

We acquire knowledge by referring to individuals who are socially or politically defined as

qualified producers of knowledge. These may be oracles in tribal societies, ministers of
religions in theocratic societies, kings in monarchical societies, scientists in technocratic
societies, and others like teachers, parents, doctors, lawyers, plumbers, subject experts and
from books. When we accept information or something as being true, then the knowledge
we acquire is said to be from an authority. It is believed that authorities have a much
specialised information or expertise and therefore we tend to believe most of the things
they tell us.

An authority has experience or unique expertise in something and is able to provide

insights and understandings that we are unable to see.

Limitations to relying on Authority

1. It is easy to overestimate the expertise of other people. You may assume that they are
right when they are not. For example, some ‘experts’ of the past measured students
readiness to school by telling the students to place their hands over their heads to touch
their ears. Their errors seem obvious now, but you can be certain that today’s experts will
not become tomorrow’s fools?
2. Authorities may not agree, and all authorities may not be equally dependable.

3. Authorities may speak on fields they know little or be plain wrong. An expert who is
informed about an area may use his or her authority in an unrelated area.

4. Using the halo effect, expertise in one area may spill over illegitimately to be authority
in a totally different area. For example, have you ever seen a movie star using his or her
fame as an authority to convince you to buy a car in television commercials ? We need to
ask: who is or not an authority?

Tradition

People sometimes rely on tradition for knowledge. Tradition is a special case of authority
that is authority of the past. Tradition means you accept something as being true because it
is the way things have always been. For example, my mother says that when you eat and
talk, you will die. When I asked about this statement, she said that she learned it from her
father when she was a child.

Limitations

1. Tradition can be distorted from generation to generation.

2. Traditional knowledge may work once in the past but not now.

3. People may cling to traditional knowledge without really understanding it.

Media Myths

Television shows, movies and newspapers and magazines are important sources of
information. For example, most people have no contact with criminals but learn about
crime by watching television shows and movies. Although many journalists try to present a
realistic picture of the world, they must write stories in short time periods with limited
information and within editorial guidelines. Unfortunately, the media tend to perpetuate the
myths of a culture. For example, the media shows that most people who receive welfare are
black (actually most are white).

Personal experience

If something happens to you or you personally see it or experience it, you accept it as true.
Personal experience or ‘seeing is believing’ has a strong impact and is powerful source of
information and knowledge. People believe what they see or personally experienced rather
than what very carefully designed research has discovered.

Limitations

2
1. Personal experience is a basis for misleading people through propaganda, cons or fraud,
magic and some advertising. Overgeneralization occurs when some evidence supports your
belief, but you falsely assume that it applies to many other situations, too.

2. Selective observation occurs when you take special notice of some people or events and
tend to seek out evidence that confirms that you already believe and ignore contradictory
information.

3. Premature closure, this reinforces the first two limits. It occurs when you feel you have
the answer and do not need to listen, seek information, or raise questions any longer.

4.Halo effect, it is when we overgeneralise from what we accept as being highly positive or
prestigious and let its strong reputation or prestige rub off onto other areas.

Scientific Method

The scientific method is a method for conducting an objective investigation. The scientific
method involves making observations and conducting an experiment to test a hypothesis.
The number of steps of the scientific method isn't standard. Some texts and instructors
break up the scientific method into more or fewer steps. Some people start listing steps
with the hypothesis, but since a hypothesis is based on observations (even if they aren't
formal), the hypothesis usually is considered to be the second step. Here are the usual steps
of the scientific method.

Scientific Processes:

Step 1: Make Observations

You may think the hypothesis is the start of the scientific method, but you will have made
some observations first, even if they were informal.

Step 2: Propose a Hypothesis

It's easiest to test the null or no-difference hypothesis because you can prove it to be wrong.
It's practically impossible to prove a hypothesis is correct.

Step 3: Design an Experiment to Test the Hypothesis

Step 4: Test the Hypothesis

Step 5: Make a decision to fail to reject (Accept) or Reject the Hypothesis

Step 6: Revise the Hypothesis (Rejected) or Draw Conclusions (Accepted)

The scientific method is illustrated with the model below

3
Characteristics of Research

The nature of educational research is analogous with the nature of research itself, which is
systematic, reliable and valid to find the ‘’truth’’ investigates knowledge and solves
problems (Wierma, 1991). Moreover, educational research process involves steps to collect
the information in order to investigate problem and knowledge. However educational
research is more complex because it can use various approaches and strategies to solve
problem in educational setting. Therefore educators can use those research findings to
improve their competences and teaching and learning.

Furthermore, the characteristics of educational research are a part of its nature. According
to Gray Anderson (1998), there are characteristics of educational research which can be
grouped into three categories namely the PURPOSE of the research, the PROCEDURES of
the research and the ROLE of the researcher. The purposes of research are to solve the
problems, investigate knowledge and establish the principles in educational phenomena.

This includes: Educational research attempts to solve a problem

It attempts to find an objective, unbiased solution to the problem and takes great pains to
validate the procedures employed.

4
 The procedure is an important characteristic of educational research which involves
collecting data with accurate observation, objective and interpretation and verification.

It is a deliberate and unhurried activity which is directional but often refines a problem or
questions as the research progresses. In short, it focuses on solving the problems and
developing knowledge.

This involves:

 Gathering new data from primary or firsthand sources or using existing data, for new
purpose.
 Demand accurate observation and description.
 Generally employ carefully designed procedures and rigorous analysis.
 Carefully recorded and reported to other persons interested in the problem based
observable experience or empirical evidence.

PURPOSE
Emphasise the development of generalizations, principles or theories that will help in
understanding, prediction and/or control.
Finally, the role of the researcher needs to be considered since researchers needs to be experts
and familiar with their field of study, using the data to develop solutions and increase
knowledge. The researchers also need to be patient and careful to use every step of
researcher’s procedure to achieve the purpose of the research. This implies –the research
requires expertise, familiarity with the fields, competence in methodology, technical skill in
collecting and analyzing the data.

CHARACTERISTICS OF RESEARCH
According to Leedy (1974) research has certain basic characteristics which include;
1. Research begins with a question, an issue or problem or a perplexing phenomenon in the
mind of the researcher. This point out that before you set out to do research, you must have in
mind a question to answer, an issue to explain, or a perplexing phenomenon to solve. You
must know that research usually arises from a question, an issue or a problem that a researcher
has observed and it puzzles him/her. If in your teaching you observe that in a particular
subject area students are not performing well, then you have a problem you can seek through
research. You can find out why students are not performing well in that particular subject and
your concern for finding out why they are not performing begins a research.

2. Research requires a plan: Research entails a clearly stated plan, including direction and
procedures. Research is a systematic activity and for it to be successful it requires a plan
known in research language as Research Proposal. This plan is very important because just as
a journey started without knowing its destination gets nowhere. As Mager(1975) puts it, if
you’re not sure where you are going, you’re liable to end up some place else.

3. Research demands a clear statement of the problem: A good research begins with a clear
statement of the problem or issue or concern. Whatever issue, problem or phenomenon you
set out to deal with in your research it must be stated clearly indicating the relevant variables
and the expected relationship if any. It is also important to note that statement of a problem is

5
preceded by background information to help readers to understand what exactly you are
studying.

4. Research deals with the main problem through sub-problems: Most at times, the main
issue or problem of concern may be too big to solve at a go. It is therefore important to
breakdown the main problem into smaller problems to which the researcher will find
solutions to. By solving the smaller problems within the bigger problems, the main problem
is solved.

5. Research seeks direction through appropriate hypothesis or questions: The problem

having being stated and sub-problems being attended to, are then stated in the form of a
hypothesis or a specific research question.

6. Research deals with fact and their meanings: Even though research involves collection of
data, it goes beyond the collection of data. In research, whatever relevant data (information)
collected must be organised into meaningful aggregate, analysed and interpreted .

7. Research is circular: The research cycle begins with a problem at hand. Next is the
formulating of specific research questions or hypothesis through the sub-problems that
related to the main problem. The process is continued by collecting of relevant data,
organising, summarising and analysing the data after which he/she will interpret the result.
The result is the identification of another problem.

Fig 1: Circular Nature of Research

6
Generally, characteristics of research include;

1. Empirical: Research is based or direct experience or observation by the researcher.

2. Logical: Is based on valid procedures and principles.

3. Cyclical: This is because it starts with a problem and ends with a problem.

4. Analytical: It utilizes proven analytical procedures in gathering the data, whether

historical, descriptive experimental and case study,

5. Critical: That is, exhibits careful and precise judgement.

6. Methodical: Is conducted in a methodical manner without bias using systematic method

and procedures.

7. Replicability: The research design and procedures are replicated or repeated in an

identical way as a safeguard against unintentional error or deception.

CLASSIFICATION OF EDUCATIONAL RESEARCH

For the purpose of identifying the different types of educational research in terms of what
they help to achieve (contribute) and how they are conducted, research are classified
generally using purpose, method and type of data.

The purpose deals with what a researcher hopes to achieve for carrying out the study and
this involves the following;

1. Basic Research: This also known as Pure Research. It is aimed at generating

fundamental knowledge and theoretical understanding about basic human and other natural
processes. This type of research is designed to contribute to theory development and
refinement of theory. Basic research is not concerned with practical applicability. That is, it
often lacks a practical application in the short term. Notwithstanding it provides a
foundation for knowledge that advances understanding in many policy areas, problems, or
areas of study. It is often conducted in highly controlled and contrived setting. Examples
include studies and how memory system works, language development and social
development. A study of how animals are used to determine principles of reinforcement
and their effect on learning can also be basic.

2. Applied research: This research is focused on answering practical questions to

provide relatively immediate solutions. It is designed to address a specific concern or do
offer solutions to problem identified by an employer, club, agency, social movement or
organisation. In order ways, applied is conducted for the purpose of applying theory and
evaluating its usefulness in solving educational problems. Applied research is usually
conducted on a quick small-scale study that provides practical results for use in short term.

7
Applied research and basic research can be compared on the following basis: Basic
Research and Applied Research Compared:

Basic Applied

Research problems and subjects 1. Research problems are narrowing

are selected with a great deal of or constrained to the demands of
employers’ freedom. sponsors.

The driving goal is to contribute 2.The driving goal is to have practical

to basic theoretical knowledge. payoffs or uses for results.

Success comes when results appear 3.Success comes when results are used by
in a scholarly journal and have impact sponsors in decision making.
others in the scientific community.

The primary concern is with the 4.The primary concern is with the ability
internal logic and rigor of research to generalize findings to areas of interest
design. to sponsors.

Types of Applied Research

There are many types of applied research but for the sake of the course, we will consider
three major types: evaluation, action and social impact assessment.

-Evaluation Research: Evaluation involves determining the worth, merit, or qualities of an

evaluation object. This type of research involves decision making regarding the relative
worth or effectiveness of two or more alternative actions. It is designed to find out whether
a programme is worth continuing, there is a new way of making a programme effective, in
other words, does it work? This type of research is widely used in large bureaucratic
organisation (e.g, businesses, schools, and hospitals, government) to demonstrate the
effectiveness of what they are doing. Evaluation research is directed towards making
decisions about the effectiveness or desirability of a programme. In most cases evaluation
research is focused on a specific location or type of programme and involves judgements
about such questions as:

-Which reading curriculum should be implemented?

-Did the new programme work?

-Should they build small hospitals or one large hospital?

-What is the impact of technology on student and teacher knowledge and attitude?

8
Often such questions require research methods that are unique to each situation. Evaluation
research has its own approach or methodology.

-Action Research: Action research focuses on solving practitioner’s local problems. It is

generally conducted by the practitioners after they have learned about the methods of
research and research concepts that are discussed in textbooks. Action research is a
collaborative approach to research that provides people with the means to take systematic
action in an effort to resolve specific problems. This approach endorses consensual,
democratic and participating strategies to encourage people to examine reflectively their
problems or particular issues affecting them or their community. Furthermore, it
encourages people to formulate account and explanation to their situation and to develop
plans that may resolve these problems. Its purpose is to solve a specific classroom or
school problem, improve practice or make a decision at a single local site. The most shared
five characteristics of action research are:

-the people being studied actively participate in the research process.

-the research incorporates ordinary or popular knowledge.

-the research focuses on issues of power.

-the research seeks to raise consciousness or increase awareness of issues.

-the research is tied directly to a plan or program of political action.

Action research tends to be associated with a social movement, political cause or advocacy
for an issue. Action research, by its nature can be conducted by teachers to solve problems
in their classrooms. For example, a teacher can conduct an action research in his/her
classroom to determine which grouping procedure results in the highest achievement for all
students.

-Social Impact Assessment Research: It estimates the likely consequences of a planned

intervention or intentional change to occur in the future. This type of research often
examines a range of social outcomes and work in an inter-disciplinary research team to
estimate the social outcomes. It usually includes quality of life issues such as access to
health care, illegal drug and alcohol use, schooling quality, employment opportunities etc.

-Research and Development: This is a type of research directed at the development of

effective products that can be used in schools. The main purpose of research and
development is to develop effective products for use in schools. Such products may include
teacher training materials, instructional materials and school management systems. The
development of such materials constitutes the research. Such materials must be first
developed, field-tested and revised until a pre-specified level of effectiveness or quality is
achieved before they are put in schools.

9
THE ROLE OF RESEARCH IN EDUCATION

Introduction

Research plays a vital role in education. Although it is true that university teachers by
virtue of their appointment are required to conduct research, personnel in other educational
settings like teachers and school administrators also read and use research in a number of
ways. This session explains the contribution of research to education in four important
ways.

Adding to already existing knowledge about Educational issues

All educators including you want to do things in a better way to ensure success of children
in their upbringing. Educators therefore strive for continued improvement of what they do.
As a teacher when you teach and your students do not perform well to your expectation;
you begin to ask why it happens so. You may want to know whether your presentation has
some weakness in terms of methodology you employed or because the students do not have
the necessary prerequisite entry behaviour. These questions are aimed at improving your
teaching. This requires addressing problems or issues relevant to teaching and searching for
potential solutions to them. It is through research that we address problems or issues with
the view to finding solutions to them. In the example given about the performance of
students after instruction and the questions raised, you may want to establish a cause and
find a solution to the poor performance through research. By this action, you add to your
knowledge and practice.

Adding to knowledge means that educators undertake research in order to contribute to

existing information about issues. it is through research that we extend as well as broaden
our understanding of a phenomena.

Research adds to our knowledge about educational issues by addressing

the gaps in knowledge, expanding, replicating and adding voices of
individuals to knowledge.

By investigating a pressing educational issue, the void in existing information is filled. For
instance, if you are to choose the most effective method among existing methods of
teaching mathematics in your school and you are not sure which is the most effective
method, research can help you address this gap in your knowledge. In terms of expanding
knowledge, when you extend research to new ideas or practices then your knowledge will
expand. That is you will get to know more about those ideas or practices.

Replicating knowledge involves testing of results with new participants or at a new

research sites. Sometimes results of a study may apply to only some category of
participants or under certain conditions or contexts. Replicating a study helps throw more

10
light on the issue and help us to understand it better. Through research, we tend to know
the perspectives and views of individuals on an issue. This helps us to appreciate the issues
involved and how people react. In all these, research contributes to our knowledge of
educational issues.

Improving practice

Research is also important because it suggests improvements for practice. In most scientific
inquiry, solutions are sought for identified problems. Based on the information from the
study recommendations are usually made as a way forward. These recommendations
suggest ways of improving upon the current status of whatever is researched. As a teacher
when you are armed with research results you become more effective as a professional.
This effectiveness translates into better learning of your students. It is true that educators
such as teachers, headmasters and guidance and counselling coordinators are helped by
research to improve their practices on jobs. Indeed as a teacher, you benefit a great deal in
your practice as a teacher from research.

While research supports improvement for practice, it also offers practicing educators new
ideas to consider as they go about their jobs. For instance, a recommended way of doing
something in the classroom based on research may not work with you. You may begin new
ways of doing it, considering the specific context. These new ways can be tested for their
validity through research. In addition, form reading research studies, you can learn about
new practices that have been tried in other settings or situations. The main point being
emphasised here is that research helps bring about new ideas to be considered when
we are doing our jobs. In addition research enables practitioners evaluate approaches
that they hope will work with an individual in educational settings. This process
involves going through to determine which results will be most useful. A three-step process
can be used. First you have to decide what you want to do in, say your classroom. Second,
you have to find out what researchers have to say (review of relevant literature). Third, you
decide which line of action to take based on the literature reviewed. For example, in
determining whether you should use the discussion or lecture method in teaching a social
studies topic. Remember that research will help you evaluate approaches that will work in
some specific settings.

In nutshell, we have discussed the main ways by which research helps in improving
practice and they are

 Research suggests improvement for practice

 Research offers educators new ideas and
 Research helps practitioners evaluate approaches that will work.
Research helps the practicing educator, like you, build connections with other educators
who are trying out similar ideas in different locations. The connections with other
educators can take the form of attendance of seminars, workshops and conferences.

11
Informing important policies

Research does not only add knowledge and improve practice but also informs policy
formation. Research informs and guides the formation of policy issues. Policy issues are
broad frameworks within which things must be done. An example of educational policy in
Ghana is the medium of instruction at the first three grades (i.e. class one to three).Any
meaningful policy framework should be based on credible information. Research creates
the condition and discussions about important issues among policy makers. In Ghana, these
policy makers may range from individuals in the Ministry of Education (MOE) and Ghana
Education Service (GES) as far as the education sector is concerned. Such individual may
collectively discuss and take positions on important educational issues. For these
individuals, research offers results that can help them weigh various perspectives when
policy makers read and review research and issues they are informed about current debate
and stance taken by other people. On the other hand, research can lead to the modification
of an existing policy.

It is important to note that, for research to be useful, the result should be clear and include
data- based evidence.

UNIT 5

THE RESEARCH PLAN (PROPOSAL)

Research like any purposeful activity requires a plan. For any researcher to be successful in
his or her attempt to conduct a good research, the researcher needs to prepare a plan. This
plan in research language is known a research proposal. After you have identified your
research idea, reviewed the research literature, determine the feasibility of your study,
make formal statement of the research questions (and hypotheses for quantitative study),
you are ready to develop a research proposal to guide your research study. In doing your
dissertation/ thesis, you may be required to develop a plan for the project. In this session,
you will learn about the essential elements of a research proposal and their relevance.

Research proposal and its importance

A research proposal is a detailed formal plan or scheme of how to carry out a study . In
other words, a research proposal describes in detail how to investigate a given problem .
Like any purposeful activity, conducting research requires a plan. The important point is
that for a meaningful research, there must be a plan so that the researcher can proceed
purposively. A research proposal is important for several reasons.

First, it makes the researcher think through every aspect of the study, including making his
or her intension known to others, the problem to be studied and the methodology to
employ. It also helps the researcher to identify the material and resources needed for the
study.

12
Second, the proposal serves as a guide for conducting the study. The research plan is like a
building plan for the mason which the individual researcher can follow in executing the
study. Without it, we may sometimes digress from what we originally thought of doing.
This does not mean that when you prepare a plan you follow it rigidly in carrying out the
study. Some modifications can be done if it will result in more credible conclusions.

Third, the preparation of the research proposal saves time for the researcher. Once the
researcher goes through the preparation of the plan, the steps to be taken, the materials to
be needed are all taken care of such that one does not get to a stage and has to pose to
really plan things and secure things at that stage. Its saves time for supervisors too, since
they have ideas about what the student is doing before hand and advice them accordingly
before the study begins.

Forth, by planning a research ahead of time, the probability of making avoidable mistakes
in carrying out the study is reduced. This is because once you set out to prepare a research
plan, you will think through some of the obstacle you are likely to come across and the
likely mistakes you can make. You will benefits from the mistakes from other people in the
area of study through reviewing their studies.

Fifth, preparation of a research proposal results in a higher quality of research. All the
planning that goes into the write up of the proposal will ensure that the final product of the
study has high level of quality. The high quality also stems from the fact that, once the plan
is made available for people to read (supervisors and peers), they may make some inputs in
terms of suggestions for the improvement of the study.

Sixth, the research proposal facilitates evaluation of the proposed study by the researcher
and by others. The evaluation is based in some cases, on whether the basic idea of the study
seems to be a good and the strength of the case that is made for the study in the proposal
itself is acceptable.

-Essentials of a research proposal

It is important to note from the onset that, agencies and institutions have their requirements
for research proposals. Nevertheless, whether you are preparing a proposal for an agency,
for your project, a proposal should be written so that its objectives, theoretical framework,
and methodology are made intelligible to a well –informed specialist and non-specialist
alike. Generally, a proposal has four main components. These are introduction,
methodology, time schedule and the budget. However for this course emphasis will be
placed on the introduction and the methodology.

Introduction of a proposal

13
The introduction section of a proposal puts the study into perspective. Its describes the
background to the problem, states clearly what specific issue , problem or question the
researcher wants to investigates, how the study relates to literature in the area of study, and
the statement of the specific hypothesis or questions to be tested or answered. Proposals
will vary in terms of emphasis with regards to the above mentioned aspect of the
introductory part of the proposal depending on whether the problem is given or not.

In cases of undergraduate project where the individual student moots the idea of study, the
issue of clearly stating the background to the problem and its statement is crucial. The
description of the background should be concise and help the reader to understand and
appreciate the specific problem to be investigated and its objectives.

A background description may be historical and include relevant literature.

In fact, in practice, you must review relevant literature before stating the background to the
study and the problem as well as the formulation of the hypothesis and or questions.

Statement of the Problem

The statement of the problem follows the description of the background of the study. It a
concise statement of the problem the study is addressing including its scope. The statement
of the research problem is very pertinent and must be considered with all the seriousness
that it deserves. A research problem can be defined as an issue in need of investigation .
Thus, problems spring from puzzle experience. However, a vague feeling that something is
wrong or that some theory or policy is not adequate does not constitute a problem. Such a
situation is an indication that a problem may exist. Before you proceed with your proposal
you must isolate, shape, and clarify the pertinent variables and relationships that give rise
to the problems’ statement.

This you can do by reviewing related literature. This will enable you put the problem in
appropriate theoretical framework or context. The statement of the problem should indicate
the variables of interest to the investigator and the relationships as well as the subject of
study. The research problem can be stated in difference ways. For example,

The problem to be investigated in this study is secondary school teachers’ attitudes towards
school-assessment in Ghana.

The purpose of the study is to investigate is the effect of the CARE/IMAGE project on
dropout rates of girls in implementing districts in Ghana.

This study is to investigate the effect of kindergarten attendance on pupils reading

performance at class one.

14
Even though, the researcher has to evaluate the entire proposal at the end, it is expedient to
evaluate the problem in terms of its appropriateness. Macmillan (1996) identified six
criteria that can be used to evaluate a research problem. These are:

Is the problem researchable? (i.e. The problem should be such that you can collect,
organise and analysed data).

Is the problem important? (i.e. significant to practice, policy or theory).

Does the problem indicate the type of research? (Example: Quantitative, qualitative or
mixed methods)

Does the problem specify the population? (i.e. the target group of study).

Does the problem specify the variables of interest? (Example: achievement, self-
efficacy, perception or motivation).

Is the problem clear? (i.e. devoid of ambiguity)- Put the issue on paper

Hypothesis / Research Questions

After stating your research problem and making sure that your problem is a good one by
evaluating it, you must formulate either the relevant hypothesis or research questions. The
hypothesis or questions must be directly related to the research problem under
consideration. You also note that the main problem is solved through sub- problems. These
sub-problems are what should be restated as research hypothesis or questions depending on
the problem. Hypotheses are educated guess or conjectures that are testable. Hypotheses
are formulated usually in quantitative research whiles research questions are used in
qualitative research and evaluations. Quantitative research however, can involve the
formulation of both hypotheses and questions. Depending upon the problem of study, the
researcher should state the specific hypotheses or questions of the study. This is in line with
the fact that, research deals with a main problem through sub-problems. It is expedient to
divide the main problem into appropriate sub-problems and formulate hypotheses or
questions to cover them. Hypotheses must be stated in a research form and a null form. It is
appropriate to state your hypotheses in both ways.

If a study is qualitative, questions and/ objectives may be formulated as it may be in

evaluation research, for example. It is worthy of note that the research hypothesis may be
stated in a directional or non-directional form. This is guided by review of related
literature. A directional form is written if you believe that the relation will be in a particular
direction.

An example of a directional research hypothesis is ‘there is a positive relationship between

secondary school students’ performance in mathematics and physics’. If you are not sure
about the direction, then the research hypothesis must be stated in a non-directional form.

15
The equivalent non-directional research hypothesis is ‘there is a relationship between
secondary school students’ performance in mathematics and physics’. The direction of your
hypothesis has implications for the probability level you will use for your data analysis,
particularly the power the test.

A one-tailed test (directional) is more powerful than a two-tailed test.

Significance of the study

Even though the problem may be clearly stated, there is the need to explain to the reader
the importance of the study. That is, it should be clear as to why the problem needs to be
studied. You need to explain how the results of the study will be useful in terms of how it
will influence practice, inform policy or contribute to theory. The review of literature is a
useful source for this.

Methodology section of the proposal

The methodology section of the proposal describes subjects/ participants (respondents) of

the study, the design, research instrument and procedures to be followed, including how
you will analyse data collected to test each hypotheses or answer each research questions.

Generally, it is required that the description of the methodology must be particularly clear
and precise to permit replication. When written well, the methodology section of a proposal
becomes Chapter 3 in most project or thesis reports, just as the introduction section
becomes Chapter 1.

In a study, a researcher may be interested in a group of elements or cases, whether

individuals, objects or events, that conform to specific criteria and to which he or she
intends to generalise the results of the research. The group becomes your target group. The
target group is the population of study. In the methodology section of the proposal, you
have to define your population (target group) of study and explain why it is relevant for the
study. An example of target group is all farmers in Cape Coast Municipality. If all
members of the target group are not accessible then it will be expedient to define the
accessible group in addition to the population of study. If the population is going to be
studied without selecting a subset (sample) of it then you need not think about the sample
for the study and this is termed as census. On the other hand, if there is going to be
selection of sample then there needs to think of the sample procedure to be used.

Sample size and Sampling

If the study will involve a sample, then you need to determine the sample size. Appropriate
sample size is needed for the credibility of the results. The sample size is also important in
making sure that your study is able to detect an impact if there is any. In sample surveys of
very large populations, one to five percent of the population will be adequate for the study
when the sample is well selected to be representative of the population . It is believed that,

16
the larger the sample the better the results of the study. For statistical analysis, a sample
size of 30 or higher is considered to be large. This brings us to the issue of sampling.

Sampling

The procedure for selecting a sample is known as sampling. A sample can be selected using
probability or non-probability method. These are as follows;

Subjects and Sampling

Probability Non-probability

Stratified Convenience/accidental/haphazard

Simple random Purposive

Systematic Quota

Cluster Snowball

The specific sampling procedure must be selected based on the characteristics of the
population and the practicality of the sampling procedure. The stratified sampling method
can be proportional or disproportional. The sampling should also be a representative of the
population of study.

The Design

The researcher needs to state and describe the proposed design. The design is the specific
strategy you will employ in collecting, analysing and reporting the research. The design
must be appropriate to enable you to collect relevant data to test your hypotheses or answer
your research questions. The problem, if stated well, gives direction as to the appropriate
design to be used.

17
Basic design in education include experimental and non-experimental.

Experimental (Involves manipulation and control of factors)

Research Design Quasi-Experimental (involves the use of intact groups)

Non-Experimental (Non-manipulation or control of variables)

Note that when you prepare a proposal, you need to state your design and show its
appropriateness to the solution of the proposed problem. In other words, your design should
match the problem. It is also useful to discuss the strength and weakness of the design,
particularly when you are preparing a project or thesis proposal.

Instrument/ Tools

The instruments that you will use to collect data must be described in the proposal. By the
instrument, we mean research tools such as questionnaire, interview, observation guides or
tests. If multiple instruments are involved in the study you must describe each. If one tool
is to be used in the study, it is important to show evidence of its reliability and validity for
the study. If no instrument is available for the study, you have to develop an instrument
then explain how it will be developed and what it will measure. Such instrument must be
pilot tested and validated before use. The process of developing the instrument must be
clearly explained in the proposal. It is important to state when each instrument will be used
or administered and for what purpose. Any instruments that you will use in the study must
be described.

Data Collection Procedure

The data collection procedure section of the methodology should concisely describe how
you would collect your data step by step using the instrument. If a baseline or pre-test data
collection is necessary, it must be indicated including how and when and by which
instrument? This applies equally to post-test data collection. If some people will assist you,
there may be the need to train them. That is, the assistants must be given some type of
orientation that explains the nature of the study and the part they will play in it. They
should understand exactly what they are going to do and how they are going to do it. For
instance, when the study involves interviewing participants, then there is a need to train
interviewers and let them take part in mock interviews until they acquire the necessary
level of skills, competencies in interviewing. This must be made clear in the proposal
including the number of assistance.

Analysis of Data

The last section of the methodology of a proposal should explain to the reader how data
collected will be analysed. For quantitative studies, the researcher may have to employ

18
 statistical procedures in analysing the data. The choice of specific choice of analysis
depends on

1. The scale of measurement to be used and

2. The characteristics of the data to be collected in terms of meeting or not meeting the
conditions/assumptions underlining the use of chosen statistics.

The use of inappropriate statistical procedure to analyse the data can result in a wrong
conclusion. Usually, it is expected that for each hypothesis, the researcher should indicate
the specific statistical he or she proposes to use to test it. In some cases, descriptive
statistics or percentages are used to analyse data.

For qualitative research, content analysis and other methods can be used to analyse the
data. Data analysis goes with interpretation of results. There must be an indication of how
you intend to interpret the result and the implication of what the result say. That is, the
proposal should include the summary of the anticipated conclusions and implications in
light of the proposal findings and previous research.

Evaluation of a Research proposal

It is useful to evaluate your proposal before submitting it to your supervisor for

consideration. You can use both informal and formal ways to evaluate the proposal.
Informally, you can review the proposal several days after you had written it. You can give
it to a colleague or an expert to review.

Formal evaluation of a research proposal involves a pilot study-a mini study or a sort of
dress rehearsal. In this case, the entire study is conducted in miniature form. Weaknesses
discovered during the study are corrected in the proposal. The proposal provides must be
written in the future tense. A complete proposal should provide relevant references.

UNIT 2 (PART OF IT)

THE RESEARCH PROBLEM AND ITS CHARACTERISTICS

The Research Problem

A research study begins as a question that the researcher would like to answer or a problem
that the researcher would like to solve. A problem may usually imply that a controversy or
difference of opinion exists. It is the desire to find a solution to or investigate this problem
(or question), which motivates the researcher to undertake the study. The problem that
motivates the researcher to undertake his/her study is what is known as the Research
Problem.

In order to understand the term research problem better, let us look at how some authorities
in the field of research have explained it. Frankel and Wallen (2000) note that, a research

19
problem is the focus of a research investigation. In fact, it is exactly what its name implies -
a problem that a researcher wishes to investigate. Polit and Hungler (1995) indicate that a
research problem is exactly that issue which someone would like to research.

A research problem is the fuel that drives the scientific process, and is the foundation of
any research method and experimental design, from true experiment to case study.

Look at any scientific paper, and you will see the research problem, written almost like a
statement of intent.

Defining a research problem is crucial in defining the quality of the answers, and
determines the exact research method used. A quantitative experimental design
uses deductive reasoning to arrive at a testable hypothesis.

Qualitative research designs use inductive reasoning to propose a research statement.

DEFINING A RESEARCH PROBLEM

Formulating the research problem begins during the first steps of the scientific process.

As an example, a literature review and a study of previous experiments, and research,

might throw up some vague areas of interest.

Many scientific researchers look at an area where a previous researcher generated some
interesting results, but never followed up. It could be an interesting area of research, which
nobody else has fully explored.

20
A scientist may even review a successful experiment, disagree with the results, the tests
used, or the methodology, and decide to refine the research process, retesting
the hypothesis.

This is called the conceptual definition, and is an overall view of the problem. A science
report will generally begin with an overview of the previous research and real-world
observations. The researcher will then state how this led to defining a research problem.

A research problem can be anything that a person finds unsatisfactory or unsetting; a

difficulty of some sort; a state of affairs that needs to be changed; anything that is not
working as well as it might. We need to note that problems involve areas of concern to
researchers. As educators, we want to improve our conditions and we want to eliminate
difficulties. Questions for which we seek answers are part of the problems we encounter.

Is there any condition you want to improve in your school, environment or community?
What are some of these difficulties you would wish to eliminate in your work place? Are
there some questions that you are seeking answers?

Characteristics of a Good Research Problem

Although selecting a research problem is often one of the most difficult steps in the
research process, it is unfortunately the one for which the least guidance can be given. No
rules have been established for making a final selection of a research problem. In fact,
problem selection is not subject to some degree of technical rules or requirements as are
research designs, measurement or statistics. Fortunately, however, some guidelines could
be offered.

1. It should ask about a relationship between two or more variables

A good problem statement of any simple study must name the variables and how they may
be related in a single sentence. For example, the problem "How many students in Cape
Coast Municipality JSSs can perform above average in mathematics" makes no attempt to
deal with a relation between variables. If however, the problem is worded, "Are boys more
likely than girls to perform better in mathematics in Cape Coast Municipality JSSs?" then it
would involve the relationship between the variables of "gender and "mathematics ability".

Give another example of a problem that deals with no relationship between two or more
variables. Reword the problem to make it ask about a relationship between two or more
variables.

2. It should be clear and unambiguously stated

The importance of a clear concise research problem cannot be overemphasized. Can you
offer any one reason why the research problem should be clear and stated in unambiguous
form? Now read on.

21
One purpose of the research problem is to communicate the purpose of the study.

This can occur only when the reader's understanding of the purpose is consistent with the
researcher. The terms in the problem must not be ambiguous. When do we say a
term/statement is ambiguous? Compare your response to this. According to Amedahe
(2002) ambiguity is seen when different people reading the same thing derive different
meanings from what is read. Thus, if a term or phrase can mean several things to different
readers, it is ambiguous.

These terms are too general and should be defined specifically so that their meanings are
clear. The terms can be given operational definitions or qualified to make them clearer. The
researcher as well as other researchers and readers must know precisely the problem that is
being investigated. For instance, terms like "students and methods" could be made clearer
by adding gender or course say female/male students or English and History students
whiles "method" could be clarified by stating the type of methodology, say, English as a
second language (ESL) methodology.

3. It should be stated in a question form

The problem statement should preferably be in a question form. Alternatively, it should be

in the form of an implicit question such as "the purpose of this study was to determine
whether….”

Let us consider the following example of statement of a problem in a question form: Does
students who have had pre-training be taught learning task more quickly that those who
have not had pre-training?

Problems stated or given in the form of implicit question include:

The purpose of the study is to find the effect of socio-economic background on academic
achievement

The study investigated whether the ability to discriminate among parts of speech increased
with chronological age and education level.

Give an example each of a problem stated in a question form and implicit question form.

4. It should be researchable
When do we say a problem is researchable?

A problem is-said to be researchable when it is possible to collect data to answer the

questions asked. A researchable problem is one that can be investigated through the
collection and analysis of data. In other words, it should be possible to construct a potential
solution to the problem that can be verified by the collection of certain evidence or
disconfirmed by the collection of other evidence. Problems that are concerned with value

22
questions or philosophical ideas are not researchable in the sense that a specific question
has a correct answer. Many interesting questions in education concern ethical issues that do
not lend themselves to easy scientific study. To be able to conduct research, the question
must lend itself to the systematic process of gathering and analysing data. Research can
assess how people feel about such issues but cannot resolve them. Consider the following
examples: “Charity begins at home” and “The evil that men do lives after them”

5. The problem should not represent a moral or ethical position

The research problem must not involve physical and psychological harm or damage to
human beings or to the natural social setting or environment of which they are part .
Questions about ideals or values are often more difficult to study than questions about
attitudes or performances.

Examples of such problems that would be difficult to study or answer clearly include
"should children be seen and not heard?"

Problems such as "Should people avoid cheating/lying under all circumstances?" and "Are
all teachers/chiefs equally inspiring?" represent moral and ethical issues and should be
avoided.

6. The problem must be significant

The solution to the problem must not be a trivial one. It should be seen as adding
information to the present state of knowledge either in theory or practice. The solution of
the problem should contribute to extension, refinement or revision of a theory or contribute
to improvement of the educational process in terms of practice or policy, Amedahe (2002).

7. The research into the problem must be feasible. The feasibility here implies:

The researcher should be able to get the necessary financial support or has the necessary
funds available as will be required by the study.

The researcher would be able to complete the study in good time. It must be noted here that
good time may depend on the duration of one's programme or the time available for one to
submit his/her research report.

The researcher has adequate knowledge of the required research design and statistical
procedures for the study.

To be sure of the goodness of your problem, McMillan (1996) identified six criteria that
could be used to evaluate a research problem. The criteria in question form are:

1. Is the problem researchable?

23
2. Is the problem important in terms of contributing to understanding of the issues being
investigated?

3. Does the problem indicate the type of research?

4. Does the problem specify the population (target group) of the study?

5. Does the problem specify the variables of interest?

6. Is the problem clear?

If your answer to each question regarding a problem statement is yes, then the problem is a
good one.

SOURCES OF RESEARCH PROBLEMS AND THEIR STATEMENT

This session is concerned with the sources that are commonly used to begin the process of
problem formulation. You should note that coming up with a good research problem is
usually a difficult and time-consuming task, There is no single strategy for identifying
research problems that works best for all the investigations.

Beginning researchers often are puzzled and perhaps even threatened by a requirement to
develop a research problem. Now pause and reflect on the following questions:

i. Where, then, do ideas for research problems come from?

ii. How can a topic be selected?

Seven sources are commonly used to begin the process of problem formation. Research
problem may be obtained from any of the following sources:

i.. Theory

ii. Experience

iii. Literature

iv. Ideas from external sources/experts

v. Replication of previous studies

Theory: A very common source for a research problem is a theory that has implication for
educational practice. McMillan (1996) defines a theory as a set of propositions that explain
the relationship among observed phenomena. Theories are, thus, general explanations
behaviours. It contains generalizations and hypothesized principles that should be subjected
to thorough and exact scientific investigation. One approach to obtaining a research
problem is to take a theory in a related area such as psychology, sociology and mathematics

24
and develop a problem that is applicable to an educational setting. A theory in this case
would suggest implications that can be further researched in educational settings.

Another way to apply theories is to directly test, revise, or clarify an existing theory. The
intention here is to develop and change the theory rather than to test its implications. Can
you think of a theory that has been tested, revised or clarified in your field/subject of
study? State a theory in your subject area, you would like to test, revise or clarify.

Experience: Every individual's everyday experience provides a rich supply of problems for
investigation. Some of the best sources of ideas come from ones interest, personal and
practical experiences. Recall some of the situations you met that you found puzzling or
problematic. If you have ever asked such questions as "Why are things done this way?", "I
wonder what would happen if..." or "What method would work better?” you may be well
along the way to developing a research idea.

As a teacher you encounter many problems in the classroom, the school, or community that
lend themselves to investigation. These problems are perhaps more appropriate for the
beginning researcher than are problems more remote from his/her own teaching experience.
There may be concerns about teaching methods, grouping, classroom management, tests
and continuous assessment or a multiple of everyday experiences. What problems do
school administrators face in course of their duty? List any three.

Administrators may face problems in scheduling, communicating to teachers/workers,

providing instructional leadership, generating public/community support, handling serious
disciplinary issues. For the beginning researcher in particular, experience is often the most
compelling source for research topics. In addition to personal experience, individuals have
interests and knowledge about their professional work that can also yield good researchable
problems.

Literature: Ideas for research projects often come from reading the relevant and related
literature. What do we mean by literature? Don't answer. Read on.

According to Amedahe (2002), by related literature, we mean all written documents

including scholarly textbooks, journals, theses, dissertations related to a general issue of
interest. The beginning researcher can profit from regularly reading current books and
journals especially those that report the results of studies in their areas of interest. Reading
published reports may help the neophyte researcher to find a problem amenable to
scientific investigation. It may also help to familiarize the beginning researcher with the
wording of the research problems and the actual conduct of research studies. Published
reports may suggest problem area indirectly by stimulating the reader's interest in a topic
and directly by specifying further areas in need of investigation.

Ideas from External Sources: External sources, here, means course lecturers, project work
supervisors, and development agencies like DFID, USAID, GTZ and other NGO's.

25
External sources can sometimes provide the impetus for a research problem. Research
topics/problems may be given as a direct suggestion from an external source. For example,
a course tutor/faculty member/project work supervisor may give students a list of topics
from which to choose or may actually assign a specific topic to be studied or researched
into. Entities that sponsor funded research such as NGO's, DFID, UNICEF, USAID and
ODA often identify broad or specific topics on which research proposals are encouraged.
For example, in recent years NGO's and the Government of Ghana have requested a variety
of AIDS - related research projects as well as research into cassava and yam at the Science
Faculty and the School of Agriculture at the University of Cape Coast.

However, even when a research area is suggested, it is better for the researcher to identify
the aspect of the problem that is of greatest interest to him/her. This is because curiosity is
a critical ingredient in a successful research.

Expert: An expert here refers to anybody knowledgeable on certain issues and can
therefore guide others with regard to those issues. Consulting and or talking to such experts
would be helpful in identifying a research problem. This is important because such experts
may have knowledge on topical and contemporary issues in their field of specialization
which they can share with others.

Replication: Have you met this word before? Look for the meaning and write it in your
jotter. Compare what you wrote with the following. A type of study is one that replicates,
that is, repeats a completed study with relatively minor changes that will further contribute
to knowledge by either confirming earlier findings or disconfirming them.

A researcher may, for instance, wonder whether a study similar to one reported in journal
article would yield comparable results if applied in a different setting or with different
subjects. Studies that are repeated for the purpose of verification are known as replication
(Polit & Hungler, 1995).

Replications are, therefore, needed to establish the validity and generalisation of previous
findings. It may also be needed to verify and extend the initial findings of an already
published work.

Borg and Gall (1989) summarised four reasons for conducting replication studies.

1. To check the findings of a major or milestone study. Replication can confirm or

disconfirm the validity of a study that produces new evidence or that reports findings that
challenge previous research or theory.

2. To check the validity of research findings with different subjects.

3. To check trends or change overtime. Replications can be used effectively to see if vital
"findings hold overtime.

26
4. To check important findings using different methodologies

Clarification of Contradictory Results: Contradictory results/findings exist on many

topics in the l i t e r a t e . Some studies indicate one conclusion, and other studies of the
same thing come to an opposite conclusion. These apparent contradictions present very
good opportunities for research. Once there are contradictions on a specific issue, there is
room to investigate it further for its better understanding.

Statement of the Research Problem

You have now known that a study cannot progress without the choice of a problem. You
have also learnt that the problem should be carefully stated in written form before
proceeding with the design of the study

Putting one's ideas in writing is often sufficient to illuminate (bring to light) the
ambiguities and uncertainties.

A good research statement of the problem should serve as a guide to the researcher in the
process of designing the study. What ingredients should then be included in the problem
statement? Now read on. The problem statement should identify the key 'study variables,
and their possible interrelationships, and the nature of the population of interest.

Researchers have alternative ways of expressing the research problem. Research problems
are expressed in two main forms:

Statement form

Question form

Statement Form

Many researchers state their research problem in the declarative form as a broad statement
of purpose such as:

The purpose of this research is to examine the relationship between the retention levels of
students who are taught addition of numbers using concrete materials (real objects) and
those who are taught without concrete materials (in abstract).Does this statement of put
pose indicate the population of interest? Compare your answer to the following.

The statement of purpose indicates the population of interest. These are students taught
with concrete materials. The independent variable is the method of teaching the addition of
numbers and the dependent variable is the retention levels of the students.

Question Form of Stating a Problem

A research problem may also be presented in an interrogative form, that is, as a question.
For example, the problem expressed in statement form above can be stated in a question

27
form as: What is the relationship between the retention level of students 'who are taught
with concrete materials and those who are taught without concrete materials? Another
example is: What is the process by which adult children make decisions regarding the
placement of their elderly parents in a nursing home?

The question or interrogative form has the advantage of simplicity and directness.
Questions require an answer and help psychologically to focus the researcher's attention on
the kinds of data that would have to be collected to provide that answer.

Give three examples each of declarative and interrogative forms of an identified problem in
your area of work/study.

NOT CAPTURED ON COURSE OUTLINE

RESEARCH HYPOTHESES AND QUESTIONS

In research, particularly quantitative research, investigators often make educated guesses

about a correct solution to a problem in the form of possible relationships or differences.
These guesses are stated in the form of hypotheses or questions. This session discusses
research hypothesis in detail. We will specifically discuss among other things, what is a
hypothesis, when researchers use hypothesis, types of hypothesis and criteria for evaluating
research hypotheses.

What is a Hypothesis?

The next step after the selection of the problem is to state a hypothesis. Tuckman (1994)
defined hypothesis as a suggested answer to the problem. According to Amedahe (2002), a
hypothesis is an informed/educated guess about the answer to a question.

It is a tentative expectation about a correct solution to a problem, descriptions, possible

relationships or differences. In research, a hypothesis is the investigator's prediction or
explanation of relationships between two or more variables. In fact, it is a tentative
explanation for certain behaviours, phenomena or events that have occurred or will occur
as part of solving the problem. It is the most specific statement of a problem. It states what
the researcher thinks the outcome of the study will be. In other words, it is the researcher's
expectation of what results will show.

The following are examples of a hypothesis. Intelligence quotient (IQ) and academic
achievement are positively related

The amount of time spent learning a task is directly proportional t o the level of
achievement.

There is a relationship between dropout rate among children in the urban and rural settings
in first cycle schools in Ghana.

28
Importance of Research Hypothesis
Amedahe (2002); Polit and Hungler (1995) and Tuckman (1994) maintained that a research
hypothesis is important in the following ways.
It provides a focus for the study.
The research hypothesis often provides a clear framework and a guide when designing the
data collection instruments and when collecting, organising, analysing and interpreting the
data.

It allows the investigator/researcher to confirm or disconfirm a theory and help advance

knowledge.

The most common use of hypothesis is to test whether an existing theory can be used to
solve a problem. In everyday situations, those who confront a problem often propose
informed hypothesis that can be tested to directly solve a problem. For example, when a
lamp/bulb fails to light when the switch is turned on, several hypotheses come to mind
based upon our understanding of electricity and our past experiences with bulbs/lamps.

What are some of these hypotheses? Write any two. Compare your response to the
following.

The plug is not properly connected to the wall outlet

The bulb is burnt out
The fuse is burnt out
There is power failure
A disconnection by the Electricity Company as a result of non-payment of bills, etc

Each of these speculations can be tested directly by checking the plug, connection,
inspecting fuse, replacing burnt bulb, by nothing whether there is light at nearby houses,
etc. A hypothesis helps the researcher to know exactly what to do.

Types of Hypotheses
There are many forms and types of hypotheses, depending on their structure, goals, nature,
how they are derived, or how they are stated. We shall consider two main ways of
classifying hypotheses. These are described briefly below.

Classification in terms of whether they are derived from inductive or deductive logic and

Whether the hypothesis is stated as a research or statistical hypothesis (declarative versus

null hypothesis)

Inductive Hypothesis
An inductive hypothesis is a generalisation based on observation. In induction, the
researcher starts with specific observations and combines them to produce a more general

29
statement, of relationship - hypothesis. Inductive hypothesis is, thus, formed from a
researcher's direct observations of certain variables, patterns, trends or associations among
phenomena or behaviour and the uses of these observations as basis for a tentative
explanation about how the behaviours are related to one another and how other variables
such as teaching methods, curricular materials, teacher behaviour and socio-economic
status. In the school situation, teachers can provide a rich source of inductive hypothesis
because they can use their rich experiences and knowledge to formulate hypotheses that
may explain observed relationships.

A limitation of the inductive hypothesis is that because they depend on local data and direct
observation by researcher, the generalisations are often restricted and hard to relate to a
broader theory or established body of applied research.

Deductive Hypothesis

The other mechanism for deriving hypothesis is through deduction. Deductive hypotheses
are derived from theory and thus the testing of them contributes to a better understanding
of the theory or its application. Theories of how phenomena behave and interrelate cannot
be tested directly through deductive reasoning. A researcher can develop scientific
expectations or hypotheses based on general theoretical principles. In deduction, general
expectations about events, posed on presumed relationships between variables are used to
arrive at a more specific expectation or anticipated observation. Although a full text on
deductive reasoning is beyond the scope of this course, the following syllogism illustrates
the reasoning process involved.

All men are mortal

Aban is a man
Therefore Aban is a mortal
In this simple example, the hypothesis is that Aban is in fact a mortal, a deduction that
could be verified.

Hypotheses are also classified as being either research hypothesis or statistical hypothesis.

Research Hypothesis
Research hypothesis, also referred to as substantive, working, declarative or-scientific
hypotheses, are statements of expected relationship between variables. It states an expected
relationship or difference between two variables. In other words, it states what relationship
the researcher expects to verify through the collection and analysis of data. It is, thus, a
declarative statement of results the investigator expects to find.
Examples of Research Hypothesis

30
"Teenagers who have been sexually abused as children are at higher risk of depression and
suicide than teenagers with no history of sexual abuse".

"Infants born to drug addicted mothers have lower birth weights than infants with non-
addicted mothers".

3. "There is significant difference in reading comprehension of primary school children

who participate in cooperative learning activity than those two receive regular instructions
only",

A research hypothesis can be directional or non directional,

Directional Hypothesis

A directional hypothesis is one in which the specific direction, such as higher, lower, more
or less, that a researcher expects to emerge in a relationship is indicated. It indicates or
states the nature of the expected relationship or differences, That is, the researcher predicts
not only the existence of a relationship but also the nature of the relationship. It is used
when the researcher anticipates the specific outcome of the study. For Example, in a study
of the two groups of pupils mentioned in the example under research hypothesis, the
expectation may be that one group may score higher or perform better than the other group.
In this case the researcher may formulate a directional hypothesis,

Let us look at the following examples of directional hypothesis.

Lower primary pupils who are given concrete teaching aid/materials in mathematics
lessons would demonstrate higher achievement than pupils who are taught without teaching
aids.

Newly trained teachers tend to be more approving of the teachers expanding role than older
teachers.

There is positive correlation between time spent on task/activities and how well the task is
performed.

Non-Directional Hypothesis

A research hypothesis can also be non-directional. A non-directional hypothesis, by

contrast, does not stipulate/state the direction of the relationship. Such a hypothesis
predicts that two or more variables are related but makes no projections concerning the
exact nature of the association. The researcher believes there will be a difference in
relationship but is unsure about the nature of it.

Examples of non-directional hypothesis

There is relationship between the age of a teacher and approval of the teachers expanding
role.
31
Older teachers differ from younger teachers with respect to approval of the teachers
expanding role.

There will be a difference in achievement between lower primary pupils who are given
concrete teaching learning materials in mathematics lessons and their counterparts who are
taught without concrete teaching learning materials.

There will be a difference in achievement when individually tutored children are compared
to those who received group tutoring.

Statistical Hypothesis
The statistical hypothesis is a statement of a relationship or difference that can be tested
statistically. Herman (1996) maintains that statistical hypotheses are translations of
research hypothesis. It allows us to rephrase our research hypothesis in such a way that we
can apply statistical procedures to them. He continued by stating that statistical hypotheses
are statements that describe the population parameters our sample data will represent if the
predicted relationship exists or does not exist.

A statistical hypothesis is nothing more than a claim made about a population –a claim that
can be subjected to testing by drawing a random sample from the population of interest.

Stating the Statistical Hypothesis

Statistical hypotheses are usually stated in what is called the "null form", hence, it is also
known as the Null hypothesis. Let us now look at the null hypothesis.

What is a Null Hypothesis?

A null hypothesis states that there is no relationship (or difference) between variables and
that any relationship found will be a chance relationship, not a true one. A null hypothesis
is, thus, a statement that differences or relationships have occurred because of chance.
Normally, this is a statement of no difference or relationship.

The null hypothesis is tested and its acceptance or non-acceptance provides support or no
support for the research hypothesis.

With regard to the example of research hypotheses given above, the corresponding null
hypotheses are?

There will be no relationship between the age of a teacher and the approval of the teachers
expanding role.

There will be no difference between old teachers and younger teachers with respect to the
approval of the teacher's expanding roles.

There will be no difference in achievement between lower primary pupils who are taught
with concrete teaching materials in mathematics and those who are taught without concrete
32
materials.

It is important for you to note that, usually, research hypotheses are formulated when a
study is more or less quantitative in nature.

Relationship between Directional and Non-directional Hypothesis

The table below shows the relationship between directional and non-directional hypotheses.

Directional Non-Directional

1. Upper primary children like school less 1. Lower primary, Upper primary
than lower primary children but more and Junior Secondary
than JSS students. School children have different
degrees of liking school.

2. Students with academic disabilities will 2. There will be difference

have more negative attitude about between the scores in an
themselves if they are in special classes attitude measure of students
than if they are placed in regular classes. with academic disabilities
placed in special classes and
such students placed in regular
classes.

3. Counselees who receive client centered 3. There will be a difference in

therapy express more satisfaction with the expression of satisfaction with
counseling process than do counselees the counseling process between
who receive traditional therapy. students who receive centered
therapy and students who
receive traditional therapy.

33
 Relationship among Research Problem, Research Hypothesis and
Null Hypothesis

The table below shows the relationship among the Research Problem, Research Hypothesis
and Null Hypothesis.

Characteristics of a Good Research Hypothesis

Research problem research hypothesis Null Hypothesis

1. What is the effect of 1. Teachers' attitude 1. There is no difference in

mainstreaming towards mainstreaming teachers' attitude towards
workshop on the will improve as a result mainstreaming measured
attitude of teachers of attending a workshop before a workshop on
towards on mainstreaming. mainstreaming compared to
mainstreaming? their attitudes after the
2. There is a positive workshop.
2. Is there a relationship correlation between
between teachers' teachers' attitude 2. There is no correlation
attitudes towards the towards the curriculum between teachers' attitudes
curriculum and and students towards the curriculum and
student achievement? achievement. students achievement.

3. Is there a difference 3. Students receiving 3. There is no difference in

in achievement highly detailed written achievement between students
between students who comments on their work receiving highly detailed
are given highly will show higher comments about their work
detailed written achievement than compared to students
comments on their students given grades receiving grades only.
work compared to only
students who are
given grades only?

Hypothesis can be in any form except in the form of a question. The form notwithstanding,

they have to meet a number of standards.

In general, hypotheses are required to demonstrate the following characteristics (Amedahe,

2002; Gay, 1991; Polit & Hungler, 1995; Sarantakos, 1998).

1. The research hypothesis should be empirically testable. Being testable means being
verifiable. That is, it can be empirically verified as right or wrong. It should be
possible to support or not to support the hypothesis by collecting and analyzing data.

34
2. A good hypothesis should normally be testable within some reasonable period of time.
For example, the hypothesis that requiring nursery pupils to brush their teeth after lunch
everyday will result in fewer people with false teeth at age 60, would obviously take a
very long time to test and the researcher and some respondents might very likely be
dead before the study was completed.

In addition to the above, McMillan (1996) has offered 7 criteria for evaluating the research
hypothesis.

The research hypothesis should be stated in a declarative form

and not in a question form. The research hypothesis should be consistent with
known facts, preview research and theory. The research hypothesis should follow from
other studies and established theories. In general, it should not contradict previous research
but rather should build on related literature. The result should contribute to an established
body of knowledge.

The research hypothesis should follow from the research

problem. Do not use variables in the hypothesis that have not been identified by the
research problem. Generally, a problem may include several variables and thus several
research hypotheses may be derived to indicate all the anticipated relationship.

The research hypothesis should state the expected relationship

between two or more variables. A hypothesis must have at least two variables
and must indicate how the variables are related.

A research hypothesis should be testable. -

The research hypothesis should be clear. Similar to the terminology used

in research problems, words, phrases, and descriptions in the research hypothesis should be
unambiguous. A clear hypothesis is easier for the reader to understand and easier for the
research to test. Vague terms and jargons should be avoided.

The research hypothesis should be concise. The hypothesis should be

sufficiently detailed to communicate what is being tested and of the same time be as succinct
(i.e. expressed in few words) as possible. A concise hypothesis is easier to comprehend.

35
UNIT THREE
LITERATURE REVIEW: SECONDARY SOURCE

Definition of Literature Review

Before planning the details of the study, researchers usually read written related materials
to the problem or topic in detail to find out what has been written about the topic or had
been done. Opinions of experts in the field, as well as other research studies are of interest
to the researcher in performing this important duty. Such reading is referred to as a review
of the literature. The review of literature involves the systematic identification, location and
analysis/"digging" of documents containing information related to the research problem.
These documents include periodicals, abstracts, reviews books, and o t h e r research
reports.

Purpose of Reviewing Related literature

Literature reviews serve a number of important functions in the research process. By
examining some of their specific functions, we hope to clarify their value.

Avoidance of Unintentional Replication

A thorough search through related research avoids unintentional replication of previous
studies. Do you remember the term replication? Hope you do. That is nice. Let’s continue.
A researcher may through the review of related literature discover that a very similar study
has already been made in his/her research problem. In such a case, the researcher must
decide whether to deliberately replicate the previous work or to change the proposed plans
and investigate a different aspect of a problem.

Review of related literature, thus, helps to determine what has been done that relates to the
problem under investigation. This leads to the avoidance of unnecessary duplication of
research work. ,

Refining/delimiting Research problem

Reviewing related literature helps researchers to delimit the research problem which results
in limiting their question and to clarify and define the concepts of their study. A problem or
research question may be too broad to be carried out, or too vague to be put into concrete
operation. For example a problem like "continuous assessment in Ghana" would be too
broad to be carried out. What is some of the aspects that can be carried out as a means of
delimiting this problem?

36
Similarly a problem like Drop out in Ghana would be both broad and vague. Literature s
review also helps in clarifying the concepts involved in the study and in clearly defining
the terms in relation to the study. For example through review, many educational and
behavioural constructs such as stress, creativity, frustration, aggression, motivation, bright,
dull, etc. could be clarified and operationally defined.

Identifying Useful Methodology Techniques

By studying related research, investigators learn which instruments that have proved
useful and those which have not been productive in investigating the particular problem
and thus guide the researcher in the study. Many research projects fail because of the use
of inappropriate procedures, instruments, research designs and statistical analysis. A
thorough examination of the methodologies of previous studies often results in finding the
reasons for the failure of past studies, as well as insight into the selection of an appropriate
methodology for one's own research. Both the success and failures of past work provide
insight for designing one 's own study.

Facilitating the Interpretation of Results of the Study

The study of related literature facilitates the interpretation of results of the study. In fact,
the review of related literature places researchers in a better position to interpret the
significance of their own results. The researcher learns how other researchers have
interpreted their results. Becoming familiar with theories of the field and with previous
research, prepares researchers for fitting the findings of their research into the body of
knowledge in the field.

Identifying contradictory Findings

A review of related literature may reveal contradictory findings that may be a fertile area in
which to conduct subsequent research.

Developing Research Hypotheses/Questions

Appropriate hypotheses or questions are usually based on a review of literature. The
literature provides a basis for a hypothesis or research question.

Desirable Replication

Review of related literature may suggest the desirability of replication to confirm

previous findings. Do you remember the term Replication? Repeating a study increases
the extent to which the research findings can be generalised and provides additional
evidence of the validity of the findings.

LITERATURE REVIEW: SECONDARY SOURCE

D e f i n i t i o n o f Secondary Sources

37
A secondary source is a description of a study or studies prepared by someone other than
the original researcher. In other words a secondary source is one that reviews,
summarises, or discusses someone else's research. Review articles that summarise the
literature on a topic are secondary source. When you have completed and written up a
review of the literature on a topic, your document will be considered a secondary
reference. If you go on to collect a new data on the same topic, however, your description
of the research problem, methods, and results of the study will be a primary source
reference for others doing a literature review. Some examples of secondary sources are
textbooks, scholarly books devoted to a particular topic, reviews of research in books or
journals, yearbooks, and handbooks.

Importance of Secondary Sources

Secondary sources are useful in providing bibliographical information on relevant primary

sources.

They are good to start with because they provide an overview of the topic, often citing
relevant research studies and important primary sources.

A limitation of secondary sources is that they are rarely possible to achieve complete
objectivity in summarising and reviewing written materials. In other words it could be
possible that the researcher or author did not accurately report the research since they
combine the information from other secondary sources and actual studies.

LITERATURE REVIEW PRIMARY SOURCE

Definition of Primary Sources

A review of literature involves identifying primary sources that investigate a problem or

idea of interest. A primary source, from the point of view of the research literature, is the
description of an investigation written by the person who conducted it. Primary sources are,
thus, original articles and reports in which researchers communicate directly to the reader
the methods and results of their study. They are reported in a wide variety of journals both
referred and non-referred. For example most of the articles appearing in journals such as
"Ife Psychologia, Edo Journal of Counseling and Ghana Journal of Health and Physical
Education, Recreation, Sports and Dance " are original research reports and therefore are
primary sources.

The Role of the Primary Source

Having read the write up on primary sources of literature review, can you think of any one
importance of this source? Write at least one importance of this-source in your jotter. Now
read on. According to Amedahe (2002), it is important to read primary sources because
they allow more informed judgment about the quality and usefulness of the information.
They tend not to be distorted through being summarised or reported in another form. The

38
primary sources of literature provide sufficient detail about research studies . It is possible
to achieve complete objectivity in summarising and reviewing written materials.

Steps to review Related Literature

I. Locating Existing Reviews and Other Information in Secondary

Sources Secondary sources like: Textbooks in most major fields of education e.g.
edueati9pal psychology, reading, special education, measurement and evaluation, etc. may
be located through the subject index of the appropriate catalogue in the library.
Scholarly Books, which are written on many topics found in textbooks, may also be good
secondary source.
Identify key Terms
Once a research of secondary sources is completed and the research problem is refined, the
investigator needs to make a list of key terms that represent the problem.
This step is accomplished by identifying the most important terms in the problem and then
thinking of other terms that are closely related. These terms are then located in appropriate
indexes to find related literature. The indexes organise the literature by subject, title, key
terms, and author. For example suppose your research problem is related to the teaching
styles and student achievement, key terms are "teaching style" and "student achievement".
Identifying the Appropriate Journal Indexes and Abstracts Many
indexes may be used to locate research on education. The most useful and comprehensive
is "Current Index to Journals in Education (CIJE)" CIJE is published monthly by the
Educational Resources Information Centre (ERIC), CIJE provides abstracts of articles from
almost 800 journals and periodicals.

Search Indexes for Primary Sources

The ultimate objective in a review of related literature Is to Identify primary sources that
investigate a problem or idea of interest. A useful primary source is reading a book as it was
originally written by the author.
Summarise and Analyse Primary Sources
Once you locate the primary source, the article or report that contains the original data
collection and analysis, you will need to read it and summarise the information it contains.

Organising the Review

The review of literature can be organized in different ways. The most common approach is
to group together studies that investigate similar topics and subtopics. The review should
be exhaustive. Generally, the review chapter should be organized into three main parts.
These are the introduction, body of the review and summary of the review. The
introduction part explains briefly the nature of the research problem and the structure of the

39
review chapter. It is useful to present the outline of the review in the introduction. The
body of the review reports what others have found or thought about the research problem.
Related studies are usually discussed together and grouped under subheadings. Major
studies are discussed in detail while less important works are not emphasized. The review
should not be reporting what others found but it should show how results from similar
studies are similar or different. The summary of the review put together the main trends
and points of the review.

The Search for Primary and Secondary Sources

The search for primary and secondary sources is made in one of three ways.

1. By computer
2. By mediated online search done by a librarian or library staff
3. Manually

By Computer
Technological advances now allow libraries to have indexes on computers which
researchers can search by themselves. This can also be done through internet connections
of the library by the researcher.

By Mediated Online Search done by a Librarian or a Library Staff

This method is similar to the computer search except that librarian or library staffs typically
perform these searches. This is done to help the researcher to get additional information or
sources not available to all users such as interlibrary facilities.
Manually Search
This procedure involves the researcher personally going to the library to search for
materials to review for his/her literature. To conduct a manual search, you must select the
most recent available print index. Using the subject headings in the index, which are
organized alphabetically, find the key terms you have previously identified. The index will
list by title, published article that are related to the key term. Read through the list of titles
and determine which of the articles or books may be useful in the review.

CRITERIA FOR EVALUATING THE REVIEW OF LITERATURE

Several criteria should be considered, when reading and evaluating the review of literature
section of research studies.

1. The review of literature should adequately cover previous research on the topic. To
satisfy this criterion, questions that must be answered include: Were important studies
ignored or included? Do the numbers of studies in the review reflect research
activities in the area?

40
2. The review of literature should cite actual findings from other studies . It is important
for the review to be based on the empirical results of previous research, not on
opinions of others about previous research or on the conclusion of previous research.
3. The review of literature should be up-to-date. The studies reviewed should include the
most recent research on the topic. This does not mean that older studies are not
relevant.
4. The review of literature should analyse as well as summaries previous studies. The
analysis may be a critique of methodology or inappropriate generalisation, an
indication of limitations of the study e.g. to certain population, instrument or
procedures, or a discussion of conflicting results.
5. The review of literature should be organised logically by topic and not by author. A
review that has one paragraph for each study usually fails to integrate and synthesize
previous research.
6. The review of literature should briefly summarise minor studies and discuss in detail
major studies. Minor studies are those that are related to one or two aspects of the
study. Major studies are directly relevant to most aspects of the study or have important
implications.
7. The review of major studies should relate previous studies explicitly to the research
problem or methods. The questions that one should ask here are: what is the
implication of the review? How can it be helpful and improve subsequent research?
8. The review of literature should provide .a logical basis for the hypothesis. If there is a
hypothesis, it should be based on the review. There should be a connection among the
problem, review and hypothesis.

9. The review of literature should establish a theoretical framework for the problem. For
basic and most applied research, the review should provide the theoretical context for
the study, so as to enhance the significance of the findings.
10. The review of literature should help establish the significance of the research.

WRITING THE REVIEW CHAPTER

It is important that one undertakes a literature review on a subject before actually s
conducting a research project. Having gone through the steps involved in identifying a t

literature review which includes identifying potential references, locating references,

reading relevant references and taken notes, organizing, analyzing and integrating the
references; the final step is to write the review. This session presents what should go into
the review chapter and how to organize the materials.
Content of the Written Literature Review

A written review of the literature should be neither a series of quotes nor a series of
abstracts. The central task is to organise and summarize the references so that they reveal

41
the current state of knowledge on the selected topics and in the context of a new study, to
lay a systematic foundation for the research. The review should point out both
consistencies and contradictions in the literature as well as other possible explanations for
the inconsistencies, for example different methods.

As has been mentioned earlier, studies that are particularly relevant should be described in
some detail. However, reports that result incomparable findings (i.e. similar findings from
different authors) can often be grouped together and briefly summarised. For example: "A
number of studies have that academic achievement is highly related to socio economic
background of the student (Ogoe, 1993, Bansah, 1994; Aboagye, 1998).

It is also important to paraphrase or summarise a report in one’s own words. The review
should demonstrate that thoughtful consideration has been given to the material. Another
point to bear in mind is that the review should be as objective as possible. Studies that
conflict with personal values should not be omitted. It is not unusual to find studies with
contradictory results. The review should not deliberately ignore a study simply because its
findings contradict other studies.

Parts of the Literature Report

After reading and taken notes on the various sources collected, researchers can prepare the
final review. Fraenkel and Wallen (1993) have identified four parts of the literature review.
These include:

1. The introduction

The introduction briefly describes the nature of the research problem and states the
research question. The researcher also explains in this section what led him/her to
investigate the question and why it is an important question to investigate

2. The Body

The body of the review briefly reports what others have found or thought about the
research problem. How is the related literature discussed in the review? Related studies
are usually discussed together, grouped under subheadings to make the review easier to
read. Major studies are described in more detail, while less important work can be
referred to in just a line or two. In most cases this is done by referring to several studies
that reported similar results in a single sentence as shown at the previous sub session.

3. The Summary

The summary of the review ties together the main threads revealed in the literature
reviewed. Have you met the expression, "the thread of a book"? Hold on. The thread of a
book is the main line of reasoning connecting the parts of an argument or finding. The
summary also presents a composite picture of what is known or thought to date.

42
Findings, at the summary section, may be tabulated to give readers some idea of how many
other researchers have reported identical or similar findings or have similar
recommendations

4. References

The last part is a reference with full reference data for all sources mentioned in the review.
There are many formats that can be used to list references, but the format used by the
American Psychological Association (APA) 1983 is the accepted format for the faculty of
education, UCC.

For instance the accurate reference for a book entitled, "Nursing Research, Principles and
Methods" authored by Denise F. Polit, Ph.D and Bernadette P. Hungler Ph.D., a fifth
edition published at Philadephia by JP Lippincott Company in 1995, will be:

Polit, D. F & Hungler, B. P. (1995). Nursing research: Principles and methods. (5th ed.)
Philadelphia: J.B Lippincott Company.

UNIT 4
RESEARCH DESIGNS

Descriptive Design: Nature and Characteristics

Definition of Research Design

Let us consider this question. What do researchers do to obtain answers to the research
questions or test-hypotheses they formulate?

A researcher's overall plan for obtaining answers to the research questions or for testing the
research hypotheses is referred to as the research design. Research design is thus a plan or
blueprint that specifies how data relating to a given problem should be collected and
analysed.

It provides the procedural outline for the conduct of any investigation. What we mean here
is that the research design spells out the basic strategies that the researcher adopts to
develop information that is accurate and interpretable. The design, therefore, incorporates
some of the most important methodological decisions that the researcher makes in
conducting a research study.

Gay (1992) notes that, the research design indicate the basic structure of the study; the
nature of the hypothesis and the variables involved in the study. The design thus indicates
whether there is an intervention and what the intervention is, the nature of any comparisons
to be made, the method to be used to control extraneous variables and enhance the study's
repeatability, the timing and frequency of data collection, the setting in which the data
collection is to take place, and the nature of communications with subjects.

43
The Descriptive Research Design

Definition

Descriptive research is research which specifies the nature of a given phenomena. It

determines and reports the way things are. Descriptive research, thus, involves collecting
data in order to test hypotheses or answer research questions concerning the current status
of the subject of the study (Gay, 1992). i

Purpose of Descriptive Research Design

The purpose of descriptive research is to observe, describe, and document situation as it

naturally occurs. It sometimes, serves as a starting point for hypothesis generation or theory
development. In descriptive research, the events or conditions either already exist or have
occurred and the researcher merely selects the relevant variables for an analysis of their
relationships. Examples are "the perception of UCC students of the new accommodation
system (in-out-out-in) at the university" and "the attitude of primary school teachers in the
Cape Coast Municipality towards the introduction of the new English syllabus". In all these
examples, the some conditions do exist and the researcher is to observe and report as it is.

Nature and Characteristics of Descriptive Research Design

Descriptive research, as we saw in the definition, is generally concerned with the present
status of a phenomenon. According to Best and Khan (1998), descriptive research is
concerned with the conditions or relationships that exist, such as determining the nature of
prevailing conditions, practices and attitudes; opinions that are held; processes that are
going on; or trends that are developed. Amedahe (2002) also maintains that in descriptive
research, accurate description of activities, objects, processes and persons is the objective,
He also noted that research is not a fact finding per se. In fact, there is considerably more
to the descriptive research than just asking questions and reporting answers. What then
does the descriptive research deal with? Hold on. Don't answer. Now read on It deals with
interpreting the relationship among variables and describing their relationships.
Descriptive research seeks to find answers to questions through the analysis of relationships
between or among variable.

The descriptive research studies have all the following characteristics,

a. They may either be quantitative or qualitative. Thus, they may involve hypothesis
formulation and testing or formulation of questions and seeking answers to them or
describing situations.

b. They use the logical methods of inductive - (the process of reasoning form specific
observations to more general rules) deductive (the process of developing specific
predictions from general principles) reasoning to arrive at generalizations,

44
c. They often employ the method of randomisation so that error may be estimated when
population characteristics are inferred from observations of samples.

d. Variables and procedures are described as accurately and completely as possible so

that the study can be replicated by other researchers.

e. They are susceptible or easily influenced to distortions through the introduction of

biases in the measuring instruments, etc. For example errors due to the use of
questionnaires or interviews might distort research findings.

f. There is the need to organise and present data systematically in order to arrive at valid
and accurate conclusions.

Steps Involved in a Descriptive Research Design

1. Clear definition of problem or question to be answered.

2. The method of data collection should then be selected. The alternative methods
available for collecting desired data in the descriptive research include administering
questionnaire, interviewing subjects, observing events, or analysing documentary both
primary and secondary sources.

3. Selection of sample for the study. A meaningful and representative sample should be
selected with regard to the population characteristics such as size, composition, and also
with regard to data analysis procedures. For example in most quantitative studies, a sample
size of 5% to 20% of the population size is sufficient for generalisation purposes.

4. Data collection then follows. It is necessary to check for completeness, validity,

reliability and comprehensiveness of data.

5. Editing, coding and analysis of data using the relevant procedures. For example, one can
develop frequency and percentage tables, and compute mean and standard deviations as
well as test hypotheses using the t-test and chi-square.

6. Interpretation of Results. Results of the analysis must then be interpreted to show

whether a solution to the problem has been found or not.

Forms of Descriptive Research

U
Descriptive research ranges from simple surveys that do little more than ask questions and
report answers about the status of something (phenomena) to studies that present explicit
statements about relationships between variables. 4

DESCRIPTIVE DESIGN: POPULATION, SAMPLE AND SAMPLING

One of the most significant issues investigators have to consider when designing a project
concerns the type an number of the respondents who will be included in the students. A

45
number of very important questions have to be answered such as: Will the
whole population of a sample be studied? If sampling is preferred, which sampling
procedure is most suitable? How large should the sample be? Is there a sampling frame
required? If yes, is one available? How representative should the sampling be? How will
possible problems, errors, and distortions be prevented? Are the required time, funds and
staffing available and if so how can they be rationally employed?

Population

In a study, an investigator may be interested in a group of people. The group becomes the
investigator's target group. Examples of such groups may include " "all
teachers", "all senior secondary school students", "all principals in training colleges in
Ghana, "all District Education Office staff in Arnansie West District of Ashanti", "all State
Registered Nurses at the Cape Coast Hospital", "all lower primary pupils in 'Takoradi
Metro". The target group about which the researcher is interested in gaining information
and drawing conclusions is what is known as the population. Polit and Hungler (1996)
defined a population as the entire aggregation of cases that meet designated set of criteria.
It must be noted that whatever the basic unit, the population always comprises the entire
aggregation of elements in which the researcher is interested.

Accessible and Target Population

It is often useful to make a distinction between the target and accessible population. The accessible
population is the aggregate of cases that conform to the designated criteria that are accessible to the
researcher as a pool of subjects for a study. The target population is the aggregate of cases about
which the researcher would like to make generalizations. Thus, the target population is the units for
which the information is required and actually studied. A target population may be all teachers in
basic schools in a region; however, the accessible population may be teachers in basic schools of
a particular education district. Researchers usually sample from an accessible population and hope
to generalize a target population.

Samples and Sampling

In many cases a complete coverage of the population in a study is not possible. What then can the
researcher do? The way out here would be to select a unit of the population to represent the whole,
This will then lead us to the process of sampling.

Definition of Sampling and a Sample

Sampling refers to the process of selecting a portion of the population to represent the entire
population. Generally, sampling enables the researcher to study a relatively small number of units
in place of the target population, and to obtain data that are representative of the whole target
population.

A sample consists of a carefully selected subset of the units that comprise the population. In most
cases researchers opt for incomplete coverage and study only a small proportion of the population.

46
This small proportion of the population is the sample. Sampling is thus the process of choosing the
units of the target population which are to be included in the study.

In sampling, the units that make up the samples and population are referred to as elements. The
element is the most basic unit about which information is collected. In social research, the elements
are usually humans.
C
Reasons for Sampling

Opting for a sample survey is guided by a number of factors. Sarantakos (1998) has identified the
following as the most common.

In many cases a complete coverage of the population is not possible.

Complete coverage may not offer substantial advantage over a sample survey.

Sampling provides a better option since it addresses the survey population in a short period of time
and produces comparable and equally valid results.

Studies based on samples require less time and produce quick answers.

Sampling is less demanding in terms of labour requirement, since it requires a small portion of the
target population.

It is thought to be more economical, since it contains fewer people and requires less printed
material, fewer general costs (travelling, accommodation, etc) and fewer experts

Samples are thought to offer more detailed information and a high degree of accuracy
because they deal with relatively small numbers of units.

Principles of Sampling

Samples are expected to be representative. In other words, the sample is selected such that
it will be representative of the population. For this reason, samples are chosen by means of
sound methodological principles. With regard to quantitative research, Sarantakos (1998)
identifies the following as the most important.

Sample units must be chosen in a systematic and objective manner.

Sample units must be easily identifiable and clearly defined.

Sample units must be independent to each other, uniform and of the same size, and should
appear only in the population.

Sample units are not interchangeable; the same units should be used throughout the study.

Once selected, units cannot be discarded.

47
The selection process should be based on sound criteria and should avoid errors, bias and
distortion.

TYPES OF SAMPLING

There are a number of procedures by which a researcher can select his/her sample to
ensure its representativeness. The specific procedure that may be employed may be guided 1
by the following:

i. The structure of the population of study.

ii. The distribution of certain characteristics within the population.

iii. Other relevant factors available or present in the population that may threaten the
validity of the study.

Sampling plans can be grouped into two categories. These are Probability Sampling and
Non-Probability Sampling.

Probability (Random) Sampling

Probability sampling involves some form of random selection in choosing the elements in
such a way that the researcher knows the probability of selecting each member of the
population. It employs strict probability rules in the selection process. Thus every member
or unit of the population or subgroup has an equal, calculable and non-zero probability of
being selected for the sample. A random selection of the sample limits the probability that
one chooses a bias sample. Indeed, a random selection is a strategy for controlling threats
to external validity. Probability sampling is the more respected of the two approaches
because greater confidence can be placed in the representativeness of probability
samples,

Non-Probability Sampling
In non-probability sampling, elements are selected by non-random methods. There is no
way to estimate the probability that each element has the chance of being selected in a non-
probability sample. Usually, not every element in the population has a chance of being
selected. This method is less strict and makes no claim for representativeness. It is
generally left up to the researcher or the interviewer to decide which sample units should
be chosen. It is generally employed in exploratory research observational research and
qualitative research.

Types of Non-Probability Sampling

Non-probability sampling is less likely than probability sampling to produce accurate and
representative samples. Despite the disadvantage that arise from their non-
representativeness, they are far less complicated to set up, are considerably less
expensive, and can prove perfectly adequate where researchers do not intend to generalise

48
their findings beyond the sample in question or where they are simply piloting a survey
questionnaire as a prelude to their main study. The major kinds of non-probability
sampling are as follows.

Convenience Sampling
Quota Sampling
Purposive Sampling
Snowball Sampling

1. Convenience Sampling
Convenience sampling involves choosing the nearest or available individuals to serve as
respondents and continuing the process until the required sample size has been obtained.
When do we employ this type of sampling? Write your response in your jotter. Compare
your answer to the following: This type of sampling is employed in qualitative research
and in other studies where representativeness is not an issue.

It is also known as "accidental sampling", "chunk sampling", "grab sampling", and

"haphazard sampling". When this sampling technique is employed all it’s of study that
the research accidentally comes across or in contact with during a certain period of time
are considered. The investigator may stand at a street corner, in front of a shopping centre
or at a university gate and interview a certain number of people passing by between, say,
11.00 and 12.00 o'clock on certain days of the week. Similarly, captive audiences such as
pupils or student teachers often serve as respondents in surveys based on convenient
sampling. For example, the faculty member who distributes questionnaires in his/her class
is using a convenience sample. Convenience samples like the example above are easy to
construct and evaluate.

The problem with convenience sampling is that available subjects might be atypical
(different from normal/usual) of the population with regard to the critical variables being
measured.

2. Quota Sampling
Quota sampling has been described as the non-probability equivalent of stratified sampling.
(Stratified sampling would be discussed under probability sampling). It attempts to obtain
representatives of the various elements of the total population in the proportion in which
they occur there. In simple terms, the researcher sets the `quota' of respondents to be
chosen from specific population groups by defining the basis of choice (gender, marital
status, ethnicity, education, etc.) and by determining its size (e.g. 60 parents of mentally
challenged children, 35 policemen, 66 graduate teachers, etc). The choice of the actual
respondents is usually left up to the interviewer.

Ways of Determining the Quotas

49
The way quotas are determined varies depending on a number of factors related to the type
and nature of the research.

For instance the researcher might advise the interviewer to survey 50 female students who
will attend tomorrow's lecture and 50 male students who will walk out of the lecture hall
immediately the lecturer closes or 20 girls who will make 90% less attendance during the
I
month of June etc.

Another way of choosing the respondents is based on strict procedures. One such method is
by choosing respondents according to their proportion in the entire population.

Relationship between Quota and Convenience Sampling

Except for identification of the strata and the proportional representation for each, quota
sampling is procedurally similar to convenience sampling. Despite its problems, quota
sampling is an improvement over convenience sampling.

3. Purposive Sampling
In purposive sampling, researchers handpick the cases to be included in the sample on the
basis of their judgment of their typicality or particularly_ knowledgeable about the issues
under study, Thus, in purposive sampling, also known as judgmental sampling, the
researcher purposely chooses subjects who in their opinion are thought to be relevant to the
research topic. In this case the judgment of the investigator is more important than
obtaining a probability sample. Purposive sampling is often used when the researcher wants
a sample of experts as in the case of a need assessment using the key informant approach.

How are subjects sampled in this type of sampling? The process of sampling in this case
would involve identification of the i n f o r m a n t s and arranging times for meeting them.

4. Snowball Sampling
In this type of sampling, researchers begin the research with the small number of individuals
who have the characteristics that they require and are available to them. The people are
then used as informants to identify others who qualify for inclusion and these (i.e., those
identified) in turn identify yet others. In other words, the researcher subsequently asks the
few respondents available to recommend other persons who meet the criteria of the
research and who might have been willing to participate in the project. If and when such
respondents are recommended, the investigator approaches them, collects the information
required and asks them to recommend other persons who might fit the research design
and willing to be studied. The process continues until the topic is saturated, that is, until
no more substantial information is achieved through additional respondents, or until no
more respondents are discovered.

When is the appropriate time to employ this method? This method is employed when lack
of sampling frames make it impossible for the researcher to achieve a probability sample
the target population is unknown; it is difficult to approach the respondents in any other
way.

50
Probability Sampling
Majority of social researchers employ probability sampling for several reasons, among
which are:

Its high reliability

Its high degree of representativeness and high generalisability of the results

The hallmark of probability sampling is the random selection of elements from the
population. What is a random selection process? A random selection process is one in
which each element in the population has an equal, independent chance of being selected.
We shall discuss the four most commonly used probability sampling methods.-This
includes: Simple random sampling; Systematic sampling; Stratified sampling; and cluster
sampling

Simple Random Sampling

Simple random sampling is the most basic of the probability designs. This type of sampling
gives all units of the target population an equal chance of being selected. The simple
random sampling is appropriate when a population of study is similar in characteristics of
interest. The sample units are selected by means of two main methods or strategies. These
are the lottery method and the random number method.

The Lottery Method

Choosing respondents by the lottery method follows a procedure that can be described in the
following steps.
Step 1: Identify or construct a sampling frame. Constructing a sampling frame is preparing
a list of the units of the target population, that is, the list of all elements in the population.
Such frames are, for instance, the electoral register in the area, students’ records and
include names and if possible addresses of sample units in alphabetical order and numbered
accordingly.
Step 2. Write names listed in sample from or their numbers on slips of paper and put in a
container.
Step 3: Mix well and remove one slip or paper at a time from the container without looking
into it
Step 4: Register or record the number or name on the slip. When a slip is selected and
recorded, it is thrown back into the container before the next one is picked. The process is
continued until the required number of respondents is recorded. If an already drawn number
is selected for a second or third time it is ignored, that is, it is thrown back into the
container.
ii. The Random Numbers Method

51
This method is similar to the lottery method, except that the container and strips are replaced
by already generated tables of random numbers. These are usually found at the back of
statistical textbooks (See Appendix B).
Choosing the sample by using the random numbers method involves the following steps.
Step 1: A sampling frame is identified or constructed as in the lottery method. Thus decide
on the sample size.
Step 2: Appropriate tables of random numbers are selected
Step 3 Numbers are picked from the tables randomly and registered, the name in the
sampling frame corresponding to the numbers constitute the sample. For example, assume
that we are interested in the attitudes of JSS teachers in Cape Coast Municipality to the
Best Teacher Award scheme in Ghana. A sample of 100 teachers are required cut of 5000
JSS teachers in the municipality. ' The following steps would then be followed.

A list of all JSS teachers' names (from the sampling frame) is obtained, numbered and
ordered accordingly.

A list of random numbers that contains all numbers included in the sampling frame (in our
example 0000-4999) is used.

The table is entered randomly, where the researcher proceeds vertically or horizontally
using the appropriate number of digits. Here the researcher starts at any point on the table
of random numbers.

With our 5000 teachers, the number of digits to be counted each time is four. Anytime the
researcher encounters numbers falling within the sampling frame the specific element
with that number is selected and recorded. For example, the four-digit numbers 1234 and
0069 are within the sampling frame. In fact, they are for the individual numbered 1234
and 69 and must be selected. However, the four digit number 6934 and 5001 are rejected
because there is no element in the population with this number. The process is continued
until the required number of teachers is achieved.

A number of additional techniques that have been devised to identify respondents in the
simple random sampling which you may meet in the course of your reading include, the
computer method, birthday, first letter and pin number.

It is important to note that the simple random sampling is appropriate when the population
of study is similar in characteristics of interest. Can you think of one problem associated
with this technique? One problem associated with this sampling method is that a
complete list of population is needed and this is not always readily available.

Systematic Sampling
This method is a modified form of simple random sampling. It involves selecting the
subjects from a population list in a systematic rather than a random fashion. A systematic
sampling consists of the selection of Kth term from a list of all elements in a population

52
beginning with a randomly selected element. For example, if there is the need to select
100 subjects from a population of 50,000.

Step1. Select/compute the sampling fraction (the Kth term) which is N/n where N is the
I
number of units in the target population and n the number of unit of the sampling. In the
t
example above N=50,000 n=100; sampling fraction is Kth = 50,000/100 = 500th. Every nth
would therefore correspond to 500 th subject/element.

Step 2. Select the first element randomly using lottery or random numbers. In this example
that would be some number between I - 500.

Step3 .Identify all numbers between 0 - N that result from adding Kth to the random
number. The process is repeated until N is reached. Suppose 240 was randomly selected as
a starting point, in step 2, the first sample chosen would be the 240 th number the next
subject would be the 740 th i.e. 240+500, the third would be 1,240 i.e. 740+500 etc.

Locate the names on the sampling frame that corresponds to the number drawn above,
example the names corresponding to 240 th, 740th etc.

These names correspond to the respondents who constitute the sample.

Systematic sampling and simple random sampling differ in that in the latter the selections
are independent of each other. In systematic sampling the selection of sample units is
dependent on the selection of a previous one.

Stratified Random Sampling

Stratified sampling involves dividing the population into a number homogeneous groups or
strata. Each group contains subjects with similar characteristics. A sample is then drawn
from each group or stratum. The sub samples make up the final sample for the study. The
division of the population into strata or homogeneous group is based on one or more
criteria, for example, sex, age economic status, professional status. For instance, group A,
or the stratum A might contain males and group B or stratum B females. In order to obtain
a sample representative of the whole population in terms of gender or sex, a random
selection of subjects from group A and B must be taken. The sample size can be
proportionate or disproportionate to the units of target population. For example, the target
population, consisting of equal number of males and females, may be divided into two
strata or homogeneous groups’ males and females.

Then if a proportionate stratified sample is drawn, 5 percent from each group may be
taken. If the researcher decides for a disproportionate sample, 5 percent of males and 10
percent of

Step I; Identify the stratification variable. The target population is divided into a number
of strata/homogeneous groups according to the number of the significant groups in the
population

53
Step 2: The sampling frames for each of the groups are identified. If not
available, relevant sampling frames must be developed.

Step 3: Employing one of the methods• discussed above simple random Sampling,
systematic etc, a sample is drawn from each group. This can be proportionate or
disproportionate to the number of units in the population

Step 4: The individual samples are merged into one, and this constitutes the sample for the
study.

Cluster Sampling

When population is large and widely dispersed, gathering a simple random sample poses
administrative problems. Cluster sampling is used when it is impossible or impractical to
sample individual elements from the population as a whole or when there is no exhaustive
list of all elements. In cluster sampling, there is a successive random selection of
naturally occurring groups or areas and then selecting individual elements from the
chosen group or areas. Thus, a characteristic of this sampling method is that first groups
of elements (clusters) are selected (e.g., schools, classes, suburbs) and then individual
elements are selected from these clusters. To choose the clusters and the respondents from
the clusters one of the methods discussed above can be employed.

Cluster sampling is considerably more economical. and practical than other types of
probability sampling, particularly when the population is large and widely dispersed.

Observation

Natural and Laboratory observation

The main difference between these two techniques lies in the type of setting in which they
unfold. In the natural observation, observation takes place in natural settings. In the
laboratory, they are performed in the laboratory.

Open and Hidden

This distinction refers to the degree to which the identity of the researcher as an observer as
well as the purpose of the study is known to the participants. In the case of the open
observation, the participants are well informed of the nature of the study and the identity
of the researcher but in hidden observation they are not.

Active and Passive Observation

This type of observation refers to the degree to which the observer is involved in the
process and purpose of observation. Active observation presupposes full engagement of
the observer in the course of the study. In passive observation, the role of the observer is
seen as being just a strict recording of data. In this ease observation is a job to be done in
an objective and neutral fashion.

54
Direct and Indirect Observation

Direct observation studies the subjects it intends to explain, for example, if the study intend
to explain the patterns of conflicts in marital homes, and observation involves named
couples, this is a direct observation. Indirect observation does not involve the subject of
study, either because the subject refuses to take part in the study, a direct observation is
not possible, or the subject is deceased. Instead, the observer/researcher observes the
physical traces the phenomena under study that the subject has left behind and make
conclusions about the subject.

The Process of Observation: Steps in observation

Observation takes place in the same form as the general research introduced earlier in this
book. The content, however, depends on the nature of observation. The following is a
brief summary of the basic steps of research as employed in the area of observation

Selection of Topic

This step includes the issues to be studied through observation. This must be an observable
social phenomenon of any level

Formulation of the Topic

This involves a specific definition of the topic, development of observation categories,

establishment of the observer-subject relationship, the type and explanation of what is to
be observed

DESCRIPTIVE DESIGN: INSTRUMENTS AND PROCEDURE

Explanation of Data and Data Collection

According to Gay (1992), all research studies involve data collection. Since all studies
are designed to either test hypotheses or answer research questions, they all require data
with which to do so. Let's now look at what data is. The data of a study are the pieces of
information obtained in the course of investigation.

What then is data collection? Data collection refers to obtaining relevant information,
either verbal or quantitative regarding the major ideas of the hypotheses or research
questions for the purpose of answering them.

Data Collection Techniques

In a descriptive research data may be obtained through a variety of techniques. Can you
name these, techniques? Write any two in your jotter. Compare what you wrote to the
following.

i. Observation
ii. Questionnaire

55
 iii. Interview
iv. Tests
v. Attitude scale
Observation

Observation is one of the oldest methods of data collection. Literally, observation means a
method of data collection that employs vision as its main means of data collection. In
observational studies researchers collect data on the current status of subjects by watching
them and listening and recording what they observe rather than asking questions about
them. An observation may take many forms/types.

Types of Observation

There are several types of observation; some are more popular than others. Some
types/forms of observation are Participant and Non Participant observation.

In general, the degree of the observers’ involvement in the observation varies from no
participation to full participation. In non-participant observation, observers study their
subjects from outside the group without becoming a part of the environment of the
observed. In, participant observation they actually become members of the group they are
supposed to be studying. They observe from inside the group and, ideally , their identity as a
researcher is not known.

By participating actively in a group, one may gain insight into the structure of the group
and obtain certain types of information not accessible in any other way. On the other hand,
by becoming emotionally involved the researcher may lose objectivity.

1. Structured and Unstructured

Structured observation employs a formal and strictly organised procedure, with a set of well-
defined observation categories. It is subject to high levels of control and differentiation. It is
organised and planned before the study begins. Unstructured observation is loosely
organised and the process of observation is largely left up to the observer to define.

2. Natural and Laboratory observation

The main difference between these two techniques lies in the type of setting in which they
unfold. In the natural observation, observation takes place in natural settings. In the
laboratory, they are performed in the laboratory.

3. Open and Hidden

This distinction refers to the degree to which the identity of the researcher as an observer as
well as the purpose of the study is known to the participants. In the case of the open
observation, the participants are well informed of the nature of the study and the identity
of the researcher but in hidden observation they are not.

56
4. Active and Passive Observation

This type of observation refers to the degree to which the observer is involved in the
process and purpose of observation. Active observation presupposes full engagement of
the observer in the course of the study. In passive observation, the role of the observer is
seen as being just a strict recording of data. In this ease observation is a job to he done in
an objective and neutral fashion.

6. Direct and Indirect Observation

Direct observation studies the subjects it intends to explain, for example, if the study intend
to explain the patterns of conflicts in marital homes, and observation involves named
couples, this is a direct observation. Indirect observation does not involve the subject of
study, either because the subject refuses to take part in the study, a direct observation is
not possible, or the subject is deceased. Instead, the observer/researcher observes the
physical traces of the phenomena under study that the subject has left behind and make
conclusions about the subject.

Selection of Topic

This step includes the issues to be studied through observation. This must be an observable
social phenomenon of any level

Formulation of the Topic

This involves a specific definition of the topic, development of observation categories,

establishment of the observer-subject relationship, the type and explanation of what is to
be observed.

Research Design

Here, the researcher will determine the subjects to be observed, select the setting for
observation, that is, the time when observation will be carried out, place in which
observation will be conducted, type of event to be studied- the researcher observing
everything, some events, routine events and subjects or persons to be observed.

Collection of Data

This involves familiarization with the setting and subjects, initial interaction, observation
and recording. Recording of data is an issue that must be considered during the planning
stage of the research. This includes three issues, namely what will be recorded, when and
how.

This refers to the methods of recording, the events to be recorded and the coding.

The method of recording varies from one type of observation to another.

57
The most common methods of recording are: Writing down information verbatim in
summary or in key words; Tape recording conversations. Video recording events; Taking
photographs

Auditing of Data

At this stage, the researcher will undertake data reduction, presentation (e.g. in tables and
graphs), cross tabulation and interpretation.

Report Writing

This involves the writing of the reports to be published in some form.

Characteristics of Good Observation

Best and Kahn (1993) maintain that observation as a research data-gathering process
demands vigorous adherence to the spirit of scientific inquiry. The following standards,
they indicated, should characterize observers and their observation.

Observation is carefully planned, systematic and perceptive. Observers know what they are
looking for and what is irrelevant in a situation. They are not distracted by the dramatic or
the spectacular.

Observers are aware of the wholeness of what is absent.

Observers are objective. They recognise their likely biases, and they strive to eliminate
their influence upon what they see and report.

Observers separate the facts from interpretation. They observe the facts and make their
interpretations at a later time.

Observations are carefully and expertly recorded.

Observations are collected in such a way as to make sure that they are valid and reliable.

Advantages of Observation
i
a. It provides information when other methods are not effective,
b. It employs a less complicated and less time-consuming procedures of subject selection.
c. It can offer data when respondents Are unable and/or unwilling to correlation-operate
or to offer information.
d. It approaches reality in its natural structure and studies events as they evolve.
e. It offers first-hand information without relying on the reports of others,
f. It is relatively inexpensive.
Limitations of Observation

 It cannot be employed when large groups or extensive events are studied

 It cannot provide information about past, future or unpredictable events.

58
  It cannot study opinions or attitudes directly.
 It is a relatively laborious and time-consuming method.
 It cannot offer quantitative generalisations on the results.
Questionnaire

The use of questionnaire is very common in the social sciences, including education. In
most cases questionnaires are employed as the only method of data collection.

59
A questionnaire consists of a list of questions or statements relating to the aims of
the study, the hypotheses and research questions to be verified and answered, to which
the respondent is required to answer by writing.

Forms of Questionnaire

(i) The Close Ended Form

Questionnaires that call for short, check-mark (list) responses are known as restricted,
structured or close-ended type. Here you mark a Yes or No, rank alternatives provided
based on how one feels about the issue, check an item from a list of suggested responses
or scale responses.

The following is an example that illustrates a close-ended question.

How would you rate the adequacy of textbooks in your school library?
Excellent
Good
Fair
Poor
Other - please specify

The open form question or the unstructured or unrestricted question calls for a free
response in the respondents' own words.

Example: What is the textbook situation in your school library? The open from
generally provides for greater depth of response. The respondents reveal their frame of
reference and possibly the reasons for their responses. But because they require greater
effort on the part of the respondents, returns are often meager. Also, an open-ended item
can sometimes be difficult to interpret, tabulate and summarise in the research report.
Many questionnaires include both open and close-ended type items.

Types of Questions and Response Modes

Different types of questions, statements and response modes can be used in a

questionnaire. Questions asked in a questionnaire may be:

direct or indirect questions

specific or non specific questions

fact or opinion eliciting questions or statements.

Direct questions ask the respondents to offer information in a straightforward manner. It

also asks respondents to offer information about him- or her-self. An example of a
direct question is: Do you believe in God?

60
 An indirect question asks the respondent to offer information about other people
assuming that in this way the respondent will indirectly tell about himself or herself. An
example of an indirect question is: Do you think that people of your status and age
believe in God nowadays? Indirect question is mainly used when the question deals
with sensitive, embarrassing or threatening issues.

A specific question elicits information on a specific issue. (E.g. how are you?) In a non-
specific form the question is posed to seek more or less general information from the
respondent. For example, instead of asking the respondent the number of years of
teaching experience, the researcher can provide intervals of year’s range of respondents
from which they indicate the intervals in which their years of experience fall.

Rules of Questionnaire Construction

It is common practice for questionnaire to be constructed according to set rules and

standards. Why is it so? Write your response for FTF discussion.

Many writers, Amedahe, (2002); Sarantakos, (1998); Best and Khan, (1993) refer to the
rules listed below.

Layout

• Questions must be well presented in the questionnaire to make it easy to read and
easy to follow.

• Questions and response categories must be easy to identify and distinguish from
other questions and response categories. For this reason, sufficient space should be
provided between the questions.

Clear instructions regarding the way of answering the question must be given. For
example the respondent may be asked to circle the appropriate number or tick the
right box.

Sufficient space should be left for the respondent to make relevant remarks if required.

Content of the Questions

-Every question must be relevant to one or more aspects of the study.

-Ambiguous, non-specific and hypothetical questions are to be avoided.

-Leading, double-barreled and presuming questions should not be employed.

-Embarrassing, personal or threatening questions should be avoided.

61
-Vague words and academic jargons should not be used.

-The language of the respondent should be employed. If possible, a simple language

should be used, without jargon, slang or complicated expressions.

-Easy flow and logical progression in the questionnaire should be assured.

-Each question should ask what it is supposed to ask.

Questionnaire Format

The questionnaire must have a professional appearance and should give the impression
of a document that deserves respect and invoke feelings of responsibility.

The questionnaire should be presented in a way that encourages the respondent to

complete and return it.

Writing on one side of the page is, for smaller questionnaires, preferable to w r i t i n g
on both sides.

Print and colour of paper and ink must correspond with the preference of respondents.

The questionnaire should be presented as a complete document with an inviting and

reassuring introductory cover letter and a concluding note containing instructions
regarding the return of the questionnaire.

The questionnaire size should be kept to a minimum, and restricted to as few as

possible.

Sufficient instructions and probes should be provided where necessary.

Pre-coded questions should be checked for possible bias and ethical adequacy.

Questionnaire Administration

Questionnaire administration involves sending or presenting the questionnaire

prospective respondents and following it up. Amedahe (2002) identified 2 main ways
of a d m i n i s t e r i n g questionnaire. These are:

 through mail - mail questionnaire and

 Personally delivered to the respondent (non-mail questionnaire).
Table 2 presents the characteristics assumptions, advantages and disadvantages of the
two types as presented by Amedahe (2002).

Table 2: Characteristics of Mail and Non-Mail Questionnaire

62
Mail Questionnaire
Non-Mail Questionnaire
It is sent by post to the respondents. The success of
the administration depends on the existence of an It is sent personally by the
efficient postal system. researcher or the assistant to the
prospective respondent
The questionnaire is filled in the absence of the
researcher. It therefore presupposes that all
respondents can read and write. Also it assumes that
there will be no need for explanations and It may be filled in the presence of
translations. the researcher or his assistant.
When it is filled in the presence of
The administration is ratter impersonal in the sense the researcher or the assistant the
that the researcher will not have any personal explanation of issues of the
contact with the respondent beyond the cover letter. respondents is possible.
This can be advantageous in that the respondent's
responses will not be influenced by the presence of
The administration is personal to
the researcher.
some extent since the
The administration is liable to non-response. The questionnaire will be delivered
non-responses accounts for the difficulty in personally to the prospective
obtaining a high return rate in mail questionnaires. respondent. The personal
This is partly due to the impersonal nature of the delivery can be advantageous
administration even though intensive follow-ups depending upon how the
may be done. presence of the researcher can
influence the response of the
respondent.
,
Non-responses can be controlled
to some extent by the researcher
by making appointment with the
respondent to collect the
questionnaire.

63
 Advantages and Disadvantages of Questionnaire

Questionnaires as methods of data collection have strengths and weaknesses and thus
advantages and disadvantages that the researcher must be aware of.

Strengths and weaknesses are factors that have a significant impact on a researcher's
decision about whether or not to use questionnaires in the study. The advantages and
limitations, which most researchers and writers consider as significant are listed
below.

Economy

Questionnaires are less expensive than other methods - interview and observation. It
can be sent through mail interview and observation cannot hence the expenses and
time involved in training interviewers and observers as well as time spent on
interviewing and observation are not involved in the use of questionnaire.

Reachability

The use of questionnaires promises a wider coverage since researchers can approach
respondents more easily than other methods. They are not affected by problems of
no-contacts.

Uniformity in Questions
-Questions are a stable consistent and uniform measure, without variation
-Can be completed at the respondent's convenience
-They offer greater assurance of anonymity.

Limitations

Limitation to Literate Population

The questionnaire as a tool for data collection cannot be used/ administered on
illiterates, and people who are too young to read and write.

Motivation of the Respondents

The use of questionnaire does not offer opportunities for motivating the respondent to
participate in the survey or to answer the questions.

Sampling
The characteristics of non-response associated with questionnaire especially the mail-
questionnaire is likely to affect the representativeness of the sample. This may result
in a biased final sample.

Lack of opportunity to collect additional information

64
Questionnaires do not provide an opportunity to collect additional information
through observation, probing, prompting and clarification of questions while they are
being completed.

Interviewing
An interview is a form of questioning characterised by the fact that it employs verbal
questioning as its principal technique of data collection. It involves posing questions to
respondents for answers in a face-to-face situation . It also represents direct attempt by
the researcher to obtain reliable and valid measures of characteristics behaviours,
attitudes, etc. in the form of verbal responses from one or more respondents.

Types of Interviews

There are many types of interviews, each of which differs from the others in structure,
purpose, role of the interviewer, number of respondents involved in each interview, and
form and frequency of administration. These types, however, fall under two main
categories. These are structured and unstructured.

Structured interviews are interviews in which the specific questions to be asked and
the order of the questions are predetermined and set by the researcher. They are based
on a strict procedure and a highly structured interview guide, which is no different from
questionnaire. A structured interview is in reality a questionnaire read by the
interviewer as prescribed by the researcher. The r i g i d structure determines the
operations of this research instrument and allows no freedom to make adjustment to any
of its elements, such as contents, wording or order of questions. The instrument of
predetermined questions is known as an interview schedule. The interview is based on this
schedule, and strict adherence to the questions and the instructions is paramount. The
process of interviewing, using this type of interview, is referred to as scheduled
interview.

Unstructured Interview
Unstructured interviews, in contrast, have no strict procedures to follow of the kind described
above. There are no restrictions in the wording , of the question, the order of question or
the interview schedule. Thus, the interviewers have the discretion to form questions on the
spot, probe into issues and in some cases follow the order dictated by the situation. The
structure of these interviews is flexible and the restrictions minimal. An instrument used
in such an interview is called an interview guide. The guide is not as detailed as the
schedule. It merely outlines the kind of information required

Individual versus Group Interview

When individual interviews are employed the researcher interviews one respondent at a
time. This is the most common form of interviewing. The interviewer asks the various
types of interview questions, - direct and follow up questions, probing questions,

65
structuring and interpreting questions, - to the individual/one respondent who in turn
answers these questions.

In group interviews, more than one respondent is involved in the interview situation.
They are also addressed by the interviewer simultaneously. In a small scale group two
people are interviewed concurrently. On a large scale a group of about 20 persons
might be interviewed. For example, a husband and a wife might be interviewed
together on a small scale whiles a class of students might be interviewed together on
a large scale.

Skills and Qualities of Interviewers

The interviewers must possess certain skills and qualities to be able to conduct a
successful interview. List any 4 skills that the interviewer must possess to be able to
function effectively. Write your answer in your jotter for FTF discussion. Now read
on.

Amedahe (2002) has identified the following skills and qualities to be followed for a
successful interview session.

Knowledgeable
The researcher must be knowledgeable on the topic/theme he/she is working on, why
is this so? This will enable the interviewer to pursue relevant issues.

Structuring
An interviewer should be able to structure the interview - introduction, process and
closing. He/she should be able to give an introduction that will capture the attention
of the interviewees. In the process of the interview he/she should be lively and not
boring in order to sustain the interview throughout the session,

Clear
The interviewer should pose clear, simple, short and straightforward questions.
He/she should be able to speak distinctly and understandably and avoid academic and
professional jargons.

Gentle
Allows subjects to complete what they are saying and tolerate pauses and accept
unconventional and provocative opinions.
Sensitive
Listen attentively to the content of what is said. Be empathetic, if need be.
Steering
The interviewer should control the course of the interview and should not be afraid of
interrupting digressions from the interviewee.

Open

66
Approach the interview with an open mind. Must be open to others opinions. Accept
whatever opinions an interviewee expresses on an issue once he/she has not digressed.

Critical
Does not take everything that is said at the face value but questions critically to test
the reliability and validity of what the interviewees tell.
The interviewer should be able to retain and recall what interviewees tell and must
be able to relate them.

Interpreting
An interviewer must throughout the interview session try to clarify and extend the
meanings of the statements made by the interviewee for confirmation or
disconfirmation.

Training of the Interviewers

Once the researcher decides to gather data through the interviews, he/she has to
decide whether he/she could conduct the interview alone. If the researcher has so
many respondents to interview that he/she cannot conduct it alone, then the need to
select and train interviewers.

Note: Both telephone and face-to-face interviewers need to be trained beforehand.

Why is it necessary to train these interviewers? It is necessary to train the

interviewers because the purpose of conducting an interview is to obtain the desired
information with maximum efficiency and minimum bias. In fact the accuracy of the
data obtained depends on the level of skill of the interviewers and their knowledge
on the type on questions to be asked and when to ask the appropriate question, hence
the need for training.

At the training, the interviewers need to know how to explain quickly the purpose of
their call and why it is important to obtain information from the respondent.

They also need to learn how to question in a way so as to encourage those they call
to respond honestly.

The need to know how to establish rapport, that is friendly and cordial relationship
with their interviewees and to put them at ease. The interviewer needs to know when
and how to "follow up" on an unusual answer or one that is ambiguous or unclear

Interviewers also need training in gestures, manners, facial expression, and dress.

Procedure for contacting respondents and introducing the study

All interviewers should have a common understanding of the purpose of the study.

Procedure for recording answer/response to open ended and closed ended questions

67
Rules and guidelines for handling the interpersonal aspects of the interview in a non-
biasing way must be clear. Interviewers should be taught on how to focus on the task
at hand and to avoid expressing their personal views on any of the questions being
asked.

Recording the Responses

The answers to questions asked during the interview can be secured against the time the
results will be analysed in three modes/ways namely a by making a mental note of the
answers by making a written record of the answers by making a tape or video tape
recording of the interview.
Mental Note: This is the process where research workers rely on the power of their
memory for storing the information obtained during an interview.
Written Records: Nwana (1992) identified two forms of written records that can be
made for the responses. In the first form, to which he referred to as data-blank, the
questions to be asked are listed on paper with blank spaces provided for the answers to
be written in by the investigator or interviewer whiles the interview is in progress.
In the second form of recording which he referred to as checklist, the questions are
listed on paper and alongside each question, a set of most probable responses are
supplied. The interviewer records the responses of the respondent merely by ticking off
the appropriate responses as they are made.
Tape Recording
This may be sound tape recording which employs a conventional tape recorder to record
on magnetic tape only the sounds i.e. the questions posed by the investigator and the
answers given by the respondent.

It could also be videotape recording which records on magnetic tape both sounds and the
pictures of the interview.

It must be noted that the respondent needs to be informed during the time of
familiarisation that you intend to record the conversation between you and him/her and
his/her permission sought for doing so. The mode of recording should also be made
known to the respondent.

Advantages and Limitations of Interviewing

The popularity of interview is often justified in terms of several of its qualities which
give it an advantage over other methods of data collection. The advantages of
interviewing include the following:

Flexibility - interviews can be adjusted to meet many diverse situations

High response rate - interviewing attracts a relatively high response rate.

68
Easy administration - interviews do not require respondents to have the ability to read.
Opportunity to observe non-verbal behaviour - such opportunities is obviously not
available when questionnaires are used.
4
Capacity for correcting misunderstanding by respondents, more complex questions, can
be used because the presence of the interviewer can assist in correcting any
misunderstanding as well as assisting answering the questions.

Limitations
Despite its advantages, interviewing is limited by a number of factors. The following
are the most important:

Interviews are more costly and time consuming than other methods such as
questionnaires.

They are attacked for the "interviewer factor" and the possible bias associated with it.

They are more inconvenient than other methods such as using questionnaires.

It is less effective than other methods when sensitive issues are discussed. For
example many people prefer to write about sensitive issues than to talk about them.

It offers less anonymity than other methods since the interviewer knows the identity,
residence, type of housing, etc.

CORRELATIONAL DESIGN
The Nature of Correlational Research

Human behaviour at both the individual and social level is characterised by great
complexity. Given the present state of social research, man understands very little
about this complexity. One approach to a better and fuller understanding of human
behaviour is to begin by leasing out simple relationships between those factors and
elements deemed to have some bearing on the phenomenon in question. The value of
correlational research is that, it is able to achieve this end.

Before we attempt to describe correlational research, it would be useful to begin by

examining correlation and its related terms, then indicate the purpose they fulfill in
statistical analysis. We said in the introduction that the primary purpose of this is to
discover relationship and in some situations control their occurrence. You have read
above that must of social science research, in general and educational research in
particular, is concerned with establishing interrelationships among variables. For
example, w e wish to know, how delinquency is related to social class background; or
whether an association exists between the number of years spent in fulltime education
and subsequent annual income; or whether there is a link between personality and

69
 achievement. Correlation includes all such research projects in which an attempt is
made to discover or clarify relationships through the use of correlation coefficients.

Definition of Correlation Research

Correlation research attempts to investigate possible relationships among variables

without trying to influence those variables. In simplest form, correlational research
attempts to determine whether and to what extent or degree a relationship exists
between two or more quantifiable variables. Correlational research is sometimes
referred to as a form of descriptive research because it describes the relationship
between variables. The way it describes the relationships, however, is quite different
from the descriptions found in other types of studies. A correlation study describes the
degree to which two or more quantitative variables are related and it does so by the use
of correlation coefficient. If a relationship exists between two variables, it means that
scores within a certain range on one variable or measure are associated with scores
within a certain range of another measure.

Correlation does not necessarily establish cause-and-effect relationship in other words,

it must be noted that, the fact that there is a relationship between variables does not imply
that one is the cause of the other Correlation studies provide an estimate of just how
related two variables are. However, regardless of whether a relationship is a cause-and -
effect relationship, the existence of a high relationship permits prediction.

Purposes of Correlational Studies

Correlational studies are carried out for two main purposes:

-To explain important human behaviours and characteristics in quantitative terms and how
they relate

Correlation studies help to classify our understanding of important phenomena through

the quantification of relationships among variables. For example, correlations found
between the rate of language acquisition and the type of language of the child (i.e., first

Language or second language) had taught researchers much about how correlational
research help in prediction of events,

A second purpose of correlational research is that of prediction. If a relationship of

sufficient magnitude exist between two variables, it becomes possible to predict a
relationship on either variable if a score on the other variable is known. Researchers have
found, for example, that a student's grade at the BECE is highly related to the grade of the
SSSCE. Hence, BECE would be used to predict SSSCE grades. We would predict that a
person who has a high grade at BECE would be likely to have a high grade at the SSSCE.

70
The variable that is used to make the prediction is the predictor variable. On the other
hand, the variable about which the prediction is made is called the criterion variable. In
the example above, the BECE grade would be the predictor variable whiles the SSSCE
grade is the criterion variable.

Basic Steps in Correlational Research

1. Problem Selection

The first step in planning a correlational study is to identify specific variables that
appear to be important determinant(s) of the complex characteristics or behaviour pattern
being studied. What is the basis on which the variables included in a correlational
research chosen? Don't answer. Read on the variables to be included in a correlational
study according to Fraenkel and Wallen (2000) should be chosen based on a sound
rationale growing out of experience or theory, Generally, three major types of
problems are the focus of correlational studies.

a. Is the variable X related to the variable Y?

b. How well does variable E predict variable C?

c. What are the relationships among a larger number of variables and what
predictions can be made that are based on them

Almost all correlational studies would evolve around one of these types of questions.

Sample Selection

The sample for a correlational study and in any type of study, should be selected
carefully and, if possible randomly, Identify the steps in selecting the sample for a
study. Compare your response to the following. The first step in selecting a sample
is to identify an appropriate population. What is an appropriate population? Read on.
An appropriate population is the one that is meaningful and from which data on each
of the variables of interest can be collected. The minimum acceptable sample size
for a correlational study is considered by most researchers to be no less than 30.
Data obtained from a sample smaller than 30 may give an inaccurate estimate of the
degree of relationship that exists. Samples, larger than 30 are much more likely to
provide meaningful results.

Instruments

The instruments used to measure the two (or more) variables involved in a
correlational study may take one of a number of the forms discussed in the previous
session of this unit (i.e., questionnaire, test, checklist, etc). They must, however, yield
quantitative data. What is a quantitative data? Do you remember our explanation of
quantitative research? If not, go back and revise it. Now read on.

71
Design and Procedure

The basic design used in correlational research is quite simple and straightforward.
It involves collecting data on two or more variables on the same group of subjects
and computing/calculating the correlation coefficient(s). The correlation coefficient
obtained indicates the degree of relationship between two variables.

Data Collection

In correlational study, the data on both variables will usually be collected within a
fairly short time. The instruments used are often administered in a single session, or
in two sessions immediately after each other. Data for relationship study can be
collected by various methods including tests, questionnaire, interviews or
observational techniques.

Data Analysis and Interpretation

In correlational study, data analysis involves computing the correlation coefficient

between or among the variables of interest. When two variables are correlated, a
correlation coefficient is produced. The coefficient is an index of relationship.
Depending on the data collected, the researcher can compute the following using the
computer or manually.

a. Pearson's Product Moment Correlation (r)

By far the most common correlation coefficient in educational research is the Pearson's
Correlational Coefficient (Heiman, 1996). The Pearson's Correlational Coefficient is
used to describe the linear relationship between two variables that are both interval and
ratio variables. The symbol for the Pearson'% Product Moment Correlation is r. The
statistical basis for r is that it compares how consistently each value of Y pairs with
each value of X in a linear fashion. The Pearson's Product Moment Correlation
Coefficient can be computed using the computer. The manual computation would be
treated in the mathematics.

Spearman Rank Correlation Coefficient (ρ)

The Spearman rank-order correlation coefficient describes the linear relationship

between two variables measured using rank scores. That is the Spearman Rank
Correlation Coefficient is used for data involving ordinal or rank-order (first, second,
third, etc.)

The Biserial Correlation Coefficient

The point-biserial correlation coefficient is used when we want to correlate the scores
from a continuous interval, or ratio variable with the scores from a dichotomous
variable (scores which has only two categories are dichotomous variables). The point-

72
biserial correlation coefficient describes the linear relationship between the scores from
one continuous variable and one dichotomous variable.

Interpreting the Correlation Coefficient

Once a correlation coefficient has been computed, the next step is interpreting it. In
other words, we must understand what the correlation coefficient tells us, once we have
computed it.

Correlation techniques are generally intended to answer three questions about two
variables or two sets of data.

First, is there a relationship between the two variables or set of data? If the answer to
this question is "Yes" then the two other questions follow. What is the direction of the
relationship? i.e., if there is a correlation, is it negative or positive? And, what is the
magnitude or strength of correlation? i.e., whether the existing correlation is strong or
weak. Existence, direction and strength of correlation are demonstrated in the
coefficient of correlation. When a correlation is found to exist between two variables, it
means that scores within a certain range on one variable are associated with scores
within a certain range of other variables. Where the two variables (or set of data)
fluctuates or change in the same direction i.e. as one increases so does the other, or as
one decreases so does the other, a positive relationship is said to exist. Correlations
showing this pattern are prefaced with a plus (+) sign to indicate the positive nature of
the relationship. Thus, +1.0 would indicate perfect positive correlation between two
factors as with the radius and diameter of a circle and +0.8 a high positive correlation as
between academic achievement and intelligence, for example.

A negative correlation or relationship on the other hand, is found when an increase in

one variable is accompanied by a decrease in the other variable.

Thus, when high scores in one variable are associated with low scores in the other
variable. Negative correlations are prefaced with a minus (-) sign thus

-1.0 would represent a perfect negative correlation as between the number of errors
children make on a spelling test and their score on the test (i.e., the higher the score
the lower the mistakes and vice-versa). A correlation coefficient of-3.0 would also
imply a low negative correlation as between absenteeism and intelligence, for
example. If a coefficient of zero (0) is obtained, the two variables are not related.

CASE STUDY

73
It involves studying an individual, one school, one decision or one district, etc Case
study has a long history in educational research and has been used extensively in
such areas as psychology and sociology of education. For example, both Fraud and
Piaget typically used case studies to develop their theories. Case studies are
investigations of an individual group, institution or other social unit. The social unit
may be a person, a family, a social group, a solid institution of a community. In other
words, case studies mostly involve the observation of individual unit e.g. a student, a
delinquent clique, a family group, a class, a school, a community, an event, or even
an entire culture, etc.

The researcher conducting a case study attempts to analyse and understand the
variables that are important to the history, development, or care of the subject or the
subjects' problems. For example, there are some students who learn a second
language rather easily. In an attempt to gain insight into why this is the case, one
such student could be observed on a regular basis to see if there are any noticeable
patterns or regularities in the student 's behaviour. The student, as well as his/her
teachers, counsellor, parents and friends might be interviewed in-department. A
similar series of observation and interviews could also be conducted with a student
who finds learning a second language very difficult. The hope here is that, through
the study of somewhat unique individual, insights can be gained that will suggest
ways to help other language students in future.

Similarly, a detailed study might be made in a single school as a way of'

conducting/undertaking a case study in the school that is noted for its successes in the
BECE examinations.

From the above examples, it can be noted that a case study can be simple and specific,
complex and abstract. A case study should also focus on bounded subject or unit that
is either very representative or extremely typical i.e. different from all others of its
type.

Purpose of Case Studies

The purposes of case studies include the following:

I. Preliminaries to major investigations. Case studies are a useful way to

explore/examine/bring to light variables, phenomena, processes and relationships
that have not been thoroughly researched and as such deserve more intensive
investigation. The information obtained in case ’ studies can be extremely useful in
the-production of hypothesis and research questions to be tested or answered more
carefully in subsequent research.
II. Case studies could be a pilot study where methods, approaches or policies are
tried out to see what the difficulties are that need to be dealt with before a main
study is undertaken.

74
 III. Provision of anecdotal evidence. A case study may provide anecdotal evidence
(i.e. a short, interesting or amusing information about the phenomena under study;
that illustrates more general findings).

A case study is mostly used when the relevant behaviours cannot be manipulated of
respondents’ physical characteristics, social qualities or behaviours, recorded data on
respondents from newspapers, schools, courts, clinics etc.

Intensive probing and analysis of interactions between factors under study. The primary
purpose of a case study is to determine the factors, and relationships among the factors
that have resulted in the current behaviours or status of the subject of the study. The case
study thus determines why and not just what.

The case study therefore probes deeply and analyses interactions between the factors that
explain present status or that influence change or growth. This intensive probing that
character rises case studies often leads to insights concerning previously unsuspected
relationships.

In-depth case studies serve as important role of clarifying concepts and variables or
clarifying ways to measure them.

Types of Case Studies

Four types of case studies are described in the following paragraphs:

1. Historical Case Study: These studies involve the tracing of the development of
an organization or system overtime. Considered under the historical case studies is oral
history. Oral history involves first person narratives that the researcher collects using
extensive interviewing of a single individual retired educationist recounting how he/she
was taught during the era of the Accelerated Development Plan (ADP) in 1951.

2. Observational Case Study: These studies often focus on a classroom, group,

teacher and pupil often using a variety of observation and interview methods as the
major research tools.

3. Situational Analysis: Situational analysis is a form of case study where special

events are studied. For instance, events like students’ unrest/strikes, teachers/workers
strike, etc. could be studied by interviewing the students, parents, teachers, workers
headmasters and other witnesses concerned.

4. Clinical Case Study: This approach aims to understand in in-depth a

particular individual, such as a child having problem with reading, a teacher with
disciplinary difficulties, or a headteacher/manager with poor interpersonal
relationship, etc.

Planning Case Study

75
 Case studies, like other research design, need to be planned before they are carried
out. Can you think of the components that must go into these planning?

Osuala (2001) has identified four main components to the research design in case
studies. Initial case study questions: Questions with regard to who, what, where,
when and how the case study would be undertaken must be clarified and stated in the
right form. It must be noted that without the initial questions, to which the researcher
wishes to find answers, no start can be made.

Study Propositions: There is the need to state clearly and shortly suggestions or
propositions that will enable the question(s) to be answered.

Unit of Analysis: This component is concerned with defining clearly and in

unambiguous terms the exact case we want to investigate. Without this the
investigator will not be limited to any defined area/scope. He/she would, thus, be
tempted to collect everything that randomly may have a bearing on the issue.

Linking Data/Data Analysis and Interpretation: This is the means to linking data to
proposition using it as a criterion for interpreting findings.

Principles of Case Study Data Collection

What should the investigator do so that case studies would yield good and reliable
results?

Use multiple sources: Most case studies are capable of producing several sources.
The use of multiple sources is the major strength of case study approach. Multiple
sources allow for the use of different approaches/methods - observation,
t
interviews/questionnaires, etc. This helps to improve the reliability and validity of
data findings. Corroboration of all these approaches makes a case study report more
convincing.

Maintain a chain of evidence: In a case study evidence should be built up from

multiple sources if possible and shown to be in line/suitable with the conclusion.

Record data: There should be an on-site recording of every data the researcher
obtains. On-site recording can range from sketchy notes to the minute detail encoded
on a tape.

It is also important that full notes be written up as soon as possible after the
observation in view of the fallibility of human memory. The investigator should also
be more concerned with remembering the substance of conversations rather than a
perfect reproduction. The actual content of any observation should include verbal
descriptions of setting, the respondents and the activities, the observed comment,
feelings, reactions and interpretations.

Skills Needed by Case Study Investigator

76
Why should the investigator require special skills in carrying out a case study design?
Write your response in your jotter. Now let’s compare the written response to the
following.

The quality of a case study data depends to a great extent on the experience and skills of
the investigator. These skills as enumerated by Osuala (2001) include the following.

Formulation of relevant and precise questions: The investigator needs to be able to

formulate or prepare relevant and exact questions that would enable data to be extracted
from the subject.

Ability to use all the sense organs: The investigator needs to be a good listener,
observing, sensing and if possible touch and feel as a means of obtaining the right data.

A grasp/in-depth knowledge of what is being studied. The investigator must have a firm
grasp of the issues he/she is studying. Why should it be so? Without this, important
points and issues can be misled or misrepresented.

Lack of bias: Another important skill to be exhibited in undertaking a case study is to

avoid being bias. The investigator should be fair and present and interpret issues exactly
as he/she has observed. Do not interpret to support a preconceived position.

The investigator must know how to observe, allowing the subject to talk freely and at
the same time be alert for something definitive that may relate to a hazy hypothesis
which they are seeking to check. Minimise direct questions and use non-directive
probing questions. Example, "What happened next?" "What do you think?" Use words,
terms and structures used by the respondent.

Analyzing Case Study Data

How do you take notes for case study? Write your answer in your jotter. Compare your
response to the following.

Notes for case study may be derived from interviews observations documents

They may be in various forms. List the forms that you know. Now read on.

Case study notes may be in the form of written or typed notes in a pad, diary or index
card, audiotapes, computer disks, video/film. According to Osuala (2001), these notes
should be organised on an ongoing process so that as the study progresses the
investigator has some sense of the direction in which it is going and the confirmation
and contradictions that are arising. As part of the analysis, all the observation and data
should be organised chronologically or by topic, and they should be closely perused i.e.
read/look in a relaxed and not very detailed and the main aspects of the data isolated.
Through this activity the data would be organised. Other approaches of analysing the
case study data involves creating flow charts and diagrams, tabulating frequency and

77
 sorting information into chronological order i.e. ordering the events by the time at
which they happened.

EXPERIMENTAL AND CAUSAL COMPARATIVE DESIGN

Definition of Experimental Research
Experimental research is one of the most powerful research designs researchers use.
Amedahe (2002) maintains that in an experimental study, the researcher manipulates
at least one independent variable, controls other relevant variables and observes what
will happen to the subjects as a result. The researcher determines which group of
subjects gets a particular treatment. According to Compbell and Stanley (1963), this is
the kind of research in which variables are manipulated and the effects of the
manipulation upon the variables are observed.

The earliest assumptions of experimental research were based upon what was known
as the law of the single variable. In its simplest form, the law of the single variable
states that: If two situations are alike in every respect, and one element is added to one
and not the other, any difference that develops is the effect of the added element, or if
two situations are alike in every respect and one element is removed from one but not
from the other, any difference that develops is would be attributed to the subtracted
, element. Experimental research is unique in two very important respects: It is the only
! type of research that directly attempts to influence a particular variable. Also, it is the
only type that can really test hypotheses about cause and effect relationship.
1
Characteristics of Experimental Research -

McMillan (1996) identified two essential characteristics of all experimental research.

Direct manipulation of the independent variable(s) and Control of extraneous
variables.
Direct Manipulation of Independent Variables
Direct manipulation of the independent variable(s) means that the
investigator/researcher has a direct control of when the
subjects receive the independent variable and how much of it each subject receives. In
an experimental research, the investigator has control over one or more factors in the
study that may influence the 'behaviour of the subjects. That is, the researcher can
manipulate a factor, and see what will happen to the responses of subjects as a result.
The experimenter does something to at least some of the subjects in the study.
Manipulation also means varying conditions for different people and determining the
their effect.

According to McMillan (1996), the purpose of manipulating a factor is to investigate

its causal relationship with another factor. For example, some investigators may be
interested in studying the causal relationship between time on task and achievement.
These investigators would then manipulate "time" by having one group of children
spend a small amount of time on the subject and a second group a large amount of

78
time. The investigators determine (manipulate) whether children are engaged a small
or large amount of time. If, for instance, the children who are engaged a longer
amount of time show higher achievement than the other children, time on task may be
causally related to achievement.

In educational research, the method of instruction, type of grouping, amount of learning

time assignments are common independent variables manipulated in experiments.

Control of Extraneous Variables

A second characteristic is control of extraneous variables. In an experiment the

researcher seeks to keep constant all subjects, all variables, conditions, events and
procedures except the independent variable. Such factors are kept constant to eliminate
them as possible explanations for the cause-and-effect relationship. In other words, the
effect, which is measured by differences on the dependent variable, should be produced
only by variations in the independent variable.

Control refers to efforts on the part of the researcher to remove the influence of any
variable (other than the independent variable), which might affect performance on
the dependent variable. Here the experimenter introduces one or more control over the
experimental situation, including the use of control group. Why is control of extraneous
variable(s) necessary?

Control of extraneous variables is necessary to conclude that the independent variable is

causally related to the dependent variable.

Amedahe (2002) maintains that two different kinds of variables need to be controlled.

The first kind is subject variables - these are variables on which subjects in the different
groups might differ. An example is reading ability of students.

The second kind is environmental variables. Examples are established by either

eliminating a possible extraneous variable constant for all groups. In controlling
variables, the researcher wants the groups to be as similar as possible, so that the only
major difference between them is the independent variable.

Some extraneous variables can be eliminated. For example, in an experiment to

investigate which of two methods of instructions is most effective, teachers with similar
experience and teaching competence and enthusiasm would be selected. Similarly, the
academic ability of pupils should be similar.

Steps in Experimental Research

The steps of the experimental research are, in principle, no different from that employed
in other areas of research. The steps are:

Selection and definition of a problem

79
Definition and selection of population, subjects and measuring instruments

Selection of a design/drawing up the experimental design

Data collection

Analysing and interpreting of data

Drawing up the conclusion

Reporting the result

Selection and Definition of Problem

Experiments begin with the selection of the problem/topic to be researched. Any

problems that can be put under experimental conditions lend itself to experimentation.
The problem amenable (lending itself) to experimentation generally should be
converted into a hypothesis. The hypothesis can be verified or refuted by the
experimental data. The variables to be investigated should be defined in operational
terms.

Defining the Population, Selecting the Subjects/Sample and Measuring

Instruments

It is important to define the population precisely so that there can be no questions about
the population to which the conclusions are to apply. The sampling
procedure/technique for obtaining the sample for the study must then be selected
having defined the population.

Selection of a Design

Experiments are designed in a number of ways. This has resulted in several types of
experiments. Basically, experimental designs include decisions related to the selection
of the subjects and arrangements of the experimental conditions that will guide the
execution of the experiments.
The selection should place primary emphasis on the question of control,
randomization and replication. It should include a clarification of such basic aspects
of the design as the place and duration of the experiment. It is generally advisable to
conduct a pilot study, because of the complexity of an experiment, in order to ensure
the adequacy of the design.

Collection of the Data

80
Collection of the data in experimentation differs from that of the other designs. It
includes a number of steps that ultimately consists of subjecting the respondents to
certain conditions and ascertaining the effects that resulted from that procedure,

Analysing and Interpretation of Result

The researcher is concerned with the operation of the factors under investigation.
He/She must be especially sensitive to the possibility that the results of his/her study
arouse through the operation of uncontrolled extraneous factors/variables. Do you
remember extraneous variable? Turn to Unit 1 Session 5 in this book.

An extraneous variable affects the dependent variable but is unknown or not

controlled by the researcher. The methods of analysis employed in experiments are
quantitative and concentrate on experimental conditions on assessing the differences
between pre-test and post-test results and on statistical techniques. Quantitative
analysis emphasises numbers, measurements, deductive logic, control and
experiments.

Drawing up the Conclusion

The conclusion of the study must be based on the findings of the study. Care must be
taken not to over-generalise the results obtained. The results also pertain only to the
conditions under which they were obtained, and since control may have changed the
natural situation, care must be taken to restrict the conclusions to the conditions
actually present in the experiment,

Reporting the Result

The study must be reported in sufficient detail so that • the reader can make an
intelligent judgement as to its validity. Validity would be looked at later

Experimental Sampling

Characteristics of a sample

Selection of the subjects in an experimental design is usually undertaken by means

of probability sampling discussed earlier on this session, precisely session 2. The
experimental sample demonstrates the following characteristics.

It includes two sub-samples namely the experimental group and the control group.

What is the experimental group? It is the group that will be exposed to the
independent variable.

The two groups must be checked for any systematic differences.

81
 The subjects in the experimental and control groups must be the same or similar.

The selection of the subjects should be free from sampling bias.

These criteria indicate that sampling in experimental designs has to meet at least two
conditions namely,

That it produces a representative sample /free from bias and distortion

That it establishes two similar sub-samples

Types of Experimental Design

Researchers use several types of experimental designs to establish relationships

between variables. -The type one could depend upon the problem being investigated.
The differences between these designs lie mainly in the number of experimental and
control groups employed in each experiment in the use of pre-tests (i.e. in the
experimental group or in both groups), and in the way in which the independent
variable is treated.

Amedahe (2002) notes that selection of a given design is dictated by such factors as
whether there will be a control group subjects will be randomly assigned to groups
each group will be pre-tested how resulting data will be analysed

Classes of Experimental Design

Gay (1992) has identified two major classes of experimental designs. These are: single
variable designs and factorial designs.

In single variable designs one independent variable is manipulated whiles in factorial

designs two or more independent variables are manipulated.

1 Simple Variable Designs

Single variable designs are classified as pre-experimental, true experimental or quasi-

experimental design depending upon the control they provide for sources of internal and
external validity. Internal and external validity is explained under experimental validity.

Which of the three classifications of single variable designs is best and hence must be
preferred?

Pre-experimental designs do not do a very good job of controlling threats to internal

validity and should therefore be avoided. The true experimental designs represent a very
high degree of control and are always to be preferred. Quasi-experimental designs do not
control as well as true experimental designs but do a much better job than the pre-
experimental designs.

82
Factorial Designs

When more than one independent variable is included in a study, whether true experiment
or a quasi experiment, a factorial design is necessary. Factorial designs are basically
elaboration of true experimental designs and permit investigation of two or more
variables, individually and in interaction with each other.

The basic designs in each category as presented briefly by Gay (1992) are given below.

Two of the pre-experimental designs involve only one group.

(a) The One-shot Case Study: The one-shot case study involves one group which is
exposed to a treatment (x) and the post tested (o). There is no pre-test. None of the
threats to validity that are relevant is controlled by this design.
(b) A high score by subjects on the post test cannot be attributed to the pressure of the
treatment since the researcher does not even know the subjects before the treatment
was administered.
(c) The One-Group Pre-test - Post-test Design. The design involves one group which is
pre-tested (o), exposed to a treatment and post-tested.
(d) The success of the treatment is determined by comparing pre-test and post-test
scores. History and maturation as a threat to validity are not controlled.
(e) Group Comparison: The static-group comparison involves at least two groups. One
group receives a new unusual treatment while the other receives a traditional or usual
treatment. Both groups are post-tested. This design too does not control for
maturation, selection mortality and selection interactions.

Experimental Design

The true experimental designs provide completely adequate controls for all source
internal and external validity. One common characteristic of all the true experimental
designs is the random assignment of subjects to groups. To qualify as a true
experimental design, there must be random assignment of subjects. In addition, all
the true designs involve a control group. Two groups, both of which are formed by
random assignment are employed in this design.
Both groups administered a pre-test of the dependent variable and one group, the
experiment group receives a treatment X w h i l e the second group, the control
group, receives no treatment. Both groups are then post-tested. The use of a
pre-random assignment and a control group are the main difference between this
design and other designs. The combinations of these activities serve to control all
threats of internal validity except pre-test treatment reaction. Thus, by use of a
control group which has all the same experiences as the experimental group other
than the experience of the treatment itself, this design controls for history, maturation
and regression. By randomising the control and experimental groups, both selection
and maturity are controlled. The design, therefore, controls many threats to validity
or sources of bias. Hope you are bothered with the terms "history", "maturation",
"regression", "selection", "mortality" and "threats to validity", do not worry. We shall
discuss this in the latter part of this unit.

83
 The best procedure to analyse data collected using this design is to simply compare
the pre-test scores of the two groups. If the pre-test scores of the group are essentially
the same, post-test scores can be directly compared using a t-independent test. If the
pre-test scores are not essentially the same then the researcher should use analysis of
covariance (ANCOVA) which adjusts the post-test scores based on any initial
differences. Data analysis involving the t-test would be treated later in this book.

Post-test - Only Control Group Design

The design is exactly the same as the pre-test - post-test control group design except
that there is no pre-test. The post-test - only control design is potentially the most useful
true design. It can be diagrammed as follows

The post-test - only control group design is ideal in that it controls all threats to validity
and all sources of bias. The design utilise two groups, one of which experiences
treatment while the other does not, thus controlling for history and maturation. Group
assignment is made on random basis, which controls for selection and mortality. In
addition, to control for a simple testing

Quasi Experimental Designs

Quasi-experimental designs are partly, but not fully, true experimental designs. They
control some but not all of the sources of internal validity. Although they are not as
adequate as the true experimental designs, because the sources of bias are not simply
controlled, they are substantially better than the pre-experimental designs with regard to
control of the threats validity.

Quasi-experimental, like true experiments, involve the manipulation of an independent

variable, that is, the institution of an experimental treatment Quasi-experimental designs
lack at least one of the other two properties that characterises true experiments,
randomization or control group

Quasi-experiment, in sum, is a study in which subjects cannot be randomly assigned to

treatment conditions, although the researcher manipulates the independent variable and
exercises certain controls to enhance the internal validity of the results. They exist for
situations in which complete experimental control is difficult or impossible.

EXPERIMENTAL VALIDITY
To make a significant contribution to the development of knowledge, an experiment must
be valid, Any uncontrolled extraneous variables, which affect performance on the
dependent variable are threats to the validity of an experiment. According to Sarantakos
(1998), validity of experiments is considered in two forms. These are internal and
external.

What is Internal Validity of an Experimental Study?

84
Imagine the results of a study show that SSS students taught by the inquiry/discovery
method score higher on a test of science on the average, than do students taught by
lecture method. The question then is, "is this difference in scores due to the difference
in the method to the fact that the two groups have been taught differently?

The researcher who conducted the study would like to conclude like this. His/hers first
inclination may be to think same. However, this may not be a legitimate interpretation.
The following possibilities might have existed.

Students who were taught the inquiry method may be good in the sciences. .

These and other factors might explain why the inquiry group scored higher on the science
test. In such a case, the researcher may be mistaken in making any conclusions.

Many studies that describe or test relationships always have the possibility that the
relationship shown in the data is, in fact, due to or explained by somebody else.

Many alternative hypotheses may exist to explain the outcomes of a study. These
alternative explanations are referred to as threat to internal validity.

Internal validity is, therefore, the degree to which observed differences on the
dependent variable are directly related to the independent variable and not to some other
uncontrolled variable. Thus where a study has internal validity, it means that any
relationship observed between two or more variables should be meaningful in its own
right rather than being due to something else. The "something else" might be one of the
number of factors. The validity of a study is threatened by certain factors. We proceed
to discuss them briefly.

Threats to Internal Validity

True experiments possess a high degree of internal validity because the use of control
procedures (manipulation and randomization) enables the researcher to rule out most
alternative explanations for the results. With quasi-experimental, pre-experimental or
ex-post facto designs the investigator must always contend with competing explanations
for the obtained results. These competing explanations referred to as threats to internal
validity has been discussed by Campbell and Stanley (1966) and Bracht and Gales
(1968). They include:

History

In research, the term History refers to events occurring in the environment at the same
time that the experimental variable is being tested. Frequently, in educational research,
events other than the experimental treatments occur during the time between pretest and
post-test observations. Such events produce effects that can mistakenly be attributed to
differences in treatment. Events like lost of a dear one, an important event taking place at

85
the time of testing, mood of testee etc can affect the performance of participants/
respondents.

Maturation

The change in the dependent variable may not be caused by the independent variable
but rather by maturation of the subjects. What is Maturation? Read on... Maturation
refers to the processes of change that take place within the subjects of an experiment over
a period of time. These changes can be physical/biological or mental which may
occur within the subject over a period of time. The views of 'subjects may
change over time and/or subjects may by the time they come to the post-test be tired,
hungry or influence by others. Such changes can produce differences that are
independent of the experimental treatment. The problem of maturation is more acute in
studies that are extended for periods of time than in brief studies. The researcher cannot
control the occurrence of maturation but can control for its occurrence.

3. Pretesting

A pretest is a measure of the independent variable given before the treatment begins.
What is the use of pretest in experiment(s). Write your response in your jotter. Now
read on .

Many experiments use a pretest to determine the initial stage of the subject with regard
to some particular variable. The effect of t a k i n g this pretest may be to increase the
likelihood that the individual will do better in the subsequent post-test, particularly
when it is identical to the pretest. A pretest can cause differences between experimental
and control groups to disappear by providing the control group with an experience more
relevant to the post-test than in the experimental treatment. 'In fact, pretesting might
sensitise subjects and predispose them to develop an interest in t e experiment and
respond atypically to experimentation.

Instrumentation

Another threat that is related to the researcher's measurements is referred to as the

threat of instrumentation. Amedahe (2002) refers to instrumentation as "unrel iability
or inconsistency in measuring instrument which may occur in several different ways.
For example, if two different tests are used for pretesting and post testing and the
nature of the test (including scoring, administration procedure time allotted etc) is
changed in some way or another, instrumentation threat may occur. Can you think of
other ways where instrumentation threats could occur? Write your response in your
jotter? Compare to the following: Instrumentation affects can occur even if the same
instrument/test is used. For example, if people taking the pretest data, especially in
observation and interview, are more experienced than those who took the second

86
 data (post-test) and vice-versa. Similarly, if subjects or test takers become bored or
fatigued in any of the administration, then these differences could bias the results.

Mortality (Loss of Subject). No matter how carefully the subjects of a study are
selected it is common to "lose" some as the study progresses. This is known as
Mortality threat. Mortality has a threat to internal experimental validity, does not
necessarily mean less of subject through death/or subject die. Mortality rather refers
to the differential loss of subjects (attrition) from the different groups that are being
controlled. Mortality occurs when subjects systematically drop out of the study or
are lost and their absences affect the results of the study. For one reason on another
such as illness, family relocation, lost of interest, motivation etc some individuals
may drop out of a study. In any study, it is desirable to obtain post-test data from all
of the subjects who were originally included in the study. When a large number of
subjects withdraw from an experimental study, those who remain to take part in the
study may be unrepresentative of the group as a whole.

The threat of attrition is ap/likely to be especially high when the length of time
between points of data collection is long. A 12 month fellow-up of subjects, for
example, is likely to produce higher rates of attrition than a one month fellow-up,

It must, however, be noted that if attrition is random (i.e. those dropping out of the
study are highly similar to those remaining in the study with respect to
characteristics that are related to the dependent variable), and then the risk of
mortality biases is low. Generally, the higher the rate of attrition the greater the
likelihood of bias.

One may then ask, what is the standard for acceptable attrition rates?

Although there is no absolute standard for acceptable attrition rates biases are
generally of concern if the rate exceeds 20%.

Selection

The selection of people for a study may result in the individuals or groups differing
A
from one another in an unintended ways that are related to the variables to be
studied. This is what is referred to as "selection bias" or a subject characteristics threat.

To explain this further, in most experiments two or more groups of subjects are
compared. One group receives one level of the independent variable, and the other
groups receive other levels of independent variable, In studies that compare groups,
subjects are randomly assigned to levels of the independent variables. This procedure
ensures that the different groups of subjects are comparable on such characteristics/
variables as age, gender, ability, socio-economic background, motivation attitudes,
interest and the like. However, in some experiments the random assignment of subjects
to groups may not be done properly. In these circumstances it is possible that the groups

87
to be compared may differ so much on such variables mentioned above. If not controlled,
these variables may "explain away" whatever differences between groups are found.

A regression threat may be present whenever change is studied in a group that is extremely
low or high in its pre-intervention performance. In other words statistical regression
refers to the 'tendency of subjects who score extremely high or low in -a pretest to score
closer to the mean of both groups on the post-test, regardless of the effect of `the
treatments. That is, very low pretest scores are likely to be higher on the post-test and
very high pretest scores are likel y to be lower on the post-test. Studies in special
education are particularly vulnerable to this threat, since the students in such studies are
frequently selected on the basis of previous low performance.

The issue of regression can be explained statistically, but for our purposes it simply
describes the fact that a group selected because of unusually low (or high) performance
will, on the average, score closer to the mean on subsequent testing, regardless of what
transpires in the meantime. In other words, a class of students of markedly low ability
may be expected to score higher on post-tests regardless of the effect of any
intervention to which they are exposed. Like maturation, the use of an equivalent
control or comparison group handles this threat.

Some examples of a possible regression threat are as follows:

Those students who score in the lowest 20 percent in a maths test are given special help.
Two weeks later, their average score on a test involving similar problems has improved.

s The University Athletic coach selects the members of his team from those who have
the fastest times during the final trials for various events. He finds that their average
time decreases the next time they run, however, which he attributes to differences in
track conditions.

PREPARATION AND ORGANIZATION OF DATA

In research we have to collect data and the data collected must be analysed to enable
us solve a problem or explain and describe a phenomenon under consideration. As
researchers, whatever decision we reach after analysing data should have conclusion
validity. Conclusion validity is the degree to which conclusions you reach about
relationships and patterns in your data are reasonable (Trochim, 2001). Conclusion
validity is important because in a study such as the relationship between two
variables, a researcher can conclude that there is no relationship when in fact there is
or he/she can conclude that there is a relationship when in fact there is not. The
problem of a researcher concluding that there is no relationship when in fact there is is
of concern to researchers. In order to avoid error in our conclusions, that is, to
achieve conclusion validity, there is the need to take appropriate steps to analyze
and interpret our data and the results.

88
Preparation and Organization of Data

From the earlier sections of this text, it is clear that you can use several tools to
collect data for your project work or research. You may use a questionnaire, an
interview guide or observation guide or schedule. The data collected in a study must
be prepared before it is analysed. Data preparation involves editing, scoring, coding
and entering the data into the computer, if a computer will be used to analyze the
data. Even if the data will be analysed manually, it must be prepared before it is
analysed. We now proceed to discuss the process or activities involved in data
preparation and organization.

Editing: Editing of research data involves (a) correcting mistakes in the data if
possible and (b) deciding whether some of the data collected should be deleted from
or kept for data analysis later.

Editing of data is intended to detect and as far as possible eliminate errors in the data
to ensure conclusion validity.

For instance, data collected using questionnaires may be incomplete. They may
contain inaccurate information which you the researcher must examine before
proceeding to analyze it. It is at the editing stage that you do this examination.
Similarly, the best interviewers are liable to make errors, omit to ask certain questions
or to record answers to some questions. Thus, interview responses must be edited
before proceeding with the analysis of the data. Such editing can be done by the
interviewers immediately after they have completed each interview. In the same vein,
before the completed questionnaires can be regarded as ready for coding, tabulation
and analysis, they should be checked for completeness, accuracy and uniformity.

For completeness, you must check and ensure that there is an answer to every
question. Sometimes, you have to fill the gap, if any, by deducing from other data on
the questionnaire. With regard to accuracy, you must try to check whether the
answers are accurate. You must look for inconsistencies.

Inaccuracy may be due to carelessness or to a conscious attempt to give misleading

answers by respondents or even the interviewer. Remember that some of the ways of
data collection are self-report procedures. .

You also have to edit your data in terms of uniformity. This process helps you as a
researcher to find out whether in interviewing, for example, questions and instructions
have been followed uniformly, if different people administered the instruments.

Scoring: The next activity in data preparation is scoring, When a study is quantitative
in nature, all instruments administered should be scored accurately and consistently.
For self-developed instruments such as questionnaires, you must develop a scoring
procedure to guide you. Steps for scoring each item and for arriving at a total score

89
must be delineated and carefully followed. For instance, if a study uses an attitudinal
scale, the negative and positive statements should be scored properly. For example,
when "strongly agree" to the statement "I enjoy school work" is scored 5 on a five
point scale the statement "I find school work boring" should be scored 1, if an
individual "strongly agree" with it. Scoring is usually not applicable to qualitative
data.

Coding: Coding is one other activity at the data preparation stage. Coding does not
mean the same thing in qualitative and quantitative research.

In qualitative research, coding refers to the process of categorizing data. That is,
coding in qualitative research means the process by which the researcher makes sense
out of text data, divides it into text or image segments, labels the segments, examines
codes for overlap and collapses these codes into themes. The codes are just labels to
describe a segment of text. Examples are "poor learners", "problem children",
"disruptions during the class", "students talking to each other',', Teach (1990) and
Creswell (1994) recommend the following steps in coding qualitative data.

Read through all of the transcriptions carefully and jot down in the margins some
ideas based on the text

Pick one document (e.g., one interview) and go through it, asking the question "what
it is about?" Write down in the margin in two or three words and draw a box around it.

Begin the process of coding the document by identifying text segments, placing a
bracket around them and assigning a code word or phrase that accurately describes the
meaning of text segment.

After coding an entire text, make a list of code words. Cluster together similar codes
and look for redundant codes. The objective is to reduce a long list of codes to a
smaller, more manageable number.

Take this list and go back to the data and circle specific quotes from participants that
support the codes

Reduce the list of codes to get 5 to 7 themes or descriptions of the setting or

participants. Themes are similar codes aggregated together to form a major idea in the
database.

From the coding and the themes, construct a narrative description and possibly a
visual display of the findings for your research report.

STATISTICAL DATA ANALYSIS: HYPOTHESIS TESTING

Choosing a Statistical Test

90
In the previous session the use of descriptive statistics as part of the analysis of data
has been presented. Although some research questions are answered in the
preliminary analysis, more complex questions often need to be addressed by making
inferences from a sample to a population. This is called inferential statistics.
Inferential statistics enable a researcher to draw conclusions, inferences or
generalizations from a sample to a population of participants. This involves
hypothesis testing. In doing this, you have to choose the appropriate statistical test to
use.

For many quantitative researchers, the use of descriptive statistics is only the
beginning of their data analysis. Generally, as researchers we are interested in
comparing responses of groups or relating two or more variables from a sample
drawn from a population. Since the scores or observations from studying a sample
are only an estimate of a larger population, researchers conduct hypothesis testing.

Hypotheses are explained as educated guesses or conjectures. Such educated guesses

must be tested. Hypothesis testing is a procedure for making decisions about results
by comparing an observed value with a population value to determine if no
difference or relationship exists between the values. Because the intent of much
quantitative research is to draw inferences or conclusions about a population from a
sample, inferential statistics are used. To understand inferential statistics, you need
to learn about the decisions that go into the selection of an appropriate statistical
test.

Generally, inferential statistics are divided into two types: parametric and non-
parametric.

Parametric statistics are statistical tests based on the premise that the population
from which samples are obtained follows a normal distribution and the parameters of
interest to the researcher are the population mean and standard deviation. These
statistics apply to hypothesis testing which we shall discuss in the next section.

Parametric statistics have certain assumptions about the observations/scores. These

assumptions are:

The variables are measured in interval scales

Scores from any two individuals in a study are independent of each other
The variables that distinguish each population are similarly distributed
The variables that distinguish each population are similarly distributed among each
population, in the case of two or more groups
Examples of parametric statistics include the t- test and Pearson Correlation
Coefficient. There are other advanced statistical parametric procedures but we will
limit you to the two.

91
The T-independent test is used when a study involves two groups with the
independent variable being categorical and the dependent variable being continuous.

The Pearson correlation coefficient is denoted by (r xy) and is applied when two
continuous variables are related to determine their degree of relationship. We hope the
term correlation is not new to you.

Non-parametric statistics, on the other hand, are statistical tests that only make the
assumption of independent observations of scores for each individual in the study. In
these tests, the data are typically measured in categorical scores on either the
independent or dependent variable. The most frequently used non- parametric test in
educational research is chi-square (X 2). It is used to test a hypothesis concerned with
category within group comparison. In the next section of this session, we will explain
the procedure of testing a hypothesis using chi-square.

Below is a few criteria to guide you in choosing an inferential statistical test.

'Determine the type of quantitative research hypothesis you would like to analyze. If
the hypothesis is to compare group means of independent groups, then the t-test may
be appropriate. If the hypothesis is concerning how two variables are related and they
are continuous variables then Pearson correlation coefficient is appropriate.

Consider the scale of measurement for your independent and dependent variable(s) in
the research hypothesis. If the scale of measurement for your independent variable is,
for example, categorical and the dependent variable is continuous and the hypothesis
demands determining differences between two groups using their means, then the t-test
will be appropriate, For categorical independent and dependent variables, chi-square
test may be appropriate.

Determine if your scores (observations) on the variables in your hypotheses are

normally distributed or whether they are skewed. If the scores are normally distributed
and the other assumptions of parametric statistics are met, then apply the appropriate
parametric statistics. Otherwise, you have to choose a non-parametric statistics, (e.g.,
X2).

It should be clear to you from the foregoing discussion that the first major step you have
to take in analyzing your data using inferential statistics is to decide the appropriate
statistics to apply, Once you have selected the appropriate statistics, you go ahead to
conduct the hypothesis testing,

Conducting Hypothesis Testing

We explained in the previous session that inferential statistic allow you to make
inferences from a sample to a population. Since we can rarely study an entire
population because of size, expense and ease of identification, we need to study
samples and draw inferences to the population. Sample data that you collect can only

92
estimate the population parameter such as the mean. Because a sample is selected
from a population, any given sample may be in error. That is the characteristics of the
sample may not reflect the population characteristics/parameters, assuming that the
sample is randomly chosen. Hypothesis testing allows you to test whether the
sample data differ from the population values.

For example, in an experiment to establish the effectiveness of two methods of

teaching a topic in mathematics, you may be interested in whether there is a
difference between sample means of the control and experimental groups.

Suppose you carry out this study using 20 subjects for the control and experimental
group and obtain a mean score of 65 for the control group and a mean score of 70 for
the experimental group. The question that arises is: Is there a statistical difference in
performance between these two samples. In other words, do the two populations differ
in terms of their performance? You can only answer this question through
hypothesis testing. We give an example later.

In a correlational study, you may be interested in whether there is a relationship

between the scores of two variables for individuals. Again, hypothesis testing allows
you to test the relationship. The rest of this session will inform you about six general
steps you will take to conduct hypothesis testing. Important terms are defined

Generally, there are six steps in hypothesis testing (Cresswell, 2002):

Establish a null and alternate hypothesis

Set the level of significance or alpha level for rejecting the null hypothesis

Collect data

Compute the sample statistic, (usually using a computer programme)

Make a decision about rejecting or failing to reject the null hypothesis

Determine the degree of differences if a statistically significant difference or

relationship is found.

Establishing a Null and Alternative Hypothesis

You may recall that the null hypothesis (H0) is a statement that no difference exists
between the averages or means of two groups or no relationship exists between two
or more variables. Examples are:

a. No difference exists between the experimental and the control programmes

b. No difference exists between privately and publicly managed schools'

achievement in mathematics in the BECE;

93
c No difference exists between the sample of teachers chosen to be interviewed and
those who were not chosen in terms of experience in teaching

d. No relationship exists between Junior Secondary School students’ achievement in

Math and English.

The alternative hypothesis, on the other hand, indicates a difference or relationship.

The direction of the difference may be positive or negative (directional hypotheses) or
either positive or negative (non-directional hypotheses). In the same win, the
relationship between two variables will be either positive or negative (directional
hypotheses) or either positive or negative (non-directional hypotheses). The
alternative hypotheses corresponding to the above null hypotheses are:

a. The experimental group's mean score is statistically higher than the control group's
means score.

b A difference exists between privately and publicly managed schools' achievement in

mathematics in the BECE.

c A difference exists between the sample of teachers chosen to be interviewed and

those who were not chosen in terms of their experience in teaching.

d A negative relationship exists between junior secondary school students

achievement in Math and English.

From the examples of hypotheses presented above, the researcher will test whether
differences observed are real or due to chance or error in the case of hypothesis (a),
(b) and (c). With regard to hypothesis (d), the researcher will test if there is real
relationship or the relationship is due to chance or error. This calls for establishing a
decision on the standard to use. This is concerned with setting the level of
significance. We explain and discuss the level of significance in the following section

Setting the Level of Significance (Alpha Level)

The level of significance or alpha level is an important concept in hypothesis testing.

We said at the end of the previous section that we need to set a standard to determine
whether there is a difference between the means of two samples or whether there is a
relationship between two variables. It is the standard that we are calling significance
level. A level of significance is a boundary in a sampling distribution for making this
decision. The three statistics we referred to earlier, namely t-test, chi-square and
Pearson correlation coefficient have their sampling distributions. The significance
level helps you to establish the boundary for any one of them that you may be using.

Specifically, a significance level (alpha level) is a probability level that reflects the
maximum risk you are willing to take that any observed differences are due to chance

94
(Cresswell, 2002). It is typically set at .05 (5 out of 100 times it will be due to chance)
for social science research. In some disciplines it is set at .01 (1 out of 100 times the
sample statistics will be due to chance). What a significance level of M5 means is that
five out of 100 times an extremely low probability value will actually be observed if
the null hypothesis is true.

In hypothesis testing, the area of the sampling distribution of the specific statistics you
are using is divided into what are called critical region and acceptance region using
the significance level. The area on the distribution (t-distribution or normal curve) for
low probability values if the null hypothesis is true is called the critical region. -If
sample data (e.g., the difference between the performance of private and public
schools in math at the BECE) falls into the critical region the null hypothesis is
rejected. This means that instead of "there is no difference as stated in hypothesis (b) in
the previous section; we find the alternative to exist. That is, "there is a difference".

Sketch of t-distribution showing the rejection and acceptance regions

Note that the sketch showing the acceptance and rejection regions indicate the rejection
regions at both ends of the curve with the alpha level of .025 each, making a total
of .05. When the critical region for rejection of the null hypothesis is divided into two
areas at the tails of the sampling distribution, as shown above, we have a two-tailed test
of significance. On the other hand, if the rejection region is placed at one-end for
rejection of the null hypothesis, we have a one-tailed test of significance. In this case
the alpha level is not divided. The specific level is used at the specific tail of the
distribution. Directional hypotheses are tested using one-tailed tests of significance
while the non-directional hypotheses are tested using two-tailed tests of significance.

It is important for you to note that a one-tailed test of significance is more likely to
reject the null hypothesis if it is false than a two-tailed test. Thus, a one-tailed test is
said to have more power than a two-tailed test.

We want to continue this section by outlining the four possible outcomes that could occur
during hypothesis testing.

The researcher can reject the null hypothesis when it is actually true. In this case, we
say that the researcher commits a Type I error. The probability of this error rate is
alpha (α)

The researcher can fail to reject the null hypothesis when it is actually false. Here, we
say, the researcher commits a Type II error. This is known as beta (β).

The researcher can reject the null hypothesis when it should be rejected because a
difference exists or a relationship exists. This is a correct decision.

The researcher can fail to reject the null hypothesis when it should not be rejected
because there was no difference or relationship. This is also a correct decision.

95
When you test your hypotheses you must endeavour to make a correct decision. Bear in
mind that in education and social sciences, alpha level of .05 is typically used to test
hypotheses.

Having set the alpha level, you must as well determine the critical value the test statistic
must attain to be significant. Each test statistic, such as t and chi-square has a
distribution. This distribution is called the sampling distribution. Its mean is called the
expected value and its variability is called the standard error.

All test statistic distributions are divided into an area of rejection any an area of
acceptance which we explained in the sketch above. The boundary between an area of
acceptance and rejection has a value based on the alpha level. This is what we are
calling the critical value.

Critical values can be found in statistical tables and statistics textbooks.

For example, for the t-distribution with an alpha of .05 and a two-tailed test, tabular
values will show that at three degrees of freedom, the critical value is 3.182. The term
degrees of freedom sounds intuitively meaningful but it is, in fact, a complex
statistical concept that is discussed in advanced texts and is well beyond the scope of
this module. With the critical value of 3.182 and calculated value higher than 3.181 or
less than - 3.182 will fall in the rejection region.

Before you collect data, you have to establish a decision point. Do not arbitrarily set
this point after you collect data. You may be fishing. This is not allowed.

Having set your alpha level and determined the critical value, then you perform the
calculation. The calculation can be done manually or by computer. Calculating your
tests using computer will give you accurate values. You could get somebody who
knows about statistical testing to do the calculation for you for a fee. The manual
calculation may tend to be tedious. You may not have the competencies in calculating
some of the statistics.

We present below the formulae for computing the t-independent test, chi-square and
Pearson correlation coefficient.

Computing the t - independent test

The t - independent test is used to test hypotheses about the difference of two means
and thus requires numerical data. The shape of a t distribution approaches the bell
shape of a standard normal distribution.

The formula for the two-sample t- test is a ratio and it is as follows:

96
Where X1 and X2 = mean of group one and two respectively

S; and S: = variance of group one and two respectively

n1 and n 2 = number of cases in group one and groups two respectively

The degrees of freedom that go with the distribution, is n1-1 plus n 2 - 1 . In other words,
the degrees of freedom for the test is n 1 + n 2 - 2.

Once you compute the t statistic you compare it with the determined critical value for
a decision-making. Note that there are a number of tests. These include t test for one
mean, two means and matched pairs. You have to choose the appropriate t test
depending on your sample and other characteristics of the sample.

97
RESEARCH REPORT: GENERAL ISSUES AND THE
PRELIMINARY
Getting Started

Planning Writing the Project Report

In the previous section, we explained the process of analyzing data, particularly,

quantitative data. When all the hard work of gathering and analyzing data is
complete you will need to write the final report. A study is not finished until it is
written up. Time has to be allowed for writing. The report is the document that tells
others, apart from the researcher, all that was done and its outcome. It is therefore,
important to be able to communicate the report in the most effective way possible.

Report writing is not or should not be a frantic activity carried out at the end of the
project. It is a process of varied stages all of which need to be recorded at the time
they are completed. Your first drafts will almost certainly need to be revised and in
some cases completely rewritten. The foundation for the report should be established
at the proposal stage of the study. To be able to write the report well and to be on
schedule, you need to plan the report writing by doing the following:

Set deadlines. It is useful to set deadlines and completion dates for different
sections and for the whole report and keeps the dates in mind. The project report,
generally, has five chapters. You must set yourself deadlines for completing each
chapter.

Write regularly. For most people writing is a difficult exercise. Because of this
most people procrastinate in writing. It is advisable to keep regular hours for writing
and to work in the same place. This advice is very pertinent to you as a distance
learner because you have been working and studying at the same time.

Write up a section of the report as soon as it is ready. That is,

some sections of the research will be ready for writing before others. For example, the
introduction chapter and the literature review chapters will be ready before the other
chapters. This is to advise that, you should write your report as you go along, when
possible.

GENERAL WRITING SKILLS IN RESEARCH REPORT WRITING

An important aspect of writing a research (project) report is being able to

communicate your ideas well to the reader. Indeed, the tools of communication that a
writer employs are words. They should be so selected and ordered as to inform the
reader about the research rather than to impress him/her with flowery language and
other forms of verbal gymnastics. The rest of this section explains some of the general
writing skills you need to bear in mind as you embark on writing a project report.

98
The Use of Appropriate Terms and Concepts: Writing a report is different from
writing a non-technical composition. You must, therefore, use appropriate terms and
concepts in research in your writing. However Jargons which are not understood by
some people should be avoided when writing a research report.

Sentence Structure: In a research report, long sentences should be avoided. A sentence

is long if it is more than three lines or uses more than two dependent clauses. Usually,
when sentences are long writers tend to lose the trend and grammatical errors occur in
the sentences. It is, therefore, advisable to write short and clear sentences to express
your ideas.

(c). A, B, C's of writing: You must apply the A, B, C's of writing when writing your
research report. These are accuracy, merit and clarity that go with critical adherence to
the laws of grammar. Accuracy involves precision of expression

Any account must be given in simple and straightforward manner so as to forestall

problems of misinterpretation by readers. It also has to do with the reliability in
reporting the finding(s) of the data. Broad generalisations and dogmatic expressions
should be avoided. It is better to use qualifications such as tends, more likely, appears,
seems and probably than to use the unmodified words.

Emphasis should be on the data rather than the investigator. In writing a research
report, your emphasis should be on the work rather than on you the investigator. In
general, references to the investigator are to be kept to a minimum. Consequently,
personal pronouns of the first person (I, me, we, us, our) and of the second person
(you, your) are not to be used.

Generally, a research report is written in the past tense. Thus, in writing your report
you have to say, for example, the purpose of the study was ... This, generally, follows
from the fact that in research or project report, the investigator reports what he or she
had done. However, there are some exceptions. Conclusions that apply to more than a
single instance are Written in the present tense, The present tense is also used when
referring to general truths and well established principles as well as tables presently
before the reader„

Avoid one-sentence paragraphs in the report. They do not convey complete ideas.

Be logical in your presentation. The logic of the argument should flow from one idea to
the next in sentences, paragraphs and chapters, one point should be fully treated before
going to the next.

Spellings, punctuations and numbers must be correct. Simplified spellings are not
acceptable in research writing. The use of the abbreviation, e.g.. for example, is only
appropriate in parentheses ( ). This equally applies to i.e., Avoid , the use of etc. It
creates doubt in the mind of some readers. Abbreviations are also not to be used in the

99
 report, except they are acceptable, generally, in writing. The expression "percent" is
spelt out in a sentence., Numbers of less than 10 that do not represent precise
measurement, are spelt out. Round numbers and numbers that begin sentences, titles
and headings are also spelt out.

That is, you should not begin a sentence, title or heading with a number. Fractions
should also be spelt out except when they are part of a larger number. Decimals should
be rounded to the nearest percent. If small and large numbers appear in a series, figures
must be used for them.

Note that a project report of dissertation/thesis is type written and double-spaced

including quotations. Quotations that are more than 40 words are inducted rather than
put in quotation marks. The pages of all quoted materials should be indicated against
the materials in parentheses (e.g., p. 18)

The Structure of a Project Report

So far we have been discussing ways by which you can write a good report and be on time
too. In this section, we present to you the general structure of a project report as
approved by the Faculty of Education, University of Cape Coast. You should note that
universities or academic institutions tend to adopt styles and formats in writing research
reports, depending upon the type of research and subject area. Thus, specific rules and
formats of reports vary from institution to institution. Most of the work depends on the
individual style of the researcher and those required by the institution. The structure
presented here are suggestions to guide you towards effective report writing but with
emphasis on the style and format approved by the Faculty of Education, University of
Cape Coast. We earlier on advised that you should obtain the Faculty's brochure for
writing projects and thesis.

This is important in the sense that when you read through the brochure, you become more
familiar with how you should do certain things with regard to the report writing.

The entire research report has three major parts. These are:
The preliminary section

The main body of the report

The reference materials

Each of these major parts contains sub-sections. We present the outline of the three
4
major parts of the research report here. In subsequent sections of this unit we explain
and describe the sub-sections.

OUTLINE OF THE STRUCTURE OF A REPORT

100
 Preliminary Section of a Project Report The essential elements of the
preliminary part of a project report are:

The Main Body of the Report

The main body of the report follows immediately after the preliminary materials.
Generally, it consists of five chapters. These are:

The Reference Materials The reference materials are in two parts, namely:

The References

The Appendixes

The Preliminary Section of the Research Report

In the foregoing section we outlined the contents of the preliminary section of the
report. In this section, we describe in detail the preliminary materials of the report and
how to present them.

The Title Page

101