Introduction To Researcch Methods (AutoRecovered)
Introduction To Researcch Methods (AutoRecovered)
The reader will also have the skills of assessing the quality of information to view other
people’s research with a critical eye. The Module introduces principles, methods, and elements
of the research process within the context of quantitative, qualitative, and mixed design
approaches. Learners will use these theoretical underpinnings to critically understand how to
define a research problem, how to develop research questions, how to do research, how to write a
thesis report or manuscript, and how to critically appraise research outputs so as to use them for
evidence based decision making.
The module sessions included in the module are introduction to research and literature
review, Quantitative research methods, Research ethics, Introduction to qualitative research,
Qualitative research methods, Research proposal and thesis report writing, and manuscript
writing. Purpose of the Module This module is prepared to standardize the teaching of research
methods course mainly for MPH/MSc and other postgraduate health sciences students. It can
also be used as a resource for anyone who is interested in research.
This session begins by describing about the term ‘research’. Different scholars define
Research differently. In 2002 Oxford English Dictionary define research as “the systematic study
of materials and sources in order to establish facts and reach new conclusions.” According to
McMillan and Schumacher (1997) define research as “a systematic process of collecting and
analyzing information (data) for some purpose.”
Discussions then move on to critical literature review steps and recognise its importance
in research work. Having provided this information, the session then focuses on the theoretical
considerations involved in the first stage of the research process: formulating the research
problem and research questions, hypotheses or objectives.
We assess the information what has been presented, create an opinion as to the relevance
and validity of the research ideas, and reach to our own conclusions based on the following
considerations:
why the research was conducted
what we understand is the question to answered
how the findings are presented
what we know about who conducted the research
what other people think about the information
how the research findings relate to us
We may choose to overlook the information, if we don’t understand what the findings
mean, we may disagree, or agree with the findings and either to adjust our contribution to curb
the problem or decide that we would have other computing priority agendas. In other words, you
may not be aware of it, but you already have some understanding of the research process, the
role of research, research concepts, and research evaluation.
Characteristics of Research
Activities to Qualify as Research Criteria for good research Descriptions Controlled
Exploring causality in relation to two variables, set the study to minimizes the effects of other
factors affecting the relationship Rigorous you must be meticulous in ensuring procedures
followed to find answers to questions are relevant, appropriate and justified. Systematic
procedure adopted to undertake an investigation follow a certain logical sequence different
steps cannot be taken in a haphazard way Some procedures must follow others Valid and
verifiable Implies that whatever you conclude on the basis of your findings is correct and can
be verified by you and others. Empirical Any conclusion drawn are based upon hard evidence
gathered from information collected from real life experiences or observations. Critical critical
scrutiny of the procedures used and the methods employed is crucial to a research enquiry. The
process of investigation must be foolproof and free from drawbacks. The process adopted and
the procedures used must be able to withstand critical scrutiny. Remarks: For a process to be
called research, it is imperative that it has the above characteristics. 2. Research Types Based on
purposes Research differ accordingly. There are three major research types such as application,
objectives, inquiry mode, and sometimes may be use as mixed method research. 2.1 Application
(Pure and Applied Research) Applied research is "hands-on", which means that researcher is
actually working with the topic while conducting the research. Generally, applied research
focuses on 10 "practical problems" such as public health issues/ clinical issues in order to come
up with solutions to better or improve an existing condition. Basic research is often called "pure"
research and is considered the foundation for applied research. It is often considered researching
to increase knowledge as opposed to applied research truly is intended to solve a problem. 2.2
Objectives (Descriptive, Co-relational, Exploratory and Explanatory) Descriptive research
(statistical research): The idea behind this type of research is to study frequencies, averages, and
other statistical calculations. The main goal of this type of research is to describe the data and
characteristics about what is being studied. Correlation research : measures the relationship
between two or more variables or gives an indication of how one variable may predict another.
Exploratory research is a type of research conducted because a problem has not been clearly
defined. Exploratory research helps determine the best research design, data collection methods,
and selection of subjects. Given that it is fundamental in nature, exploratory research often
concludes that a perceived problem does not actually exist. Explanatory research explores
"why," and attempts to explain as the purposes of explanatory research. It builds on exploratory
and descriptive research and further identifies the reasons for something that occurs. It looks for
causes and reasons. 2.3 Inquiry mode (qualitative and quantitative): From the process adopted to
find answer to research questions – the two approaches are: Structured approach
Unstructured approach Unstructured approach (qualitative): The unstructured approach to
inquiry is usually classified as qualitative research. This approach allows flexibility in all aspects
of the research process. It is more appropriate to explore the nature of a problem, issue or
phenomenon without quantifying it. Main objective is to describe the variation in a phenomenon,
situation or attitude. E.g., description of an observed situation, the historical enumeration of
events, an account of different opinions different people have about an issue, description of
working condition in a particular industry. Structured approach (quantitative): quantitative
research is usually classified as structured approach to inquiry. Here everything that forms the
research process includes objectives, design, sample, and the questions that you plan to ask of
respondents is predetermined. It is more appropriate to determine the extent of a 11 problem,
issue or phenomenon by quantifying the variation. For example it is deals with measurement of
attitudes, behaviors and perceptions of study subjects. In short quantitative research address the
following issues: Means the data is analyzed in terms of numbers. Involves the collection of
numerical data. Predicts and explains data in the form of statistical analysis. Uses the
numerical method to analysis and interpret the results. Finds out the relationship among
quantifiable variables and the results are inferred. learning activity 1.1 The Instructor can ask the
students to think and plan for doing research as their graduation requirement on different public
health research topics such as factors affecting increasing early childhood mortality, factors
increasing maternal mortality. Suppose that the instructor asks students to plan for doing research
on factors affecting increasing mortality on early childhood in certain locality, the instructor
would give directions and set the limits for this planning. For example: you have one million
Ethiopian birr to do this research and you can do this research in one year period; and you have
to spend all the money but cannot over spend. You research design could be prospective or
retrospective follow up . Now ask the students to think of their personal and research context,
research outcomes and policy implications and limitations etc. and ask them to rationally plan
the process. Give them ten minutes to plan. Comment 1.1 Now ask them how they planned it.
As some of the students share how they planned it, through this discussion establish that the
planning involves a complete research process. One has to see the problem and objectives. One
has to do literature review by gathering information from different literatures and discuss with
their colleagues those who have the expertises. The process also involves research methodology.
For instance, some students will directly go to the literature review and understand what is
already know about their research questions and what is unknown in the subject matter. Ask
them in the same way how they would analyze gathered information collected from the literature
review they have done and share that it is their data analysis that leads to their conclusions and
decision making. 12 Learning Activity 1.2: Identify the Research Types from the Following
After the introductory lecture on types of research, the instructor will provide the participants
with a worksheet which contains the following statements: 1. If you are going to study the
relationship between job satisfaction and career development of health professionals working at
the university of Gondar, this is an example of _____________________________. 2. If you are
going to find out reasons of staff turnover in the Hospital, this is an example of
_______________. The instructor ask the students to fill in the blanks with appropriate
statements about the types of research. Comment 1.2 1. Correlation research 2. Explanatory
research 3. Research as a process Research process as impression of linearity: Research can be
seen as a series of linked activities moving from a beginning to an end. Research usually begins
with the identification of a problem followed by formulation of research questions or objectives.
Proceeding from this the researcher determines how best to answer these questions and so
decides what information to collect, how it will be collected, and how it will be analyzed in order
to answer the research question. At each operational step in the research process you are required
to choose from a multiplicity of methods, procedures and models of research methodology which
will help you to best achieve your objectives. 3.1 Steps in linearity Research Process:
Formulating the Research Problem Extensive Literature Review Developing the objectives
Preparing the Research Design including Sample Design Collecting the Data Analysis of
Data Generalization and Interpretation Cyclical or iterative Research Process: On the other
hand, research process is often an iterative process whereby the process of conducting the
research will give rise to new ideas which, in turn, feed back into the data collection and analysis
stage. Decisions made early in the research process are often revisited in the light of new insights
or practical problems encountered along the way. 13 Figure 2 cyclical or iterative Research
Process Sources: 1. Through the process of designing your research, consideration of both
practical and conceptual issues may force to reconsider original research question. 2. Difficulties
with access to research sites or participants may cause to reconsider questions or methods. 3.
Issues arising during data collection may suggest that additional data are required or reveal
problems with the original research question. 4. Problems or new questions arising from analysis
of data collected so far may result in a need to collect more data, sample elsewhere or employ a
different technique. 5. Finally you progress to the report writing stage. Regardless of the route
taken subsequently, research should start with the problem and the research questions. If the
intention of research is to answer your questions, it follows that choice of method should develop
from question: choose the method that can best provide the information you need to answer your
research question given the resources available to you. This is one reason why it is very
important to be clear as to what you are asking. As you can see there are numerous choices to be
made within the research process. Planning your research involves the consideration of four
overlapping themes. The conceptual approach – the philosophical underpinnings of research 14
Research design – how data collection is organized Data collection techniques – how data are
collected Sampling – from whom data are collected These aspects of research planning can be
represented as forming different layers of a research ‘onion’. Figure 3: The ‘onion’ diagram of
research choices Each layer of the onion presents a different set of choices regarding research
philosophy, research approach, method and so on. All research involves choices at all these
levels, though these choices are not always made explicit. Choices further into the centre of the
onion are often, but not always, contingent on those made further out. Other Considerations
Choices in research planning must also take into account the following: The types of
information outputs required – who needs the information and for what purposes Research
resources – time, funds, facilities, staff, and access Ethical considerations – for example, within
the research plan is it possible to obtain informed consent from all participants, does the plan 15
involve any risks to the safety of the researcher, can the researchers assure the confidentiality of
all information given? 4. From Research Idea to Researchable Problem 4.1 Sources of research
ideas Ideas for research problems or topics can arise from a range of different potential sources
such as personal or professional experience, a theory, the media, or other research studies. 1.
Personal professional experience Our daily personal or professional experience lead to identify a
problem that we would like to have a solution. Alternatively, we may encounter a question or
questions that we would like to try and answer. For example, on a personal level, you may prefer
the taste of organically produced vegetables and thus you may speculate if people general in
Ethiopia prefer the taste of organically produced vegetables to those produced non-organically.
The research topic is a study into taste preferences and the question ‘do people in general in
Ethiopia prefer the taste of organically produced vegetables to those produced nonorganically?’
2. Theory Theories are ideas about how things relate to each other. Theories may be general,
commonly held beliefs (such as, man is mortal, therefore Abebe is Mortal ) or more technical
ideas (for example, that global warming is causing a change to the timing of the seasonal
responses of the change of cultivation in Ethiopia). There are many ways of expressing theories,
some are very formal, others are informal. Here are some examples: The idea that distance
learners have different needs than on-campus students is a theory. Theories may be useful in
suggesting interesting questions and generally guiding fieldwork, but should not restrict us from
exploring alternative explanations. The end result of the research process is knowledge. 3.
Literature and Media There are many sources of literature, such as journal articles, books, and
newspapers. When searching and reading literature it is possible to encounter gaps in information
and knowledge, and problems for which there is currently no solution. These may provide a good
basis for research. We are also flooded with information presented by the media, such as
television, which again might give rise to research ideas. 16 Considerations in Selecting a
Research Problem These help to ensure that your study will remain manageable and that you will
remain motivated. 1. Interest: a research endeavour is usually time consuming, and involves hard
work and possibly unforeseen problems. One should select topic of great interest to sustain the
required motivation. 2. Magnitude: It is extremely important to select a topic that you can
manage within the time and resources at your disposal. Narrow the topic down to something
manageable, specific and clear. 3. Measurement of concepts: Make sure that you are clear about
the indicators and measurement of concepts (if used) in your study. 4. Level of expertise: Make
sure that you have adequate level of expertise for the task you are proposing since you need to do
the work yourself. 5. Relevance: Ensure that your study adds to the existing body of knowledge,
bridges current gaps and is useful in policy formulation. This will help you to sustain interest in
the study. 6. Availability of data: Before finalizing the topic, make sure that data are available. 7.
Ethical issues: How ethical issues can affect the study population and how ethical problems can
be overcome should be thoroughly examined at the problem formulating stage. 4.2 Formulating
the research problem Once the general topic or problem has been identified, this should then be
stated as a clear research problem, that is, taken from just a statement about a problematic
situation to a clearly defined researchable problem that identifies the issues you are trying to
address. It is not always easy to formulate the research problem simply and clearly. In some
areas of scientific research the investigator might spend years exploring, thinking, and
researching before they are clear about what research questions they are seeking to answer.
Many topics may prove too wide-ranging to provide a researchable problem. Choosing to study,
for instance a Public health issue such as maternal mortality ratio, does not in itself provide a
researchable problem at particular district level. The problem is too wide-ranging for one
researcher to address because it required large denominator. Therefore it will be time and other
resources unfeasible and the results from such a study would consequently lack depth and focus.
17 Steps in formulation of a research problem : Working through these steps presupposes a
reasonable level of knowledge in the broad subject area within which the study is to be
undertaken. Without such knowledge it is difficult to clearly and adequately ‘dissect’ a subject
area. Table : Show Steps in formulation of a research problem Steps Descriptions Step 1 Identify
a broad field or subject area of interest to you Step 2 Dissect the broad area into sub areas. Step 3
Select what is of most interest to you. Step 4 Raise research questions. Step 5 Formulate
objectives. Step 6 Assess your objectives. Step 7 Double check your research question 4.3
Statement of Research Problem An adequate statement of the research problem is one of the
most important parts of the research. Different researchers are likely to generate a variety of
researchable problems from the same situation since there are many research issues that can arise
out of a general problem situation. Your research will be able to pursue only one in depth. For a
problem statement to be effective in the planning of applied research it should have the following
characteristics. 1. The problem reflects felt needs 2. The problem is non-hypothetical, i.e. it must
be based on factual evidence 3. It should suggest meaningful and testable hypotheses – to avoid
answers that are of little or no use to the alleviation of the problem 4. The problems should be
relevant and manageable Formulating the research problem allows you to make clear, both to
yourself and the reader, what the purpose of your research is. Subsequent elaboration of method
should be oriented to providing information to address that problem. The problem statement is
therefore a very important device for keeping you on track with your research. It is also one
means by which your research will be evaluated – does the research address the problem as
stated. 5. Literature Review Literature review is an important part of any research. It gives a
foundation for the research by enlightening the researcher about the process. You need to
provide evidence that you are aware of the current trends and issues in your area of interest and
cognizant of the current state of knowledge on the subject. 18 5.1 Literature review as sources of
information Any research, no matter the scope of the research, requires reading about: what
other people have done in the area of your interest, how they have done it, what are the gaps
in the research in that area Any new generation of knowledge is based on previous and existing
knowledge. To write about your findings you need to have information either to support or refute
your arguments. Ideally this reading should form the basis for choosing your research
methodology. Reading continues and spills over into the stage of data collection and analysis.
Some reading is naturally required at the stage of interpretation of the qualitative or quantitative
data to embed your research in its settings. Moreover, every new articles you read will have a
cumulative effect on the existing body of knowledge in your mind, pushing to you reinterpret or
rethink some of your previous assumptions or ideas. But there is a limit to the changes you can
make in your literature review. The best way is to prepare a first draft and then polish it at the
stage of data interpretation when you will be reading your literature review again to present your
analysis in a coherent manner with cross-references related to your literature review. 5.2
Important Reasons to review literature Some reasons for including a literature review in your
research paper/thesis are: Essential preliminary task in order to acquaint yourself with the
available body of knowledge in your area of interest. Literature review is integral part of entire
research process and makes valuable contribution to every operational step. Reviewing literature
can be time-consuming, daunting and frustrating, but is also rewarding. Its functions are: 19
Summary points of functions of Literature review Reason for Literature review Descriptions
Bring clarity and focus to research problem Better understand the subject area To
conceptualize research problem clearly and precisely To understand the relationship between
research problem and the body of knowledge in the area Contextualization shows awareness of
the work going on in field and identifies the niche to occupy Improve your methodology If
similar procedures and methods others used to the ones that you are proposing, to learn what
procedures and methods have worked well for them, what problems they have faced with them.
To select a methodology that is capable of providing valid answer to your research questions
Broaden your knowledge It ensures to read widely around the subject area to conduct the study
As you are expected to be an expert in your area of study, it helps fulfill this expectation. It
also helps to understand how the findings of the study fit into the existing body of knowledge.
Contextualize your findings How do your answer research questions compare with what others
have found? What contribution have been able to make in to the existing body of knowledge?
How your findings different from those of others? To answer these questions go back to
literature review It is important to place findings in the context of what is already known in
field of enquiry. 5.3 Procedure for reviewing the literature: 1. Search for existing literature in
your area of study; 2. Review the literature selected; 3. Develop a theoretical framework; 4.
Develop a conceptual framework. 5.4 Sources of literature review Search for existing literature:
To effectively search for literature in your field of enquiry, it is imperative that you have in mind
at least some idea of broad subject area and of the problem you wish to investigate, in order to
set parameters for your search. 20 Articles in Journals Books Internet Research Reports
Government Documents Abstracts Reviews Unpublished Theses Electronic Research
Information Center/Social Sciences Citation Index/Dissertation Abstracts Index Most of this
material is now available on the Internet. Libraries in established higher education institutions
are another rich resource for locating the required materials. 5.5 Review the literature selected:
After identifying books and articles as useful, the next step is to start reading them critically to
pull together themes and issues that are associated. If you do not have a theoretical framework of
themes in mind to start with, use separate sheets of paper for each article or book. Once you
develop a rough framework, slot the findings from the material so far reviewed into that
framework, using a separate sheet of paper for each theme of that framework. As you read
further, go on slotting the information where it logically belongs under the theme so far
developed. You may need to add more themes as you go. Read critically with particular
reference to the following aspects: Note whether the knowledge relevant to your theoretical
framework is confirmed beyond doubt. Note the theories put forward, the criticisms of these
and their basis, the methodologies adopted and the criticisms of them. Examine to what extent
the findings can be generalized to other situations. Ascertain the areas in which little or nothing
is known-the gaps that exist in the body of knowledge. 5.6 Writing up the literature reviewed In
order to comply with the first function of literature review i.e. to provide theoretical background
to your study: List the main themes that have emerged while reading literature Convert them
into subheadings. These subheadings should be precise, descriptive of the theme in question, and
follow a logical progression. 21 Now, under each subheading, record the main findings with
respect to the theme in question, highlighting the reasons for and against an argument if they
exist, and identify gaps and issues. In order to comply with the second function of literature
review i.e. contextualizing the findings of your study requires you to very systematically
compare your findings with those made by others. Quote from these studies to show how your
findings contradict, confirm or add to them. It places your findings in the context of what others
have found out. This function is undertaken when writing about your findings i.e. after analysis
of your data 5.7 Presenting the Literature Review Literature review is specific piece of writing
designed to inform readers of a topic, establish credibility as a researcher, and argue the need for,
and relevance of the research work. Most find it difficult task that takes patience, practice, drafts,
and redrafts (O’Leary, 2004:84) O’ Leary’s quote encapsulates the essence of the literature
review. You have to be certain that the review is adequate and sufficient. The literature review
should have a definite structure based on the arguments that you want to make or the task that
you want to accomplish through it. Prepare a blue print of the structure of your research. This
can be revised as your thinking evolves but a basic outline would help you to begin writing. You
can also use it to fill in the bits and pieces as your reading advances or your theoretical
perspectives develop. Use empirical evidence to substantiate your argument (Gay, 2000). This
will authenticate your carriage. Be judicious in the selection of relevant literature and avoid the
temptation to include every single source that you have accessed. Moreover, be careful to be
recent and up to date in citing. With the internet revolution, the pace of knowledge generation
and meaning making has multiplied manifold and access to the internet is also within reach. One
valid suggestion for writing a good literature review is to read a few well written literature
reviews. Because it will give an idea about how to go about it. Finally, literature review is an
ongoing process that overlaps other stages of your research. It will then support your arguments
and contribute to your analysis and interpretation of the data. In public health, literature review is
usually a separate chapter. Sometimes it is built into the introduction section. However, in
sciences, the literature review is usually part of the introduction. The role of literature remains
the same both in sciences and public health. 22 5.8 Sample Literature Review If you are doing
action research, a good way to begin your literature review would be to define what it is. Action
Research Action research can be defined as the process of studying a public health situation to
understand and improve the quality of public health services…. It is a systematic and orderly
way for public health experts to observe their practice or to explore a problem and a possible
course of action…. Action research is also a type of inquiry that is pre-planned, organized, and
can be shared with other public health experts. The world average infant mortality rate is about
42/1000 live births; and it varies from 3/1000 live births in some highly developed countries
such as Japan and Sweden to about 182/1000 live births in least developed countries such as
Afghanistan and Sierra Leone ( UN Inter-agency Group, 2011, Jupitor,2008). Some of the world
highest rates of infant mortality are found in sub-Saharan Africa, especially in West and Central
Africa where 92 deaths per 1,000 live births in 2009 ( UN Inter-agency Group,2011). In Eastern
and Southern Africa, infant mortality has fallen 33 per cent since 1990, from 103 deaths per
1,000 in 1990 to 69 per 1,000 in 2009. Table 1 summarizes the reduction of infant mortality rates
between the years 1990 and 2009 in different regions of the world. The development has been
slowest in subSaharan Africa as compared to other regions of the world. Moreover, the declines
of infant and Perinatal mortality were not uniform and varied greatly in size, timing, and pace in
this region. Table - Reduction of infant mortality rate (IMR) during 1990-2009 in the world
Region Reduction of IMR /1000 live births Sub- Saharan Africa 109 to 81/1000 West and
Central Africa 118 to 92/1000 Eastern and Southern Africa 103 to 69/1000 Middle East and
North Africa 57 to 32/1000 South Asia 89 to 55/1000 East Asia and the Pacific 40 to 21/1000
Latin Americans the Caribbean 41 to 19/1000 Central and Eastern Europe 42 to 19/100
Industrialized countries 8 to 5/1000 The world 62 to 42/1000 Source: Child Mortality report inter
agency group 2011 23 In Ethiopia, infant and Perinatal mortality rates remain high compared to
some eastern African countries. National level infant and Perinatal mortality rate was estimated
59 /1000 live births and 46/1000 births respectively (EDHS, 2011). Substantial variations in
infant mortality rates were also observed across the regional states in the country. Learning
Activity 1.3 After the delivered the lecture, if possible, take the students to the e-library; if not
possible, arrange 5 to 6 Published Articles. Make each of these Journal articles should be related
to one general theme. Use 5-6 themes such as women empowerment and higher education in
Ethiopia, gender equality issues, primary education, and socioeconomic issues and women
empowerment in Ethiopia. Give these sets of research ideas to different groups. Ask each
students to develop a research topic for himself/herself and then do a brief literature review using
the set of literatures provided to his or her group. Comment 1.3: Please read related literature to
the topics given above and synthesis literature part of your research topic. please follow steps
given in the main body of the literature to prepare your literature review. 5.9 Conceptual
frameworks Conceptual frameworks are sometimes called “research” or “theoretical”
frameworks. Conceptual frameworks take a broad view of the program itself in order to clarify
the relationship of its activities and its main goals to the context in which it operates. It is also a
useful tool for identifying and illustrating a wide variety of factors and relationships that may
affect program success. The design should show the interrelationships between all factors that
are relevant to achieving the program’s goals. These factors can be systems, organizations,
government or institutional policies, infrastructure, population characteristics, or other features
of the operational landscape that may help or hinder the program’s success. Constructing a
conceptual framework clarifies your assumptions about the causal relationships between
significant features of the program context, clarifying aspects that your planned intervention may
affect as well as other factors beyond your control. Identifying the variables that factor into
program performance, and organizing the explicit ways they interact with each other sets the
stage for outlining the objective results you can reasonably expect from your program activities.
Clarifying all of these issues is a critical step toward designing valid measures for analyzing, or
evaluating, the success of those interventions. A common tactic here is to ‘unpack’ your idea or
problem thus generating a range of possibilities before narrowing down on one or two themes.
Following the suggestions of Punch (1998) steps could be: 24 1. write down the all the concepts
involved, and all the sub-questions you can think of pertaining to the issue. Reading around your
research idea will help to generate questions and information and to identify themes and potential
information sources 2. subdivide your questions where possible; split wide general questions into
smaller ones 3. begin to order questions and develop focus: group questions together under
common themes, separate general and specific questions 4. start to trim by selecting those
questions that you wish to deal with, consider the resources that will be available to you 5.
collate these thoughts within a loose conceptual framework this shows how questions and themes
are related and may help guide your thinking at a later stage. This process of thinking wide and
then focusing and delimiting your questions, should result in a handful of research questions that
you wish to investigate. These may still need further modification to render them answerable;
they may need to be operationalize. Note: there are no right or wrong answers in such an
exercise; the purpose is to get you thinking about as many facets of your research idea as
possible. It should also cause you to question some of the concepts you might previously have
accepted as given. Learning activity 1.4 Develop conceptual framework for Get into groups of 4
or 5 students. • Develop conceptual framework for utilization of family planning/maternal and
child health services. • Identify factors that influence health service utilization • Arrange factors
into a conceptual framework • After 30 minutes, a member of each group will share the
framework with all participants Comment 1.4: See conceptual framework annexed at the end of
this module. 6.0 Managing the literature The bibliography/reference at the end of a recent article
or book can provide you with an adequate reading list of most relevant materials related to the
topic. Once 25 the relevant literature has been located and found through the sources suggested,
the next step is to manage it. This requires efficient and selective reading. Once you try locating
the relevant literature, you will discover that it is available in vast quantities. Now the primary
task is to pick out the material that is actually related and relevant to your research area. This
requires you to be a proficient reader who can get the gist of things quickly as you will have to
go through a lot of reading. In an article published in a journal, the first thing to do is look at the
abstract or summary of the article. This will give you an idea whether it is relevant for you. In
the case of a book you should look at the list of contents, the blurb, the summaries usually given
at the end of the chapters and the introduction. This will tell you quickly if any part is pertinent
for you. The next step is to follow a clear system of keeping track of your reading references.
You need to create a management system that will incorporate your sources with all relevant
details including a note about where you found that article or book. This means that you put
down that the article/book was on the shelf or in the library you have created. A number of
computer programs like Endnote, Procite, or Reference Manager are available to keep track of
your references. This will save a lot of hassle and futile finger biting at the end when you
desperately need a page number for a quote and cannot remember where you had found the
material. These programs can automatically format references in any number of styles once the
basic details have been entered. It is also useful to annotate your references. You can write brief
notes on the Index Cards. This will provide you with a methodical and organized review of
materials that you have read. It is a record of the accuracy, relevance, and quality of the sources
you have consulted. Your critical comment, that is your informed and considered evaluation, can
also be added. These annotations can be varying in length and content depending on the
pertinence of the material reviewed. It will minimize the time when you have to incorporate them
in your writing. 7. Formulating Research Questions, Hypotheses, And Objectives Research
questions If you have gone through the ‘brainstorming’ process described above you will
probably have a number of questions which are close to your requirements as research questions
they appeal to your interests and are likely to be answerable within your time and resource
constraints. Phrase your questions so that they are simple and clear. It is usually best to pose only
one or two questions. Do not pose more than three questions at a time because it leads to much
larger research project. Start questions with terms like, how, who, what, why, and when.
Questions should be specific, not vague. Example research questions What factors determine
the quality of laboratory services? 26 Is there a relationship between the training of laboratory
personnel and the quality of their diagnosis? Is the quality of laboratory diagnosis influenced
by high workload on laboratory personnel? The questions above represent wide differences in
scope and complexity and hence will make very different demands on research resources. Some
of these questions might later need to be modified to fit resources (abilities, time, finance,
equipment) while still addressing research problem. Research hypotheses What is a hypothesis?
A hypothesis is a tentative answer to a research problem that is advanced so that it can be tested.
When do we use a hypothesis? It is appropriate to use a hypothesis when you are testing a
theory. Definition of theory is very broad – it could be ‘an idea about how things relate to each
other’. If you have an expectation of how your research question will be answered (the outcome)
then it is fair to say you have a theory in mind. If you ask of your research question ‘What is the
expected outcome?’ and have an answer, you can ask why? What is my thinking behind this
prediction? This is essentially the theory that you will be testing. If you are not able to predict the
answer to your question then your approach is not one of theory testing and you should not
proceed with developing hypotheses to test. Your research questions remain as such. This will be
the case if your research is descriptive or exploratory in nature. Developing a hypothesis from a
research question Developing hypotheses requires that you identify one character, variable of a
sampling unit that causes, affects, or has an influence on, another character, variable of the same
or other sampling units. The character, variable that affects other variables or sampling units is
called the independent variable. The character, variable which is affected by the independent
variable is called the dependent variable or response variable. Note that although for the
purposes of research methodology some variables may be called ‘dependent’ when investigating
their relationship with other ‘independent’ variables, this does not imply the existence of a causal
(as compared with associative) relationship unless strict rules of research design are followed.
This issue is discussed in more detail later in the module. 27 Good hypotheses There are two
criteria for good hypotheses. One, hypotheses are statements about relationships between
variables. Two, hypotheses carry clear implications for testing the stated relationships. These
criteria mean, then, that hypothesis statements contain two or more variables that are measurable
or potentially measurable and that they specify how the variables are related. Research without
hypotheses In exploratory research our base knowledge of a subject may be so low that we
cannot formulate meaningful hypotheses. Nonetheless, exploratory research should be guided by
a clear sense of purpose. Instead of hypotheses, the design for the exploratory study should state
its purpose, or research objectives as well as criteria by which the exploration will be judged
successful. For example, if we would like to start and encourage mothers to make use of family
planning services, we may first need to know the social norms or cultural issues of the
community before we begin making meaningful hypotheses. We can state that our exploratory
study would have the purpose of generating hypotheses about personal characteristics which
correlate with the adoption/rejection of use of family planning services, the composition of the
decision-making unit, and the factors which influence the decision either to adopt or reject.
Success would be measured in terms of generating testable hypotheses. Interpretative research,
which seeks to develop knowledge through understanding meaning, does not usually proceed
with hypotheses. Research objective(s) What are the research objectives? In general, research
objectives describe what we expect to achieve by a project. Research objectives may be linked
with a hypothesis or used as a statement of purpose in a study that does not have a hypothesis.
Even if the nature of the research has not been clear to the layperson from the hypotheses, s/he
should be able to understand the research from the objectives. A statement of research objectives
can serve to guide the activities of research. Consider the following examples. Objective: To
describe what factors mother take into account in making such decisions as whether to use a new
family planning services or not. Objective: To develop a budget for reducing pollution by a
particular enterprise. In the above examples the intent of the research is largely descriptive. In
the case of the first example, the research will end the study by being able to specify factors
which emerged in mothers decision. 28 In the second, the result will be the specification of a
pollution reduction budget. These observations might prompt researchers to formulate
hypotheses which could be tested in another piece of research. So long as the aim of the research
is exploratory, ie to describe what is, rather than to test an explanation for what is, a research
objective will provide an adequate guide to the research. Examples of research statements
Session summary In this introductory session of the module, you have learned about the different
definitions of research, why research is done, where it is conducted and what researchers use. At
the end of the section, you have acquired basic knowledge about research methods and skills
based on your educational background and professional experiences. You can easily differentiate
among perspectives on and approaches for classifying different research types. The classification
will help you understand different concepts related to research described in later sessions. In the
literature review, you have learned what a literature review is, and why review of literature is
necessary for any research. You have also been introduced to various sources for conducting
literature review. It is also discussed how literature review can be managed well. To facilitate
learning, a brief sample of annotation and a sample literature review are also provided. 29
Learning activity 1.5 Learning activity 1.5.1 Match the research types with the following
descriptions Learning activity 1.5.2: Assessment would be done through the outcome of the
activity of writing literature review as provided in the example above for your thesis desertions.
Learning activity 1.5.3: Identify the independent and dependant variables in the following
situation: A university teacher is interested in determining the best way to teach introductory
technology course, and ensure that her students have learned the material. Learning activity
1.5.4: Formulate hypotheses for the following two situations: Situation 1: What effects does
viewing violence on television have on boys? 30 Situation 2: A sanitation department is nearby
and the smell is coming from the same direction. The sanitation department does a series of
steps to process the sewage and wastewater. Hypothesis: If
__________________________then___________________________. 31 Session two:
Quantitative Research Methods Session introduction This session describes how researchers
shall collect, organize and analyze data that are relevant to the purpose of the study. There is a
need to organize the methods section into logical subsections that illustrate the steps to be
followed during collection, organization and analysis of data. During proposal development, you
need to describe what you want to do using future tense. Original methods shall be described in
detail otherwise give references if you want to apply standard methods. When writing the
methods section, selection of the study design appropriate for the study objectives is usually the
initial task. The design itself guides how you should proceed with data collection, management,
and analysis. Mechanisms of maintaining the quality of data shall be considered from the
inception of the research project to the final write up of a scientific paper. If the methods section
of a proposal is carefully written, there will be little work later on during the write up of a thesis
report or manuscript because usually it is a matter of changing to past tense. Learning outcomes
When you have studied this session, you should be able to: Identify the study design(s)
appropriate for different research objectives Describe to which population groups study
findings could be generalized Describe the minimum number of study units required for
different studies and how they can be selected Describe the variables of a study and how they
can be measured Describe the different data collection technique(s) Develop a questionnaire
for data collection Describe what issues need to be considered during the analysis of data
Describe the mechanisms of maintaining the quality of data 2.1 Choosing study design Study
design is a logical model that guides the investigator in the various stages of the research
process. Once the researchers identify the study design, they will know how to proceed
conducting that research. There are different quantitative and qualitative study designs with their
own advantages and limitations. Selecting the appropriate study design shall be the first task of
the researchers after they know their objectives. This session will not present the details of the
different study designs. The learner shall refer the Epidemiology module to learn about the 32
quantitative research designs. Additionally, it is good if you refer other modules on qualitative
research to learn about the qualitative study designs. But, Session five of this module also
includes qualitative methods where there is slight touch of qualitative designs The following
criteria may be considered during selection of study design(s) 1) Research problem 2) Personal
experience 3) Resources available for the study 1) Research problem This is the major criteria
that should be considered during selection of study designs. For example quantitative designs
would be appropriate if the interest is: estimating the prevalence of a specific condition in a
specific locality assessing associations between the different independent variables and a
dependent variable Assessing the effectiveness of a new drug, vaccine etc.... Major quantitative
designs include Cross sectional, Case control, Cohort, and Intervention studies. On the other
hand qualitative designs would be appropriate when: • the topic is new • the topic has never been
addressed with a certain sample or group of people, and • existing theories do not apply with the
particular sample or group under study Qualitative approaches allow room to be innovative and
to work more within researcher-designed frameworks. They allow more creative, literary-style
writing. The major qualitative study designs include ethnography, phenomenology, grounded
theory, participatory action research, and case study. 33 A mixed methods design is useful when
either the quantitative or qualitative approach by itself is inadequate to best understand a research
problem. Mixed methods are becoming popular. The integration of quantitative and qualitative
methods permits a more complete and synergistic utilization of data than do separate quantitative
and qualitative data collection and analysis The problems addressed by social & health science
researchers are complex. The interdisciplinary nature of research calls for mixed methods.
Formation of research teams with individuals having diverse methodological interests and
approaches is required. Mixed Methods Research Designs can be used for the following
purposes: 1) Validate findings using quantitative and qualitative data sources. 2) Use qualitative
data to explore quantitative findings. 3) Develop survey instruments. 4) Use qualitative data to
augment a quantitative outcomes study. 2) Personal experience Researcher’s personal training
and experience influence the choice of design. For example, an individual trained in technical,
scientific writing, statistics, and computer statistical programs and familiar with quantitative
journals would most likely choose the quantitative design. On the other hand, individuals who
enjoy writing in a literary way or conducting personal interviews, or making up-close
observations may prefer qualitative approach. The mixed methods researcher is an individual
familiar with both quantitative and qualitative research 3) Resources available for the study Diff.
types of study design are not considered equal in the strength of evidence they provide.
Investigators may not be able to select the design that gives a high level of evidence. The design
that gives a high level of evidence may not be feasible to do or ethical to do. For example
conducting prospective cohort study produces more valid data than cross sectional study or the
traditional type of case control study. Since conducting prospective cohort study is time
consuming and costly, researchers with 34 limited time and money may opt to apply study
designs which can be completed within short time period and minimal cost 2.2 Population In
epidemiological research, measurements are taken from few people and estimates are derived
from these measurements. All kinds of errors prior, during and after the study can be responsible
for bias in the final results. This bias can be caused by measurement errors, as well as through
poorly chosen source and study populations. Bias can also be introduced during the sampling
procedure. Given that epidemiological research is generally carried out with a limited number of
people, the generalizability of the results could be limited by these types of bias. A good sample
from a well-described study population is therefore very important. When planning research,
researchers need to consider the different groups of people ranging from target population to
sample population. Target population refers to the entire group of individuals or objects to which
researchers are interested to generalize the conclusions. But, because of practicalities, entire
target population often cannot be studied. The target population usually has varying
characteristics and it is also known as the theoretical population. The target population consists
of all the individuals at large, for example in Ethiopia, with the same characteristics as the actual
sample that will be included in the study. Sometimes target population is interchangeably used
with source population. Source population is the underlying population from which the subjects
will be drawn. E.g patient population of a hospital. Study population is the population to which
the researchers can apply their conclusions. This population is a subset of the target/source
population and is also known as the accessible population. It is from this accessible population
that researchers draw their samples. E.g Female patients who are older than 50 years admitted
with a diagnosis of diabetes mellitus. Sample population is a population selected and included in
the study. Samples are subsets of study populations used in research because often not every
member of study population can be measured. However, the results drawn from the investigation
of the sample are interpreted and applied directly to the study population. 35 2.3 sampling
techniques &Sample size determination 2.3.1 Research and Sample The term sampling refers to
strategies that enable us to pick a subgroup from a larger group and then use this subgroup as a
basis for making inferences about the larger group. The researcher's goal is always to generalize
about the population based on observations from the sample. Sampling strategies not only make
it possible to collect data from a smaller number of respondents, but these strategies also make it
possible to go into greater depth with this smaller number by asking more and deeper questions
or by following up the structured questions with more open-ended or qualitative questions than
would be possible with a larger group of respondents. When using such a subgroup to make
decisions about the larger group, the subgroup must be as closely representative of the larger
group as possible. Sampling techniques are useful when we want to know how a large group
would be described with regard to several variables, but there would be major added costs,
narrow restrictions on the number of questions that could be asked, or some other difficulty in
administering the data collection procedure to every member of the target population. If we
wanted to find out how many graduate students in entire university were interested in taking a
certain advanced course next semester, it would be difficult and expensive to ask every graduate
student this question. Even if we had a list of all their names, it would be difficult to get them all
to reply to a questionnaire. Finally, we would probably have time to ask only one question, and it
would be unlikely that we would be able to uncover the qualitative reasons behind interest or
lack of interest in the course. Similarly, if we wanted to know how many number of patients
showed progress after taking a certain drug in Ethiopia, we would find it a lot more convenient to
administer our test to a sample of patients than to interview every patients. This will bring us to
sampling 2.3.2 Sampling Strategy The manner in which a sample is drawn is an important factor
in determining how useful the sample will be for making inferences about the population from
which it is drawn. It is quite possible to have a very large sample upon which no sound decision
can be based. This occurs because the respondents in the sample are not really similar to the
population about which we want to make generalizations. To be useful, the sample must be
representative of the population about which we wish to make generalizations. 36 2.3.2.1 What is
Sampling? Sampling is the act, process, or technique of selecting a suitable sample, or a
representative part of a population for the purpose of determining parameters or characteristics of
the whole population. When dealing with people, it can be defined as a set of respondents
(people) selected from a larger population for the purpose of a survey. The purpose of sampling
is to draw conclusions about populations from samples, we must use inferential statistics which
enables us to determine a population`s characteristics by directly observing only a portion (or
sample) of the population. We obtain a sample rather than a complete enumeration (a census ) of
the population for many reasons. Obviously, it is cheaper to observe a part rather than the whole,
but we should prepare ourselves to cope with the dangers of using samples. There would be
difficulties to measure the whole populations because: The large size of many populations
Inaccessibility of some of the population: some populations are so difficult to get access to that
only a sample can be used. E.g. prisoners, people with severe mental illness, disaster survivors
and so on. The inaccessibility may be associated with cost or time or just access.
Destructiveness of the observation: sometimes the very act of observing the desired characteristic
of the product destroys it for the intended use. Good examples of this occur in quality control.
For instance, to determine the quality of a fuse and whether it is defective, it must be destroyed.
Therefore if you tested all the fuses, all would be destroyed. Accuracy and sampling: A sample
may be more accurate than the total study population. A badly identified population can provide
less reliable information than a carefully obtained sample. 37 Figure 1.1: Schematic presentation
of the process of sampling 2.3.2.2 Steps in Sampling Design There are steps that we need to
follow to get in to the respondents . The steps that we need to do are that: What is the target
population? Define the target population and study population. What are the parameters of
interest? Define the parameters of interest of the study. What is the sampling frame? Select the
sampling frame. What is the appropriate sampling method? Determine which sampling method
we are going to use depending on the setting of the population and the purpose of the study.
Plan procedures to select the sampling unit Determine the size of the sample which will be
selected from the population. Select actual sampling unite Conduct field work What you want
to talk Population Sample Sampling Frame Sampling Process What you actually observe in the
data Inference 38 2.3.3 Advantage and Disadvantage of Sampling Advantage of Sampling
Sampling is a must in some situations: to diagnosis the patient for a certain diseases we don't
need to take the whole blood of the patient. In addition, if we want to inspect eggs, bullets and
missiles, we don't need to test all. This is possible only in sample study. Thus sampling is
essential when the units under study are destroyed. It saves time: Research without sampling
might be too time consuming. The population to be studied may be too large or unlimited that
it is almost impossible to reach all of them. Sampling makes possible this kinds of study.
Sampling reduces the study population to a reasonable size that expense are greatly reduced.
Sometimes the experiments are done on sample basis. The fertilizers, the seeds and the
medicines are initially tested on samples and if found useful, then they are applied on large scale.
Most of the research work is done on the samples. Sampling saves the source of data from
being all consumed. Sample data is also used to check the accuracy of the census data.
Disadvantage of Sampling If sampling is biased, or not representative or too small the
conclusion may not be valid and reliable In research the respondents of the study must have a
common characteristics which is that basis of the study If the population is very large and there
are many sections and subsections, the sampling procedure becomes very complicated If the
researcher does not possess the necessary skill and technical knowhow in sampling procedure,
then the outcome will be devastated. 2.3.4 Types of Sampling There are many methods of
sampling when doing research. One of the most important decisions that any researcher makes is
how to obtain the type of participants needed for the study. The sample that we draw for our
study determines the generalizability of our findings. As we have seen before, in most situations,
it is impossible to study an entire population. We typically study a subset of people drawn from a
larger population and use inferential statistics to make an inference from the sample back to the
population. The validity of that inference depends on how representative the sample or subset is
of the population from which it is drawn. Our goal, as researchers, is to 39 obtain the most
representative sample that we can. Some sampling strategies can get us pretty close to the
population, others have problems that might result in a biased sample. As we learn about
different sampling strategies, let's use the population drawn below as our starting point. When
we draw our sample, we want to have a good representation of all of the kinds of people in the
population. In General there are two methods of sampling: 1. Probability Sampling Method 2.
Non-Probability Sampling Method 2.3.4.1 Probability Sampling Method Probability sampling
strategies typically use a random or chance process, although there are important exceptions to
this rule. Random sampling is a strategy for selecting study participants in which each and every
person has an equal and independent chance of being selected. What does it mean to be
independent? The researchers select each person for the study separately. Let us say you were
asked to participate in an experiment, enjoyed it, and told your friends to contact the researcher
to volunteer for the study. This would be an example of non-independent sampling. We assume
that friends share similar values and by recruiting your friends to be in the study, the sample
might represent you and your friends but not the whole population of interest. The "equal
chance" and "independent" components of random sampling are what makes us confident that
the sample has a reasonable chance of representing the population. Where do we start? When we
use probability sampling, we begin by defining our population. Once we have done this, we must
have some sort of record or directory to use to select individual participants from the target
population. The sampling frame is the population as it is defined and available through records.
There are a number of probability sampling techniques that can be used depending on the types
of the population complexity we want to study. 40 Figure 2: Schematic presentation of
probability sampling methods A. Simple Random Sampling Simple random sampling is the most
straightforward of the random sampling strategies. We use this strategy when we believe that the
population is relatively homogeneous for the characteristic of interest. This sampling method has
the following properties. The population consists of N homogeneous subjects. There is frame
for the population The sample consists of n subjects or objects. All possible samples of n
subjects are equally likely to occur Procedures to select the sample How do we actually take a
random sample? The specific procedures that you follow may vary depending on your resources,
but all involve some type of random process. Depending on the complexity of the population, we
can use different tools to select n samples from the frame. These are lottery method, table of
random number (they are available in the appendix of many research methods and statistics
textbooks) or computer generated random number. Lottery method is appropriate if the total
population is not too large, otherwise if the population is too large then it will be very difficult to
use lottery method. Sampling Methods Probability Sampling Techniques Nonprobability
Sampling Techniques Simple Random Sampling Systematic Sampling Stratified Sampling
Cluster Sampling Multistage sampling 41 Table of random number or computer generated
random number is the feasible method to be used. If computer is accessible, it will be very easy
to use computer generated random number to select the sample. What you need to do is that
specify the total population(N), the sample size (n) and the range of the frame (minimum and
maximum number.) At the end the computer will display the randomly selected list of numbers
according to the sample you specified. That will be your sample. There are a number of
computer programs which can generate the list of the sample for simple random sampling. Of
this, openepi is one of the free software that we can access from the web. The required
information to use this program are: lowest value, highest value, and the sample size you want to
draw. Figure 3: OpenEpi software to generate random numbers Example 1: Consider the
following condition. The total number of patients who visit Gondar University Hospital for the
last six months in the year 2015 were N. We want to see the proportion of TB among those
patients who visited the hospital. So if we thing that those patients who visited the hospital
within the specified time period are homogeneous with respect to the variable of interest and list
of the patients are available, then we can use simple random sampling to select the sample, n. B.
Systematic Random Sampling A method of selecting sample members from a larger population
according to a random starting point and a fixed, periodic interval. We use this sampling method
when the population is homogenous and frame is not available. 42 Typically, every nth member
is selected from the total population for inclusion in the sample population. Systematic sampling
is still thought of as being random, as long as the periodic interval is determined beforehand and
the starting point is random. It is frequently chosen by researchers for its simplicity and its
periodic quality. The procedure involved in systematic random sampling is very easy and can be
done manually and the results are representative of the population unless certain characteristics
of the population are repeated for every nth individual which is highly unlikely. Hence, in the
absence of frame, this method will be the appropriate choice. Steps in systematic sampling:
Define the population Determine the desired sample List the population from 1 to N
Determine K, where k = N/n Select a random number between 1 and k, let us denote this
number by a Starting at a, take every Kth number on the list until the desired sample is
obtained. Then the selected list will be a, a + k, a + 2k, a + 3k, …., Note that: Since systematic
random sampling is a type of probability sampling, the researcher must ensure that all the
members of the population have equal chances of being selected as the starting point or the initial
subject. The researcher must be certain that the chosen constant interval between subjects do
not reflect a certain pattern of traits present in the population. If a pattern in the population exists
and it coincides with the interval set by the researcher, randomness of the sampling technique is
compromised. C. Stratified Random Sampling Stratified random sampling is used when we have
subgroups in our population that are likely to differ substantially in their responses or behavior
(i.e if the population is heterogeneous). In stratified random sampling, the population is first
divided into a number of parts or 'strata' according to some characteristic, chosen to be related to
the major variables being studied. For example, you are interested in visual-spatial reasoning and
previous research suggests that men and women will perform differently on these types of task.
So, you divide your sample into male and female members and randomly select equal numbers
within each subgroup (or "stratum"). With this technique, you are guaranteed to have enough of
each subgroup for 43 meaningful analysis. Often we used simple random sampling to select a
sample from each strata after stratification. Figure 3: Characteristics of population and sample
for stratified sampling Steps involve in stratified sampling method: Define the population
Determine the desired sample size Identify the variable and subgroups (strata) for which you
want to guarantee appropriate representation (either proportional or equal) Classify all
members of the population as a member of one of the identified subgroups Randomly select
(using simple random sampling) an appropriate number of individuals from each subgroup. Then
the total sample size will be the sum of all samples from each subgroup. There are two methods
to get the study subject from each subgroup, proportional allocation or equal allocation. We use
proportional allocation technique when our subgroups vary dramatically in size in our
population. For example, we are interested in risk taking among college students and suspect that
risk taking might differ between smokers and nonsmokers. Given increasing societal pressures
against smoking, there are many fewer smokers on campus than nonsmokers. Rather than take
equal numbers of smokers and nonsmokers, we want each group represented in their proportions
in the population. The subsample which will be taken from each subgroup can be determined by:
Let N be total population and N1, N2, …, Nk be the subtotal population for strata 1, 2, …, k
respectively. Moreover, let n be the total sample size and n1, n2, …, nk be the subsample for
strata 1, 2, …, k respectively in which: N = N1 + N + 2+, …,+NK n = n1 + n2+,…, +nk Then the
subsample ni which will be selected from subgroup Ni can be allocated equally or
proportionally. Proportional allocation cab ne done: 44 N nxN n i i Where i = 1, 2, 3,…, k
Figure 4: Schematic presentation of the distribution of sample size for each strata The higher the
population in the subgroup, the higher the sample size will be. However, equal allocation will be
used if the total population from each subgroup is approximately equal. Advantage of stratified
sampling over simple random sampling: It can provide greater precision than a simple random
sample of the same size. Because it provides greater precision, a stratified sample often
requires a smaller sample, which saves money. A stratified sample can guard against an
"unrepresentative" sample We can ensure that we obtain sufficient sample points to support a
separate analysis of any subgroup. The main disadvantage of a stratified sample is that it may
require more administrative effort than a simple random sample. 45 D. Cluster Random
Sampling Cluster sampling is useful when it would be impossible or impractical to identify every
person in the sample. If the study covers wide geographical area, using the other methods will be
too costly. The idea is, divided the total population in to different clusters and then the unit of
selection will be cluster. Therefore, total population in the selected cluster will be taken as the
sample. Figure 5: Characteristics of population and sample for cluster sampling Steps in cluster
sampling are: Define the population Determine the desired sample size Identify and define
a logical cluster (can be kebele, Got, residence, and so on) Make a list of all clusters in the
population Estimate the average number of population number per cluster Determine the
number of clusters needed by dividing the sample size by the estimated size of the cluster
Randomly select the required number of clusters (using table of random number as the total
number of clusters is manageable) Include in the sample all population in the selected cluster.
46 There are different classification of cluster sampling; Figure 6: Illustration of cluster random
sampling. The main advantage of cluster sampling is that it saves time and money and hence it is
advisable to use this method if the study covers wide geographical area. Consider the following
graphical display: E. Multistage Random Sampling This is the most complex sampling strategy.
The researcher combines simpler sampling methods to address sampling needs in the most
effective way possible. For example, the administrator might begin with a cluster sample of all
schools in the district. Then he might set up a stratified sampling process within clusters. Within
schools, the administrator could conduct a simple random sample of classes or grades. By
combining various methods, researchers achieve a rich variety of results useful in different
contexts. 2.3.4.2 Non-Probability Sampling Method Most probability sampling strategies have a
random or chance component, though there are some important exceptions. It is this random
component that gives us confidence that our sample is a reasonably good representation of the
population. This random component can be time-consuming and expensive. Hence, in the
presence of constraints, the alternative sampling method is nonprobability sampling method.
Nonprobability sampling strategies are used when it is practically impossible to use probability
sampling strategies. This typically occurs because of time and expense constraints and the lack
of an adequate sampling frame. 47 Nonprobability sampling is also used when the frequency of
the behavior or characteristic of interest is so low in the population that a more targeted strategy
is needed to find sufficient numbers of participants for the research. Figure 7: Schematic
presentation of non-probability sampling methods A. Purposive Sampling Purposive sampling
targets a particular group of people. When the desired population for the study is rare or very
difficult to locate and recruit for a study, purposive sampling may be the only option. For
example, you are interested in studying cognitive processing speed of young adults who have
suffered closed head brain injuries in automobile accidents. This would be a difficult population
to find. Figure 8: Schematic presentation of purposive sampling Purposive sampling
Conveniences sampling Quota Sampling Snowball Sampling Nonprobability Sampling
Techniques 48 The major problem with purposive sampling is that the type of people who are
available for study may be different from those in the population who can't be located and this
might introduce a source of bias. B. Convenience Sampling Convenience sampling selects a
particular group of people but it does not come close to sampling all of a population. The sample
would generalize only to similar programs in similar cities. Convenience sampling is widely used
in student research projects. Students contact professors that they know and ask if they can use
their classes to recruit research subjects. Convenience sampling looks just like cluster sampling.
The major difference is that the clusters of research participants are selected by convenience
rather than by a random process. C. Judgment Sampling The researcher selects the sample based
on judgment. This is usually and extension of convenience sampling. For example, a researcher
may decide to draw the entire sample from one "representative" city, even though the population
includes all cities. When using this method, the researcher must be confident that the chosen
sample is truly representative of the entire population. D. Quota sampling It is the nonprobability
equivalent of stratified sampling. Like stratified sampling, the researcher first identifies the
stratums and their proportions as they are represented in the population. Then convenience or
judgment sampling is used to select the required number of subjects from each stratum. This
differs from stratified sampling, where the stratums are filled by random sampling. E. Snowball
sampling It is a special nonprobability method used when the desired sample characteristic is
rare. It may be extremely difficult or cost prohibitive to locate respondents in these situations.
Snowball sampling relies on referrals from initial subjects to generate additional subjects. While
this technique can dramatically lower search costs, it comes at the expense of introducing bias
because the technique itself reduces the likelihood that the sample will represent a good cross
section from the population. What we need to do in case of snowball sampling is that first
identify someone who meets the criteria and then let him/her bring the other he/she knew. 2.3.5
Sample size determination Determining the sample size for a study is a crucial component of
study design. The goal is to include sufficient numbers of subjects so that statistically significant
49 results can be detected. Statistically inconclusive findings make it difficult to determine
whether a particular treatment or intervention was effective and to identify directions for future
studies. Studies with insufficient subjects also may result in potentially important research
advances that go undetected. In statistical language, these studies are referred to as "under-
powered." That is, the probability that they will detect an existing treatment effect is lower than
optimal. Among the questions that a researcher should ask when planning a survey or study is
that "How large a sample do i need?". The answer will depend on the aims, nature and scope of
the study and on the expected result. All of which should be carefully considered at the planning
stage. In general, sample size depends on: The type of data analysis to be performed The
desired precision of the estimates one wishes to achieve The kind and number of comparisons
that will be made The number of variables that have to be examined simultaneously How
heterogeneous the sampled population is The type of outcome variable under study should also
be taken in to account. There are three possible categories of outcome variables. The first is
where the variable of interest has only two alternatives response: yes/no, dead/alive,
vaccinated/not vaccinated and so on. The second category covers those outcome variable with
multiple, mutually exclusive alternatives responses, such as marital status, religion, blood group
and so on. For these two categories of outcome variables, the data are generally express as
percentages or rates. So we can use percentage to compute the sample size. The third category
covers continuous response variables such as birth weight, age at first marriage, blood pressure
and cerium uric acid level, for which numerical measurement are usually made. In this case the
data are summarize in the form of means and standard deviations or their derivatives. The
statistical method appropriate for sample size determination will depend on which of these types
of outcome variables the investigator is interested in. Deciding on a sample size for qualitative
inquiry can be even more difficult than quantitative because there are no definite rules to be
followed in qualitative study. It will depend on what you want to know, the purpose of the
inquiry, what is at stake, what will be useful, what will have credibility and what can be done
with available time and resources. You can choose to study one specific phenomenon in depth
with a smaller sample size or a bigger sample size when seeking breadth. 2.3.5.1 Sample Size for
Single Population Mean This is the condition in which the research question is about mean
(continuous variable). Three questions must be answered to determine the sample size: 50 1.
Standard deviation of the population: It is rare that a researcher knows the exact standard
deviation of the population. Typically, the standard deviation of the population is estimated: a.
from the results of a previous survey, b. from a pilot study, c. from secondary data, d. from
judgment of the researcher. 2. Maximum acceptable difference: This is the maximum amount of
error that you are willing to accept. That is, it is the maximum difference that the sample mean
can deviate from the true population mean before you call the difference significant. 3. Desired
confidence level : The confidence level is your level of certainty that the sample mean does not
differ from the true population mean by more than the maximum acceptable difference.
Commonly we use a 95% confidence level. Then the sample size determination formula for
single population mean is defined as: 2 2 2 2 ( ) Z n Where n is the sample size for
single mean α=The level of significance which can be obtain as 1-confidence level.
σ=Standard deviation of the population w=Maximum acceptable difference Zα/2=The value
under standard normal table for the given value of confidence level 2.3.5.2 Sample Size for
Single Population Proportion This is the situation in which the variable of interest is categorical.
Three questions must be answered to determine the sample size for single population proportion:
1. Best estimate of population proportion of the variable of interest : Make your best estimate of
what the actual percent of the survey characteristic is. The possible source of this proportion are:
a. From the results of a previous study, b. From a pilot study, c. Judgment of the researcher. 51 d.
Simply taking 50% 2. Maximum acceptable difference: This is the maximum percent difference
that you are willing to accept between the true population rate and the sample rate. Commonly
we use a difference of 5 percent. That is, if your survey finds that 25 percent of the sample has a
certain characteristic, the actual rate in the population may be between 20 and 30 percent. 3.
Desired confidence level: How confident must you be that the true population rate falls within
the acceptable difference. This is the same as the confidence that you want to have in your
findings. If you want 95 percent confidence, you should enter 95. This means that if you took a
hundred samples from the population, five of those samples would have a rate that exceeded the
difference you specified in the previous question. 52 Then the formula for the sample size of
single population proportion is defined as: 2 2 2 ( ) (1 ) Z p p n Where n is the sample
size for single mean α=The level of significance which can be obtain as 1-confidence level.
p= Best estimate of population proportion w=Margin of error or maximum acceptable
difference - The margin of error (w) measures the precision of the estimate - Small value of w
indicates high precision - It lies in the interval (0%; 5%] - For p close to 50%, w is assumed to be
close to 5% - For smaller value of p, w is assumed to be lower than 5% Zα/2=The value under
standard normal table for the given value of confidence level - For α = 0.05 the Z0.025 = 1.96 -
For α= 0.1 the Z0.05 = 1.645 and so on Note that: The reason behind taking 50% proportion for
the variable of interest is that this is the value which gives the maximum product when we
compute the sample size. Hence try to find all possible effort to get the estimated proportion
from previous study. Example: A survey is being planned to determine what proportion of
families in a certain area are medically indigent. It is believed that the proportion cannot be
greater than 0.35. A 95% confidence interval is desired with w= 0.05. What sample size should
be selected for the study? Solution: 350 (0.05) (1.96) 0.35 (1 0.35) ( ) (1 ) 2 2 2 2 2 x x
Z p p n The size 350 is not the final sample for the study. There are conditions that we have
to account. Different Some Considerations 53 The final sample size should consider the
following conditions: Nonresponse, lost to follow up, lack of compliance and so on. Depending
on the sensitivity of the research problem, up to 15% nonresponse can be accounted. Consider
the total size of the population (N): if N < 10000 then finite population correction formula can be
used which is defined by: N n n nf 0 0 1 Where nf=final sample size, n0=sample size from
the above formula and N=total population. Design effect: Is a correction of bias in the variance
introduced in the sampling design, by selecting subjects due to the use of clusters. The design
effect can be calculated after study completion, but should be accounted for at the design stage. It
compares the variance of simple random sampling and variance of cluster sampling; n P P Var
SRS (1 ) ( ) ( 1) ( ) ( ) 2 k k p P Var Cluster i Then Design effect can be computed as
Var SRS Var cluster Design ( ( ) Effect Where p=global proportion, pi=proportion of the
ith cluster, n=number of subjects k=number of strata The design effect is 1 (no design effect)
when taking a simple random sample. The design effect varies using cluster sampling; it is
usually estimated that the design effect is 2 in cluster sampling. 54 2.3.5.3 Sample size for two
population Equal Sample size for the Difference of proportions Here the objective of the study is
to check whether there is significant difference between two proportions coming from two
different population. The sample size to be taken from each group is assumed to be equal. Thus
the sample size which will be randomly taken from each group can be determined by: 2 2 1 2 1 1
2 2 1 / 2 ( ) ( )( ) (each group) p p p q p q Z Z n Where: p1 =current estimate of
population proportion P1 (Non-exposed or control group) q1 = 1-p1 p2 = current estimate of
population proportion P2 (exposed or treated group) q2 = 1- p2 Zα/2=the Z value
corresponding to the alpha error. This is the value at two tailed. Zβ=the Z value corresponding
to the beta error. The Z value for beta is always based on one tailed test. So if beta is 0.05, 0.1,
0.2, or 0.3, then the corresponding z values are 1.65, 1.28, 0.85 and 0.52 respectively Example :
An investigator wants to determine if the fatality rate among cancer patients undergoing standard
therapy differs from the case fatality rate in cancer patients receiving new treatment. The case-
fatality rate among cancer patients undergoing standard therapy is 0.80, and is 0.60 for cancer
patients receiving a new treatment. Find the required sample size to test a hypothesis that the
case-fatality rate differed between groups at the stated level of significance, alpha = 0.05, and
desired power of the test, 0.90. Solution: From the given information, - p1 =0.60 - q1 = 1-
0.60=0.40 - p2 = 0.80 - q2 = 1- 0.80=0.20 - Zα/2 = 1.96 55 - Zβ = 1.28 388.4 (0.8 0.7) (0.7 0.3
0.8 0.2)(1.96 1.28) ( ) ( )( ) (each group) 2 2 2 2 1 2 1 1 2 2 1 / 2 x x p p p q p
q Z Z n T hus, 389 standard treatment users and 389 new treatment users needed for the
study 2.3.5.4 Sample Size using Statistical Software As an alternative method, statistical
softwares calculate the sample size required for the study. There are many softwares which can
be used to determine sample size including Openepi, and EPI-Info. These are commonly used in
area of health because of their simplicity to use and understand easily. 56 2.4 Variables of the
study & operational definition A variable is defined as a characteristic of a person, object or
phenomenon, which can take on different values. When one tries to assess associations between
variables, there is a need to identify dependent and independent variables. Response (dependent,
outcome) variable is a variable which can be affected by explanatory/independent variables. On
the other hand, explanatory (independent, predictor) variable is a variable that explain the
response variable. Many variables can easily be measured but for some variables it is sometimes
not possible to find meaningful categories unless the variables are made operational with one or
more indicators. Operationalizing variables means that you make them measureable. For
example, if a researcher reported that 60% of the women are knowledgeable on family planning
methods, readers may have different understanding on this report unless there is standard
definition for it or operationally defined. Knowledge is not usually assessed by asking only a
single question. There is a need to develop a series of questions. The answers to these questions
form an indicator of someone’s knowledge on the issue which can then be categorized. For
example, if 10 questions were asked, you might decide that the knowledge of those with 0 to 3
correct answers is poor; 4 to 6 correct answers is satisfactory; and 7 to 10 correct answers is
good. 2.5 Data collection methods/techniques The commonly used data collection
techniques/methods include: • Using available information (Document review) • Observation •
Interviewing (face-to-face) • Administering written questionnaires • Focus group discussions
2.5.1 Using available information (Document review) Analysis of the information routinely
collected by health facilities is very useful for identifying problems. Researchers can also
identify other sources of available data like newspapers, published case histories etc. The major
advantage of available data is that collection is inexpensive. But, as a limitation, it is sometimes
difficult to gain access to records or reports. Additionally, data may not always be complete and
precise enough, or is too disorganized. 57 2.5.2 Observation Observation is a technique that
involves systematically selecting, watching and recording behavior and characteristics of living
beings, objects or phenomena Observation of human behavior can be undertaken in different
ways: A) Participant observation. Observer takes part in the situation he or she observes Eg. A
nurse hospitalized with fracture who now observes patient care practices in the ward she is
admitted B) Non-participant observation. Observer watches the situation, openly or concealed,
but does not participate Observation can give additional, more accurate information on behavior
of people than interviews or questionnaires. As observations are time consuming, they are most
often used in small scale studies. Observations can also be made on objects like observing the
presence or absence of latrines and the state of cleanliness 2.5.3 Interviewing It involves oral
questioning of respondents, either individually or as a group. Answers can be recorded by
writing them down, tape-recording, or a combination of both. Interviews can be structured,
unstructured, or semi-structured Structured interviews (standardized interviews). This type of
interview is often used in quantitative research. There is low degree of flexibility during the
conduct of interview. The same sets of questions are asked, in the same order, using the same
words, to different interviewees. Unstructured interviews These are interviews without
predetermined set of questions. It is often used in combination with observation Researchers and
interviewees talk freely. Interview is flexible and highly responsive to individual differences and
emerging new information. Researchers have to generate relevant questions based on their
interaction with the interviewees Semi-structured interviews (Guided Interview) In this type of
interview, researchers prepare interview guides that consist of a set of questions to initiate
discussion. Researchers generate other questions in interesting areas of inquiry during the
interviews. It is widely used as the qualitative interview method Qualitative Interview (In-depth
interview) Advantages of in-depth interview • Usually yield richest data • Permit face-to-face
contact with respondents • Provide opportunity to explore topics in depth 58 • Flexible according
to particular individuals or circumstances - allow explaining or clarifying questions
Disadvantages of in-depth interview • Expensive and time-consuming • Need well-qualified
(highly trained) interviewers • Interviewee may distort information due to the desire to please
interviewer • Flexibility can result in inconsistencies across interviews • Volume of information
too large; may be difficult to transcribe and reduce data 2.5.4 Administering written
questionnaires (self-administered questionnaire) Written questions are presented that are to be
answered by the respondents in written form. A written questionnaire can be administered in
different ways, such as by sending questionnaires by mail; gathering all or part of the
respondents in one place at one time, giving oral or written instructions, and letting them fill out
the questionnaires; and hand delivering questionnaires to respondents and collecting them later
Advantages of self-administered questionnaire • less expensive • permits anonymity & may
result in more honest responses • does not require research assistants • eliminates bias due to
phrasing questions differently with different respondents Disadvantages of self-administered
questionnaire • Cannot be used with illiterates • there is often a low rate of response • questions
may be misunderstood 2.5.5 Focus Group Discussion (FGD) One or two researchers and several
participants meet as a group to discuss a given research topic. This discussion is considered as
focused because it is guided by a set of specific topics (questions). Participants need to be
generally homogenous on particular characteristics of relevance to the topic. FGD encourages
group interaction where participants can influence and be influenced by others. FGD can be used
to complement other methods. It is important to develop culturally relevant questionnaire. The
number of participants in FGD should be small enough for everyone to have a chance to talk and
large enough to provide diversity of opinions (Usually 8-12 participants in one group) FGD is
preferable to that of in-depth interview when: 59 • Group interaction will help address the
research question. It bring out diverse points of view • Breadth of data is more important than
depth • Topic is not sensitive Advantages of FGD • Do not discriminate against people who
cannot read and write • Encourages participants reluctant to be interviewed • Participant
interaction helps weed out false and extreme views • Help researcher to know expressions and
slang Disadvantages of FGD • The results are more subject to interpretive bias and error •
Analysis process is more time-consuming • Investigator risks getting too much unnecessary
information 60 2.6 Questionnaire design A questionnaire is a research instrument consisting of a
series of questions for the purpose of gathering information from respondents. Researchers need
to consider two important procedures at the outset of constructing a questionnaire. These are
planning and piloting. Planning During planning, it is important to: list the topics of interest in
relation to the aims of the study collect appropriate and tested questions and scales list
additional items and response formats that need to be developed Piloting The developed
questionnaire should be pretested on a sample of about 30-50 people depending on the
complexity of the items. The following issues need to be addressed in the pilot study/pretest • Is
each question measuring what it is intending to measure? • Is the wording understood by all
respondents • Are the instructions on self-administered questionnaires understood by all
respondents? • For closed questions, is an appropriate response available for each respondent? •
Are any questions systematically or frequently missed, or do some questions regularly elicit un-
interpretable answers? • Do the responses suggest that the researcher has included all the relevant
issues in the questionnaire? Questionnaire layout The questionnaire should be visually easy to
read and comprehend. The first few lines of a questionnaire usually include: • the label
‘confidential’, • the respondent’s serial (identification) number (to preserve anonymity), 61 • the
title of the study • a brief introduction Lower case letters should be used for texts. The
instructions for the respondent or interviewer should also be given clearly at the beginning for
example, whether answers are to be ticked, circled, written in or combinations. A thank you
statement should be given at the end of the questionnaire. Any filter questions that do not apply
to some respondents must be clearly labeled and all interviewers and respondents must
understand which question to go to next. Instructions about filter questions/skips are usually
printed for interviewers in the right-hand margin of the questionnaire. The following are
examples of filter question and skip 1a. Have you ever chewed chat? Yes__ No__ GO TO
QUESTION 2 1b. are you currently chewing chat? Yes__ No__ Question numbering and topic
ordering Questions must be numbered (1,2 etc), and sub-questions clearly labeled (e.g 1a, 1b
etc). A question and its responses should never be split over two pages, as this can lead to
confusion. The order of questions is very important that questions should not skip backward and
forward between topics. Each section of the questionnaire should form a module and be topic
based (e.g questions should be grouped together by subject). Provide linking sentences when
moving to new section/modules on the questionnaire for example, the next questions ask about
some personal behaviors Question forms Question form refers to format of the question (closed
or open-ended), and type of measuring instrument (e.g single items, batteries of single items or
scales). The comprehensiveness of response choices for closed questions is important. Closed
questions can be dichotomized (e.g yes/no response choices), multiple response (no restrictions
on the number of responses that can be ticked) or scaled (with one response code per response
frame permitted). 62 Structured interviews involve the use of fixed questions, batteries of
questions and/or scales which are presented to respondents in the same way, with no variation in
question wording and with closed qes (pre-coded response choices). Some structured
questionnaires will also include open-ended questions, to enable respondents to reply in their
own words. The more structured approach is only suitable for topics where sufficient knowledge
exists for largely pre-coded response formats to be developed Semi-structured interviews include
fixed questions but with no, or few, response codes, and are used flexibly, often in no fixed
order, to enable respondents to raise other relevant issues not covered by the interview schedule
Unstructured interviews are comprised of a checklist of topics, rather than fixed questions, and
there are no pre-codes Open questions Open-ended questions are essential where replies are
unknown, too complex or too numerous to pre-code. They are also recommended for developing
questions and measurement scales. The information collected is only limited by the respondent’s
willingness to provide it. It can be very informative as a method, but demanding for the
respondents. Most interview questionnaires include a combination of open & closed questions.
Open questions following closed questions are useful for probing for clarification of reasons and
explanations. Closed questions following open questions are of value on topics about which little
is known and where people are likely to be uncritical or influenced by social desirability bias if
presented too soon with response choices Disadvantages of open questions • replies can be
distorted by the coding process back in the office • They can be time consuming and difficult to
analyze • require more skilled interviewers and coders Closed questions Closed questions are
preferable for topics about which much is known. They are quicker and cheaper to analyze. Self
administered (e-g postal) questionnaire should be restricted to closed question because most
respondents will not bother to write their replies to open-ended questions. Pre-coded responses
always carry the risk that respondents’ replies are forced into inappropriate categories. Care is
needed when one is choosing the response choices for closed ended questions. There should be a
category to fit every possible response, plus include ‘other (specify)’ category if it is 63 felt that
there may be some unknown responses. Pre-coded numbers such as age groups need to be
mutually exclusive, comprehensive and unambiguous Form and under-reporting Pre-coded
response choices may be helpful if under-reporting is likely to be a problem. The list of
alternatives must be comprehensive to prevent under-reporting. With questions asking about
knowledge, open-ended questions are preferable to the provision of response choices in order to
minimize successful guessing. Postal questions should also be avoided when asking questions
about knowledge as they give respondents the opportunity to consult others, or to look up the
answers Question items, batteries and scales Single item questions Single item measures use a
single question to measure the concept of interest. Single item questions are imperfect indices of
attitudes or behaviors, as responses to one question can only be partly reflective of the area of
interest. Responses can also be affected by other factors, including question wording, social
desirability bias and interviewer bias, all of which can lead to measurement error Batteries
Batteries of questions are a series of single questions, each relating to the same variable of
interest. Each item is analyzed and presented individually, not summed together Scales Scales
involve a series of items about a specific domain that can be summed (sometimes weighted) to
yield a score. A more valid measure than a single item questions or batteries of single items is
obtained, because any individual item error or bias tends to be cancelled out across the items
when averaged or summed. Items on the scale should differ in content i.e they should all express
a different belief about the area of interest, or different aspects of the behavior so that they will
not all be limited by the same types of errors or question bias. Scales also permit more rigorous
statistical analysis. Researchers usually aim to provide single score, partly because they are
easier to analyze and apply. The simplest method of combining scale items is to add the item
response scores. For knowledge questions each correct answer can be given a value of 1 and
each incorrect answer allocated 0, and the items added to form a score. With scaled responses a
numerical value can be attached to each class, such as strongly agree=4, agree=3, disagree=2,
strongly 64 disagree=1. Many investigators allocate a middle scale value to ‘no opinion’ (e.g
‘neither agree nor disagree) responses, as in Likert Scale. In this case, the scale values would be:
strongly agree=5, agree=4, neither agree nor disagree = 3, disagree=2, strongly disagree=1. The
crude addition of scores, which results in all items contributing equally to the multi-item scale
score, makes the assumption that all items are of equal importance. If some items are regarded as
more important than others, they should be weighted accordingly (their scores are multiplied by
X to enable them to count more) Likert scale Likert scales are the most popular for the
measurement of attitude. It contains a series of ‘opinion’ statements about an issue. The person’s
attitude is the extent to which he or she agrees or disagrees with each statement, usually on a
five-point scale. It is conventional for high numbers to signify favorable evaluation. Scoring is
reversed where necessary. The total attitude score is the sum of the responses 65 2.7 Planning
data analysis & choosing statistical techniques 2.7.1 Data processing Data processing refers data
entry into a computer, data checks and correction, variable coding, and data cleaning. The aim of
this process is to produce a relatively clean dataset. Data Entry Data entry concerns the transfer
of data from a questionnaire to a computer file. It is converting data into a form that can be read
and manipulated by computers used in quantitative data analysis. The collected data should be
entered in to a data entry form which is prepared in one of the software. It is possible to enter
data in almost all statistical softwars. However, some of these softwars do not have controlling
mechanism for mistakes during when we create data entry template. Softwars like EPI-Info and
EpiData have the feature of mistake controlling mechanism, whereas SPSS, STATA, R or SAS
are powerful for analysis. Thus, it is advisable to use one of this software for data entry and then
export the data to other software for analysis. Before analysis, data must be checked for errors,
outlying observation, missing observation, and data that needs coding must be coded. If there are
outlying or/and missing observations, handling mechanism should be used. Deleting such
observation before formal statistical test will introduce bias. Data coding In general computers
are at their best with numbers. Some statistical packages cannot analyze alphabetic codes, some
cannot understand open ended responses. We must translate variables through coding. Coding is
assigning a separate (nonoverlapping) numerical code for separate answers and missing values.
For instance, instead of using male and female for the variable gender, it can be indicate by
number 1=Male and 2=Female. Numerical variables can be also transformed in to categorical
variable through coding. When you transform numerical variable to categorical, the choice of the
cutoff value should not be arbitrary. Rather, the cutoff values need reference and scientific
justifications. Data Cleaning No matter how carefully the data have been entered some errors are
inevitable. Data cleaning is need after data entry process is completed. The aim of this process is
to produce a clean set of data for statistical analysis. 66 2.7.2 Plan for analysis Data analysis plan
is critical because it tells the reader what analysis will be conducted to examine each of the
research question. In the data plan, cleaning, transformations, and assumptions of the analyses
should be addressed, in addition to the actual analytic strategy selected. We do data analysis to
get answer for the research question we plan answer. The study may have more than one
research question which can be stated as: - Primary research question - Secondary research
question The appropriate methods of data analysis are determined by your data types and
variables of interest, the actual distribution of the variables, and the number of cases. Different
analyses of the same dataset may reflect or represent different aspects of the underlying data
structure. Plan of Data analysis could contain any combination of the following types of data
analysis strategies: • Exploratory data analysis: This type of data analysis often occurs when a
program is new, and it is unclear what to expect from the data. • Descriptive data analysis: The
most common type of data analysis, this approach will summarize your findings and describe the
sample. • Inferential: Inferential statistics allow us to draw conclusions about the larger
population from which the sample is drawn. These powerful techniques may be able demonstrate
if a change has occurred as a result of your program. The purpose of analyzing data is to obtain
usable and useful information. The analysis, regardless of whether the data is qualitative or
quantitative, may: • Describe and summarize the data. • Compare variables • Identify
relationships between variables. • Forecast outcomes 2.7.2.1 Descriptive statistics When data
have been clean and checked, the next explanatory step is to examine the distribution of each
variable. The distribution of each variable should be explored where distribution refers to the
way data are distributed, in order to draw conclusions about a set of data. With continuous
variables, the aim is to determine whether or not normality may be assumed. If not we may wish
to consider transforming the variable, or may wish to categorize the variable for analysis (e.g.
age groups). With categorical variables, we obtain the frequency distribution of each variable.
Based on the distribution, we may wish to amalgamate certain categories for analysis. The
probability distribution of a categorical variable tells us with what probability the variable will
take on the different possible values. 67 The characteristics of categorical variables can be
described by charts or frequency table. There are variables which need to be described by charts
and there are variables which are appropriately described using table. Table 1: Nutritional status
of Adult people living with HIV/AIDS on ART at Achefer Woreda,June 2015,(N=315). Variable
Frequency % BMI(kg/m2) =50 10 19 1.84(0.84, 4.024) 2.326(1.016,5.32 4) INH Prophylaxis
Yes 3 30 2.27(0.712,7.22) * No 64 259 1 WHO clinical T staging at switch I/II 20 107 1 III 29
143 0.845(0.47,1.49) 0.96(0 .52, 1.745) IV 18 39 1.59(0.83, 3.04) 2.58(1.297, 5.14) CD4 cell
count =100cells/mm3 20 114 1 NRTI at first line ART start D4T 29 104 1 AZT 15 137
0.86(0 .5,1.47) * TDF 13 48 1.8(0.916, 3.56) * 78 Note that: The covariate age, WHO stage, and
CD4 cell counts were significant predictors of treatment failure. 2.3 Data quality assurance The
term quality refers to being fit for the purpose. Data quality assurance includes all those planned
and systematic actions necessary to provide adequate confidence that a data product will satisfy a
given set of quality requirements. Many aspects of data collection can impact the quality of the
data, including completeness and clarity of questionnaires, the interviewer's delivery, the
accuracy of mechanical instruments, and technicians' measurement techniques. Other errors can
be introduced into the data after the data are collected, during transcription, at data entry, and
data manipulation for analysis. Minimizing all of these potential sources of error is very
important the planning and implementation of any study. Quality assurance Quality assurance
consists of those activities undertaken prior to data collection to ensure that the data are of the
highest possible quality at the time they are collected. These activities include the development
of the study protocol, development of the data entry and data management systems, training of
data collection personnel, and testing of data collection procedures. The major component of
quality assurance is the development of the study protocol and the creation of manuals
documenting the protocol. Careful design of data collection instruments, including content,
format, and step-by-step instructions for completion of the instrument are necessary. Procedures
should be designed to ensure that the data being produced are reliable and valid. Limitations of
various types of instruments should be considered during development so that potential problems
can be identified and mitigated. All aspects of the protocol must be documented in a manual of
operations. This document should be viewed as the official study reference document for data
collection staff and should contain all details of data collection procedures. Collection of
accurate data is only the first step. It is equally important that errors are not introduced in the
process of converting the data to electronic format. For some forms it may be desirable to have
the form reviewed and edited before data entry. The purpose of this review would be to ascertain
that the form was complete, that skip patterns were followed, and that any data values that
seemed inconsistent or looked like possible errors were checked before being entered. Double-
data entry is one strategy to catch and correct any data entry errors from the original entry. 79
Once all protocols have been developed and documented, the next step is to train study personnel
to implement the procedures. Training result in standardization and can lead to reduced costs
over time. The final step in the quality assurance process is to test all procedures that have been
developed. This is often done by means of a pilot study in which the entire protocol is performed
on volunteers who are similar demographically to the anticipated study subjects. The pilot study
should encompass all aspects of the protocol Quality control Once data collection begins, the
quality control procedures developed as part of the quality assurance process must be
implemented. The goal of these quality control procedures is to identify and correct errors in the
data both during and after data collection. The quality assurance procedures developed during the
planning phase for monitoring performance must be implemented. Frequently, studies establish
procedures for identifying problems but fail to establish adequate procedures for ensuring that
the problems are resolved. Items to monitor may include the following: 1) Recruitment.
Coordinating person or group shall check whether enrolled participants meet all eligibility
criteria (e.g., gender, education status etc). 2) Performance of data collectors. Data should be
analyzed regularly to assess the performance of data collectors. Evidence that a specific data
collector’s data deviate significantly from the group as a whole should be reported and
investigated by the quality control coordinator/supervisor. 3) Data clean-up. Data should be
analyzed routinely for problems such as extreme or inconsistent values. In longitudinal studies,
for example, if a participant's weight is entered as 60 kg one year and 98 kg the next, it is likely
that one of these values is an error which should be checked. 4) Replicate measures. Multiple
measures taken at one point in time can be used both to identify possible data errors and to
calculate more accurate measures. For example, multiple blood pressures are often taken during
one study visit and averaged in order to give a more accurate value of the participant's blood
pressure. Combining multiple measurements by averaging serves to provide more accurate
estimates of true data values and to minimize measurement error. 80 Session summary Research
methods section of a protocol is the section which shall describe how researchers collect,
organize and analyze data that are relevant to the purpose of the study. When writing the
methods section, selection of the study design appropriate for the study objectives is usually the
initial task. Selection of a specific study design depends on the research problem, personal
experience, and resources available for the study. Research problem is the major criteria that
should be considered during selection of study designs. Quantitative study designs would be
appropriate when the interest is quantifying the occurrence of a problem and assessing
associations between dependent and independent variables. On the other hand, qualitative
designs would be appropriate when the topic is new or has never been addressed with the
particular group under study. A mixed methods design is useful when either the quantitative or
qualitative approach by itself is inadequate to best understand a research problem. Researcher’s
personal training and experience also influence the choice of design. An individual trained in
technical, scientific writing, statistics, and computer statistical programs and familiar with
quantitative journals would most likely choose the quantitative design. On the other hand,
individuals who enjoy writing in a literary way or conducting personal interviews, or making
upclose observations may prefer qualitative approach. Investigators may not be able to select the
design that gives a high level of evidence. The design that gives a high level of evidence may not
be feasible to do or ethical to do. The population to be included in the study should be carefully
selected so that as far as possible it should be able to generalize the study findings to a bigger
population i.e target (source) population. Selection of adequate sample by probability sampling
techniques would enable to generalize the findings to the source population. Sample size should
be large enough to address the different objectives of the study. When one tries to assess
associations between variables, there is a need to identify dependent and independent variables.
For some variables which can’t be easily measured, for example by asking a single question,
operational definition is necessary. You need also to identify a data collection technique which is
appropriate for your study purpose and provide more valid data. The commonly used data
collection techniques include document review, observation, interviewing, self administered
questionnaire, and focus group discussion. The most commonly used data collection tool is
questionnaire. Questionnaires need to be carefully prepared and pre-tested/piloted. A variable
can be measured by using single item question, batteries of questions, or scales. The plan for data
management shall reflect issues like which software packages will be used for data entry and
analysis, what will be analyzed, which statistical tests will be used, which measures of
association will be computed, what methods will be applied to control confounding etc. You
should also describe how you will be assuring the quality of data. The activities applied for
quality assurance include careful development of the study protocol, development of the data
entry and data 81 management systems, training of data collection personnel, and testing of data
collection procedures/tools. 82 Learning activities for session 2, Part 1 Read the following thesis
proposal carefully and answer the questions that follow. Some parts of the proposal are
deliberately omitted for the sake of the exercise Research proposal title: Determinants of Non-
adherence to Antiretroviral Therapy Among HIV- Infected Adults In Aksum Town, Northern
Ethiopia: Unmatched Case – Control Study Objectives: To identify determinants of non-
adherence to Antiretroviral Therapy among HIV infected adults on ART. METHODS AND
MATERIALS Study design: Institution based unmatched case control study design will be
carried out. Source Population: Source population will be HIV/AIDS positive adults who are on
ART in Aksum town. Study Population: The study population will be HIV/AIDS positive adults
who are on ART in Aksum town during the study period. Sample size calculation: The sample
size will be computed by Kelsey formula in Epi Info version7 StatCalc for un-matched case
control study. 5% type I error, 80% power and 1:2 ratio of non-adherent cases to adherent
controls will be used. From previous studies, percent of controls exposed, Odds Ratio and
percent of cases exposed to determinant of non-adherence to ART will be employed to carry out
sample size based on table-1 below.