Research methods notes, Lecture notes of Research Methods in Psychology

Download Research methods notes and more Lecture notes Research Methods in Psychology in PDF only on Docsity! RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 1 INTRODUCTION Research is one of many different ways of knowing or understanding. It is different from other ways of knowing, such as insight, divine inspiration, and acceptance of authoritative dictates, in that it is a process of systematic inquiry that is designed to collect, analyse, interpret, and use data. Research is conducted for a variety of reasons, including to understand, describe, predict, or control an educational or psychological phenomenon or to empower individuals in such contexts. What is research? Depending on who you ask, you will likely get very different answers to this seemingly innocuous question. Some people will say that they routinely research different online websites to find the best place to buy goods or services they want. Television news channels supposedly conduct research in the form of viewer polls on topics of public interest such as forthcoming elections or government-funded projects. Undergraduate students research the Internet to find the information they need to complete assigned projects or term papers. Graduate students working on research projects for a professor may see research as collecting or analysing data related to their project. Businesses and consultants research different potential solutions to remedy organizational problems such as a supply chain bottleneck or to identify customer purchase patterns. However, none of the above can be considered “scientific research” unless: (1) it contributes to a body of science, and (2) it follows the scientific method. METHODS OF ACQUIRING KNOWLEDGE There are many methods of acquiring knowledge. These methods can be grouped into two main types: a) Everyday/Non-Scientific ways of knowing b) Reasoning and c) Scientific method. a) Everyday/Non-Scientific ways of knowing These methods are based on faith, accepting things at face value. It is relying on knowledge that has not been tested. These methods include the following: i) Authority One of the most common sources of knowledge is the authorities in different spheres of knowledge. In many societies people rely on the wisdom of elders who are recognized as having better understanding of the world than ordinary members of society. Thus, statements and pronouncements by experts in various areas of knowledge are seldom challenged or questioned. Examples of such people are elderly people in rural areas, heads of religious organizations and dictators. A major weakness of this method is that authorities in most cases tend to make false statements in order to justify and preserve their status. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 2 ii) Method of tenacity or custom/tradition Many people tend to believe things because people in their society regard them as the truth, even when there are clearly conflicting facts. They even go to the extent of inferring “new” knowledge from beliefs that may be false. They know to be true because they hold firmly to it, because they have always known it to be true.” As a result, some people hold to certain things as true because most people in society assume those things to be true. iii) Mystical method In this method the correctness of the knowledge is assumed to reside in the supernatural source. The knowledge producers such as traditional medicine men and diviners are an authority because they claim that they are able to receive and decipher messages from ancestral spirits. To convince people that they actually communicate with the spirits of dead ancestors, they use rituals, ceremonies and unusual language. Like authority, this approach is based on faith. iv) Method of intuition Intuition is the perception, or explanation or insight into phenomena by instinct. In other words, it is the ability to gain new knowledge without conscious reasoning or rational process. v) Experience Human beings learn new knowledge through their experiences in life. People tend to believe that what is in their mind and that a social encounter they have had is generally true. For example, a person who has been swindled by a policeman believes that most policemen are dishonest. b) Reasoning (Rationality) Reasoning is the second category of methods used by human beings to understand their environment. Reason is a source of knowledge from which human beings derive universally valid judgments that are consistent with one another. By reasoning is usually meant the ability to expound one’s thoughts logically and to make conclusions. Rationalists believe that knowledge is innate in human beings and pure reason is sufficient to produce verifiable knowledge. Though they acknowledge the contribution of senses to knowledge in the form of bare facts and isolated impressions, they believe that the intellect interprets and organises these bits and pieces of information into what we can call reliable and significant knowledge. There are two types of reasoning: deductive reasoning and inductive reasoning. Deduction is a form of reasoning that moves from a general statement to a particular instance. It draws implications from one or a number of basic axioms or assumptions. The syllogism is an example of deductive reasoning. The major premise of a syllogism takes the form of a general statement, such as “All men are mortal.” This is a universal statement, which RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 5 researcher is one who can traverse the entire research cycle and can handle both inductive and deductive research. It is important to understand that theory-building (inductive research) and theory-testing (deductive research) are both critical for the advancement of science. Elegant theories are not valuable if they do not match with reality. Likewise, mountains of data are also useless until they can contribute to the construction to meaningful theories. Though both inductive and deductive research are important for the advancement of science, it appears that inductive (theory-building) research is more valuable when there are few prior theories or explanations, while deductive (theory- testing) research is more productive when there are many competing theories of the same phenomenon and researchers are interested in knowing which theory works best and under what circumstances. Theory building and theory testing are particularly difficult in the social sciences, given the imprecise nature of the theoretical concepts, inadequate tools to measure them, and the presence of many unaccounted factors that can also influence the phenomenon of interest. It is also very difficult to refute theories that do not work. Unlike theories in the natural sciences, social science theories are rarely perfect, which provides numerous opportunities for researchers to improve those theories or build their own alternative theories. Conducting scientific research, therefore, requires two sets of skills – theoretical and methodological – needed to operate in the theoretical and empirical levels respectively. Methodological skills ("know-how") are relatively standard, invariant across disciplines, and easily acquired through doctoral programs. However, theoretical skills ("know-what") is considerably harder to master, requires years of observation and reflection, and are tacit skills that cannot be “taught” but rather learned though experience. Properties of Scientific Research The following are the main properties of scientific research: 1) Scientific research is empirical. Only knowledge gained through experience or the senses – touch, sight, hearing, smell or taste is acceptable. The empirically oriented social scientist goes into the social world and makes observations about how people live and behave. However, Nachmias (1992) cautioned against interpreting empiricism in the narrow definition of the five senses – touch, smell, hearing, listening and seeing. (p. 8). 2) It is systematic and logical. Observations are done systematically one at a time, starting with description, explanation and finally prediction. In addition, the correct order must be followed. 3) It is replicable. Since the observation is objective anyone carrying out a study in the same circumstances should come up with the same findings. 4) Research is self-correcting. It has in-built mechanisms to protect investigators from error as far as is humanly possible. In addition research procedures and results are open to public scrutiny by other researchers. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 6 5) Scientific research is question-oriented. It is directed by a research question or problem and several specific questions. These questions might spring from observation of natural or social phenomena, a practical concern or gaps in what is reported in previous research studies and other scholarly literature. 6) Scientific research is public. Because findings from scientific research may be used to make decisions that affect people and society at large, scientific research must be open to public scrutiny and examination and criticism by other scholars. 7) Scientific research is cyclical. It proceeds in stages starting with the research problem, followed by research design, measurement design, data collection, data analysis and generalizations or tentative answers and starts all over again by asking new questions for further research. 8) Scientific research is self-critical. It critically examines its strengths, limitations and weaknesses, and discovers and reports its validity and reliability. 9) Researchers strive to overcome their personal biases as much as possible. They do this by clearly defining the phenomena being studied and using research procedures to study those phenomena that other scholars will agree are accurate. 10) Scientific research is objective. Empirical evidence is assumed to exist outside of scientists themselves. However, in the usual sense of term (to mean observation that is free from emotion, conjecture, or personal bias), objectivity is rarely, if ever, possible (Singleton and Straits, 1999, p. 30) 11) Quantitative researchers strive to make their findings generalizable to the target population. Generalizability is achieved through the selection of representative samples. 12) One major goal of conducting scientific research is to accumulate evidence over time that can be used to validate or disapprove commonly held notions about social reality. For these reasons, researchers attempt to design their studies in such a way that other researchers can replicate their research findings. Unlike everyday methods of knowing, the entire process of inquiry can be reproduced by other researchers. Purpose of Social Science Research 1. The main purpose of research is to discover new knowledge by the discovery of new facts, their correct interpretation and practical application. 2. Research aims at describing a phenomenon. Accurate description forms the basis for other purposes of research. 3. Enable prediction. Prediction is the ability to estimate phenomenon A, given phenomenon B. 4. Control. Control is concerned with the ability to regulate the phenomenon under study. 5. Explanation of phenomenon. Explanation involves accurate observation and measurement of a given phenomenon. In order to explain a phenomenon, one RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 7 should be able to describe it, predict its occurrence and observe factors that cause its occurrence with certainty and accuracy. 6. Enable theory development. This involves formulating concepts, laws and generalizations about a given phenomenon. Research is also conducted in order to confirm or validate existing theories. 1.1 NATURE OF RESEARCH Research involves the systematic collection and analysis of data about learning and teaching for the purpose of description, explanation and prediction of the object or behaviour of the study. Role of research 1. Research gives an accurate account of the characteristics of particular programmes in education. 2. It identifies relevant and appropriate messages and contents for different target groups. 3. Identifies constraints to the implementation of an educational programme. 4. Can be used to determine actions and innovations that have an impact on the target groups. 5. Provides useful data for programme planning. 6. Gives new knowledge on how specific educational programmes are working. Classification of Research Research can be classified according to its purpose or method used/procedures employed. Classification of research by purpose Under this classification, research may be basic research or applied research. i) Basic research It is directed towards increase in knowledge. When successful, basic research results in a fuller understanding of the subject matter under study and the generation of theories. The primary aim of the investigator is not to produce data for practical use, but to enhance understanding of the subject matter under study. ii) Applied research Applied research is directed towards practical applications of knowledge and when successful results in directives for development of blueprints. It is conducted for the purpose of applying or testing theory and evaluating its usefulness in solving problems i.e. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 10 This is a combination of quantitative and qualitative research paradigms. A researcher can adopt mixed method research because of the following reasons: 1. Researcher’s competency in both qualitative and quantitative research 2. The nature of the research problem 3. Time available 4. When dealing with sensitive issues that numbers cannot explain 5. When in-depth data is required N/B There is no research that is purely quantitative or qualitative. Most researches adopt the mixed method research paradigm. Basic elements in Research 1. Population Population refers to an entire group of individuals, events or objects having common observable characteristic e.g. all form four students in the country, all teachers in public primary schools in a district etc. 2. Sample A sample is a relatively small group of individuals/objects chosen in a scientific fashion to represent a relatively large group of individuals (population) which the researcher is interested in studying e.g. eighty public primary teachers in a particular district. 3. Sampling Sampling is the process of selecting a number of individuals for a study in such a way that the individuals selected represent the large group from which they were selected. 4. A constant A constant is a characteristic of objects, people or events that does not vary e.g. temperature at which water boils is a constant. 5. Variable A variable is a characteristic of objects, people or events that can take on different values. It can vary in quantity such as height, weight or in quantity such as skin texture, gender, occupation, type of school, marital status. Types of variables i) Independent and dependent variables An independent variable is one which the researcher manipulates (controls) in order to determine its effect or influence on another variable. Also called predictor variable. A RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 11 dependent variable is one whose occurrence is influenced by the independent variable. E.g. the influence on hours studied (Independent variable) on performance in a statistic test (dependent variable). ii) Discrete and continuous variables A discrete variable takes finite/exact/whole values e.g. number of children, number of classrooms. Continuous variables take any value within an infinite series of possible values e.g. weight, height etc. iii) Extraneous and control variables These are variables which influence the results of a study when they are not controlled for. However, if the extraneous variable is included in the study, it is referred to as a control variable. The introduction of a control variable in a research study increases the validity of the data and therefore it leads to more convincing generalizations. 6. Concept A concept is an interpretation/idea/image of a particular phenomenon e.g. professional qualification, performance, intelligence, gender, religious sister etc. 7. Conceptual definition This is defining a concept using other concepts. This approach is not good in research. 8. Operational definition This is defining a variable as used in context. One defines the variable according to how it will be measured. It is the description of the operation that will be used in measuring the variable. 9. Research problem A problem is a question of interest which can be answered through the collection of data. It is an intellectual challenge that requires answers through collection and analysis of data. A problem is not a difficulty being experienced. 10. Hypothesis This is a tentative answer to a research problem. In quantitative research, a hypothesis is expressed in terms of a relationship between two or more variables. 11. Theory A theory is a set of concepts and the interrelations that are assumed to exist among those concepts. Hypotheses are developed from theories. Stages in the Research Process The research process is cyclic and involves the following stages: RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 12 1. Selection and definition of the research problem a) Selecting the topic for research and identification of the research problem b) Reviewing related literature c) Formulating the research problem d) Formulating research questions, objectives and hypotheses e) Defining concepts 2. Selection of the research design and methodology a) Choice of research design b) Description of the sample and sampling procedures 3. Measurement design a) Identification of data collection procedures b) Construction of research instruments 4. Data collection Collection of data using a variety of research instruments 5. Data analysis a) Data organization b) Data processing c) Quantitative and qualitative data analysis 6. Generalization a) Interpretation of data b) Conclusion and recommendations c) Production of a research report RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 15 Multilingualism as a factor as a means towards tolerant citizenship. Stating Research Questions and Hypotheses A research problem is usually broken down into a set of concrete research questions, objectives and hypotheses that will be investigated separately. It should be noted that one cannot have both research questions and objectives in the study; use only one of them. Research Questions Research questions refer to questions which a researcher would like answered by undertaking the study. The difference between research questions and objectives is that a research question is stated in a question form while an objective is a statement. Example: EFFECTIVENESS OF THE SECONDARY SCHOOL HIV-AIDS CURRICULUM IN KENYA Main question What is the effectiveness of the secondary school HIV-AIDS curriculum in Kenya? The following research questions were raised to provide answers to the research problem.  What are the objectives and content of the HIV-AIDS curriculum?  What are the attitudes of teachers and learners regarding HIV-AIDS curriculum?  What new knowledge have the learners acquired from the HIV-AIDS curriculum?  To what extent are the teaching and learning resources to implement the HIV- AIDS curriculum sufficient?  Is there a relationship between the teaching and learning of HIV-AIDS and the management of the pandemic?  What can be done to make the curriculum more efficient in dealing with the pandemic? Research questions are the sub-research problems and are arrived at by fragmenting the big questions into small parts each treating a specific aspect of the main problem. The questions may be formulated using the words: • What……………….. • How………………… • To what extent…….. • Is there…………….. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 16 Hypotheses A hypothesis states possible relationships or differences between two or more variables or concepts. They are derived from existing theories, previous research, personal observations or experiences. A study can have one or more hypotheses depending on the variables in the study. Types of hypotheses Hypotheses can be null hypotheses or alternative hypotheses. i) Null hypothesis/statistical hypothesis A null hypothesis states that there is no relationship or difference between variables or concepts. It is symbolized H0 e.g. H0: There is no significant relationship between students’ performance in Mathematics and Physics. ii) Alternative hypothesis An alternative hypothesis states that there is a relationship or difference between the concepts or variables. It is symbolized by H1 e.g. H1: There is a significant relationship between students’ performance in Mathematics and Physics. An alternative hypothesis can be directional or non-directional. Alternative directional hypothesis It specifies the nature of the relationship or difference between variables e.g. H2: There is a positive relationship between remedial tuition and pupils’ achievement in Mathematics Alternative non-directional hypothesis It is also referred to as research hypothesis. It simply states that there is a relationship or difference between variables but does not show the direction in which the variables are related or differ e.g. H3: There is a difference in the performance of national examinations between children from rural primary schools and children from urban primary schools. Qualities of a good hypothesis RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 17 1) Must state clearly and briefly the expected relationship or difference between variables 2) Must be based on sound rationale derived from theory, previous research or professional experience 3) Must be consistent with common sense or generally accepted truth 4) Must be testable- data can be collected to support or fail to support hypothesis 5) Must be testable within a reasonable time 6) Must be related to empirical phenomena 7) Must be simple and concise 8) Must be consistent with the purpose statement and objectives Conceptual Framework/Theoretical Framework Conceptual Framework It is a model that shows the interrelatedness of the independent and dependent variables of the study. This is either presented graphically or diagrammatically. The purpose of conceptual framework is to enable the reader to quickly see the proposed relationships between the variables. The diagram should be followed by a narration explaining it for clarity. Figure 2 BASIC EDUCATION Primary Education Secondary Education POLICIES TO BE FORMULATED Secondary Education made Compulsory Distribution of Teaching- Learning resources in schools Adequate physical facilities in schools EXPECTED RESULT Transition rates from Primary to Secondary school increased EFFECT All Primary school graduates access Secondary Education WHAT SHOULD BE DONE Build more Secondary Schools and expand the existing ones RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 20 EXPERIMENTAL AND QUASI-EXPERIMENTAL DESIGNS The effects of a programme on the participants can be measured accurately only if the researcher knows what would have happened without the programme. Because it is not possible for the researcher to observe what would have happened to the participants themselves if they had not enrolled in the programme, a group of non-participants in the same population must be identified. The effects of the programme are then ascertained by comparing the behaviour of the group that is particularly in the programme (the treatment of the experimental group) with that of the selected group (the control group). The allocation of individuals to the experimental and control groups should be at “random”. Incorporation of randomization into the design is so important that designs are categorized according to the way in which the subjects are selected: random assignment (classified as an experimental design) and non-random assignment (classified as a quasi-experimental design). EXPERIMENTAL DESIGNS Experimental designs are used to study cause and effect relationships among two or more variables. The main difference between a true experimental design and other designs is the fact that research units are assigned to the treatment and control conditions at random. The main purpose of experimental research is to study causal links; to determine whether a given variable x has an effect on another variable y, or whether changes in one variable produce changes in another variable. The three essential elements of an experimental design are: 1. Randomization – The researcher assigns participants to different groups. 2. Manipulation – The researcher does something at least to some of the participants in the research. 3. Control – the researcher introduces one or more controls over the experimental situation. In experimental studies one group, the experimental or treatment group, receive the treatment the other group; a control group receives a neutral treatment. The two groups are compared before the treatment using a pre-test. After the treatment, a post-test is administered to the two groups. This situation is presented diagrammatically below. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 21 Figure 4 Pre-test Post-test Post-testPre-test Randomly selected Experimental Group Randomly selected Control Group Before After Treatment Applied According to Frey, Botan, Friedman & Kreps (1991, p.156) three requirements are necessary for establishing a causal relationship between an independent and dependent variable. All three requirements are necessary for inferring causality; none is sufficient in and of itself. These are: 1. The independent variable must precede the dependent variable. 2. The independent and dependent variable must be shown to co-vary, or go together. 3. The changes observed in the dependent variable must be the results of changes in the independent variable and not some other unknown variable. The true experimental designs are the most exact method of establishing cause and effect relationships. They are called true experimental designs because they provide adequate controls for all sources of internal invalidity. The experimental method was developed in the natural sciences where consistent causal relationships are easy to establish. However, the accumulated experience shows that in education and the social sciences, causal relationships are difficult to measure and true experimental designs are rarely used. True experimental designs The true experimental design is considered the most useful design to demonstrate programme impact if conditions of randomization in selection of participating units and in the assignment of treatment and control conditions at random are met. Research units can be individuals (students, teachers, parents etc.), groups of individuals, institutions, regions etc. There are three true experimental designs: i) the pre-test-post-test control group design ii) the post-test-only control group design; and iii) the Solomon four-group design The Pre-Test-Post-Test Control Group Design The subjects are randomly divided into two groups. One of these groups receives the treatment. The other does not. In this design a pre-test is administered to the experimental group before it receives the treatment. The same pre-test is administered to RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 22 the control group. At the end of the treatment, a post-test is administered to both groups. The design is shown below: Figure 5: The pre-test post-test control group design P re -tes t P re -tes t P ost -te s t P ost -te s tT rea tm en t B e fo re A fte r E xp er im en ta l G ro u p C on tro l G ro u p For example, if a researcher wants to determine the effectiveness of a technique of teaching Christian religious education to secondary school students (7th Grade) he should randomly assign students to two groups; an experimental group and a control group. Both groups should be pre-tested, the new technique used to teach the experimental group for say a month, and both groups post-tested. Both the pre-tested means and post-test means can then be compared by using the T-test to determine if there is any systematically significant different between them. The weaknesses of this design are: • Influence of exposure to books, radio and television programmes; • In the case of a long period of time between pre and post-test, the students could have naturally matured; and • The use of the same test can influence the results. You should counteract those influences by keeping the time between tests short, monitoring outside influences but withholding scores until after the post-test. The Post-Test-Only Control Group Design There are circumstances under which a pre-measurement period is not practical. The post- test-only control group design omits the pre-tested groups altogether. The design is shown below: RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 25 01 x 02 03 04 PROCEDURE 1. Assign subjects to the treatment and control groups. 2. Administer a pre-test to subjects in the two groups. 3. Apply the intervention (treatment) to the experimental group. 4. Administer post-test to the two groups. STRENGTH It may be possible to use an existing sample group as experiment and control group. LIMITATIONS Since subjects are not assigned to the experimental conditions, the two groups are not equivalent. It is difficult whether or not causes other than the treatment may have been responsible for differences in the changes of the two groups. 2. Time-Series Design Sometimes it is not possible for researchers to provide a control group against which to compare a treatment group. In this design, a series of pre-tests and post-tests are given to establish intra-group comparisons. The interrupted time-series is used to detect changes in levels and rates. 3. Multiple Time-Series Design This design combines the non-equivalent control group with the time-series design. This is done by assigning subjects non-randomly to treatment and control groups and measuring both groups using a series of pre-tests and post-tests. This design may be diagrammed as follows: RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 26 4. Ex-Post Facto Design This design has been assigned the name of static group comparison. In this design data are obtained for a treatment and comparison group after the treatment has occurred. For example, an evaluator might wish to determine the effects of an AIDS Education project by comparing the sexual behaviour of students who have been exposed to an AIDS education curriculum with that of students who have not participated in the project. This design is presented diagrammatically below: There are two kinds of ex-post facto research. Co-relational study and criterion group study or causal-comparative study. In a co-relational study, the researcher attempts to show a relationship between the antecedents of a present condition and the present condition and the present situation. The design may be presented as follows. “In the (causal-comparative) approach the researcher compares the subjects in which the variable is present with similar subjects in whom it is absent. The design may be presented as follows LIMITATIONS The following are the main limitations of ex-post facto research. 1. Subjects cannot be randomly assigned to treatment and control groups. 2. It is impossible to manipulate the independent variable. 3. Causes are often multiple and not single. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 27 One of the most serious dangers of ex-post facto and causal-comparative research is the post hoc fallacy, the conclusion that because two factors go together one must be the cause of the other the effect. MULTIPLE TIME SERIES DESIGN This design combines the non-equivalent control group with the time series design. This is done by assigning subjects non-randomly to treatment and control groups and measuring both groups using a series of pre-tests and post-tests. The distinguishing feature of the quasi-experimental design is that the research units have not been randomly selected and often have not been randomly assigned to treatment conditions. INTERNAL AND EXTERNAL VALIDITY OF RESEARCH DESIGNS Validity refers to the extent to which the results obtained from the analysis of data actually represent the phenomenon under study. There are two types of validity relating to research design: internal and external validity. INTERNAL VALIDITY A research design has internal validity if the outcome of the research is a function of the phenomenon being studied rather than other causes not systematically dealt with in the study. Internal validity asks whether a research study is designed and conducted so that it leads to accurate findings about the phenomenon being investigated. A study possesses internal validity if its findings follow in a direct and unproblematic way from its methods. The results are sustained by the design itself and cannot be explained by other considerations. EXTERNAL VALIDITY External validity refers to the degree to which the results of an experiment may be extended to other samples from the same population and to other populations, (Mason and Bramble, 1997, p. 113). A study has external validity if the cause and effect noted in the study can be generalized beyond the study, individuals, settings and occasions. External validity asks whether the conclusions from a particular study can be applied to other people and other contexts. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 30 representative of the target population. As a result, the results would not be generalizable to the large group. 3. Reactive effects of experimental arrangements. These effects occur when the findings of the study may not be directly applicable to the real life setting. Two reactive effects of experimental arrangement are discussed briefly below. (i) Hawthorne Effect The term Hawthorne effect refers to any situation in which the experimental conditions are such that the mere fact that the subject is participating in an experiment or is receiving special attention tends to improve performance (Borg & Gall, 1983, p.215). For example, in experimental research, in education, when individuals are aware that they are participating in an experiment they may perform better or differently. In this case the behaviour of the experimental groups departs from what is considered normal for the target group. (ii) John Henry Effect The John Henry Effect refers to a situation often found in Research in which a control group performs above its usual average when Social Science placed in competition with an experimental group that is using a new method or procedure that threatens to replace the control procedure. Multiple-Treatment Inference When multiple treatments are applied subsequently, subjects are likely to experience cumulative effects. COPING WITH THREATS TO INTERNAL AND EXTERNAL VALIDITY OF EXPERIMENTAL AND QUASI-EXPERIMENTAL DESIGNS 1. To avoid bias in assignment to the experimental and control groups, eligible subjects should be assigned at random to the two groups. 2. To avoid bias in the assessment of final results; the assessor will be blind to the intervention group. 3. To avoid bias in behaviour of the participant, the intervention group assignment will not be known to the participants, i.e. the participant will also be blind. The survey design is the best design for collecting systematic factual data for decision making and is an efficient method of descriptive information regarding characteristics of the population and the current practices and conditions (Kerliner 1973). RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 31 A study is said to be double blind if neither the assessor nor the participant is aware of the intervention allocations. A study is called single blind if only the assessor or the participant is aware of the allocations. How to deal with threats Internal Validity Threat Solution History Maturation Selection Experimental mortality Hawthorne and John Henry effects • Random assignment • Run random assignment • Run experiment and control at the same time. • Random assignment • Random assignment • Monitor both groups • Set up treatments as courses given for credit. • Make treatment and control conditions as unobtrusive and normal as possible. • Use placebo-disguise or minimize the: i) Newness of the programme under study. ii) Make both groups feel that they have special experimental status. SURVEY RESEARCH DESIGNS The survey method is used to describe people and their beliefs, attitudes and behaviours. Survey researchers often describe characteristics of respondents for the purpose of building theories or generalizations about the population they represent. Respondents are asked questions concerning their beliefs, attitudes and behaviour. The following designs are normally used by survey researchers: • Correlational designs • Cross-sectional survey designs • Longitudinal survey designs RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 32 Survey designs – are used to collect descriptive data regarding the following: • Characteristics of existing populations, e.g. public opinion, attitudes and preferences and interests. • Current practices, conditions or needs. • Performance of specific programmes, products or organizations. A survey research is a study in which data are collected from the members of a sample for the purpose of estimating one or more population parameters (Jaeger, 1988). It deals with the incidence, distributions and interrelations of educational variables. Its purpose is to describe specific characteristics of a large group of persons, objects or institutions. Survey research does not emphasise the diverse aspects of a single case but rather the frequency or number of answers to the questions by different people. The emphasis therefore shifts from answers to all questions given by an individual (case study) to answers to one question given by all respondents (survey). Longitudinal survey is the description of a sample over an extended period of time. It involves collecting data from a sample at different points in time, in order to study changes or continuity in the sample characteristics (Cohen, 2000). They are where phenomena are studied over time either continuously or repeatedly. For example, if you want to establish the behaviour change patterns of learners who are taught the HIV-AIDS curriculum then you conduct a survey at various stages in the learning levels such as Form 1, 2, 3 and 4. Cross sectional survey is a research where data are obtained at one point in time, but from groups of different ages or at different stages of development. For example, if you are interested in studying how students’ attitudes towards mathematics changed from standard seven to form three, you could select a sample of students at each class level and administer a questionnaire to all of them on the same date or within a narrow range of dates. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 35 • Private archival records such as archives, diaries, and letters • Direct response to data using interviews guides and observation guides. ETHNOGRAPHY The word “ethnos” is a Greek word for “a people” or “a cultural group.” The focus of an ethnographic inquiry is therefore the culture (a collection of behaviour patterns and beliefs that constitute standards for deciding what is, what can be, how one feels about it, what to do about it and how to go about doing it) of a group of people. Ethnographers become part of the people they are studying. Thus, the method used by ethnographers is participant observation. This implies that the researcher is immersed in the culture under study. For instance, if you want to study the culture of the Abagusii, Akamba, Agikuyu, Abaluhya, Luo, Nandi etc. people then you have to live with them, participate in their day to day activities. In other words, be a part of them. GROUNDED THEORY This design is used to develop theory that is grounded in data systematically gathered and analyzed. There is normally a continuous interplay between data collection and analysis. Even though Strauss and Corbin who developed the grounded theory came up with three steps to be followed when analyzing data. The researcher should be sensitive to the non- linear nature of qualitative analysis. The researcher actually moves back and forth with the data, analyzing and then collecting more data, and then analyzing some more. The steps rarely occur in a linear fashion but recur as often as is necessary to reach the appropriate conclusions. Step 1: Open - coding It is the part of analysis that pertains specifically to naming and categorizing phenomena through close examination of data. Thus, the researcher must take a part an observation, a sentence, or a paragraph and give each discrete incident idea, or event, a name or label that stands for or represents a phenomenon. Step 2: Axial coding The researcher puts the parts of the data identified and separated in open coding back together to make connections between categories. During this phase, the researcher builds a model of the phenomena that includes the condition under which it occurs (or does not occur), the context, in which it occurs, the action and interactional strategies that describe the phenomena and consequences of these actions. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 36 Step 3: Selective coding This last stage is essential for those researchers who intend to develop a theory from the data. It involves the process of selecting one main category (the story line) and relating the other categories to it. The researcher validates the hypothesized relationship with the data available and fill in categories that need further refinement and development. During this phase of analysis, the first step is to identify the core category or story line, and then relate the subsidiary categories to the core through a model or paradigm which includes an explication of the conditions, strategies, and consequences identified in the axial coding phase. The researcher validates the theory by grounding it in the data; if necessary, the researcher can seek additional data to test the theory (Mertens, 1998), PHENOMENOLOGY Patton (1990) observes that the term phenomenology has been widely used until it’s meaning has been confused. The term has been viewed as a paradigm, a philosophy or a perspective. Others view the term to mean qualitative methods or naturalistic inquiry. A phenomenological inquiry focuses on the question, “What is the structure and essence of experience of this phenomenon for these people?” From its philosophical origins, in the works of Edmund Husserl, phenomenology is the study of how people describe things and experience them through their senses. The main focus here is that “we can only know what we experience” (Patton, 1990:69). Notice that our understanding comes from our sense (touch, smell, taste etc) experience of the phenomenon. However, that experience must be described, explicated and interpreted. For the phenomenologists, there is no separate (or objective) reality for people. What is for them is what people know their experiences are and mean. What does this imply? This implies: • What is important to know is what people experience and how they interpret the world (Subject matter of phenomenological inquiry) • The only way for us to really know what another person experiences is to experience it for ourselves (Methodological) BIOGRAPHY The design is used for the study of an individual. A biography study is the study of an individual and her/his experiences as told to the researcher or found in documents and archival materials. Denzin (1989a) defines the biographical method as the ‘studied use and collection of life documents that describe turning point moments in an individual’s life. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 37 Mertens (1998) quoted Yow, (1994) who said that Biographers should help the reader understand the way the individual sees (or saw) herself or himself “the inner struggles and motivation, the way psychological make up influenced the subject’s interpersonal relationships, the interpretation the subject gave to life’s events”. Biographies generally attempt to unite an individual story with the meaning of the times in which a person lived. Some biographers are based on oral histories in which the researcher gathers personal recollections of events, their causes and their effects from an individual or several individuals. Hence, writing a biography often involves extensive interviewing that could stretch over years. QUALITIES OF GOOD QUALITATIVE RESEARCH Qualitative research has been criticized by quantitative researchers for its lack of validity and reliability. Qualitative researchers such as Lincoln and Guba (1985) and Creswell and Miller (1997) have proposed naturalist’s alternatives that parallel traditional quantitative approaches to validity and reliability. According to Lincoln and Guba (1985, p. 300), the following terms are the naturalist’s equivalents for internal validity, external validity, reliability and objectivity: credibility, transferability, dependability, and confirmability. Each of these terms is described briefly below. Credibility This is the extent to which the findings of inquiry are considered to be believable. Several techniques are used to ensure the credibility of qualitative research. These include prolonged engagement, persistent observation, triangulation, peer debriefing and member checking and audit trial (Lincoln and Guba, 1985). Prolonged engagement and persistent observation involves investing a sufficient amount of time to build trust, with participants, understanding their culture and test for misinformation. Triangulation involves the use of numerous methods of data collection and sources. There are three types of triangulation: methodological triangulation, source triangulation and investigator triangulation. Methodological triangulation is applied when the researcher uses two or more methods of data collection to measure variables. For example, a questionnaire could be supplemented with in-depth interview, observation and existing records. Source triangulation is utilized when comparing the information which is given by the source at different times and in different situations. For example the information given by the headteacher during the informal conversation could be compared with the information he or she gives in the actual interview. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 40 combinations of persons for a particular sample size are equally likely. Note that each person is chosen at random, as compared with systematic sampling where only the first one is (see below). You can't produce a simple random sample without a full list of the population. b) Systematic random sampling This involves taking every nth name from the population list. After deciding on the sample size needed, you divide the total number of names on the list by this sample size. So, if a sample of 50 is required from a population list of 2,000 then every 40th (i.e. 2000 /50) person is chosen. You then randomly choose a number that is less than n (i.e. less than 40 in the example) to start and take every nth name from the list until you have the required sample. For the sample to be representative, this method relies on the list being organized in a way unrelated to the subject of the survey. Although this may seem to be a simple and straightforward way of drawing a probability sample, it has certain statistical peculiarities. Whereas the initial chance of selection of any person is the same, once the first person has been chosen, most persons will have no chance of inclusion and a few will be automatically selected. Similarly, most combinations of persons are excluded from the possible samples that might be chosen. This might be important if the ordering in the list is organized in some way (possibly unknown to you). Working in schools, where class lists are arranged alphabetically, you could find that a sample is almost exclusively made up of students sharing a relatively common surname. Both random and systematic sampling require a full list of the population. Getting this list is often difficult. Hence, if there is any possibility of ordering in the list messing up your systematic sample, you may as well go for a random sample as the extra effort is minimal. c) Stratified random sampling In this method, the population is first divided into two or more strata (subgroups), and then a given number of subjects (individuals) are randomly selected from each population subgroup. The subgroups can be boys or girls, rural or urban etc. The number selected from each subgroup should be proportional to the number of subjects in the subgroup. d) Multi-stage random sampling This involves random sampling at each stage in the sampling process e.g. supposing that one wants to select secondary schools for study in Kenya. A typical multi-stage sampling would be as follows: i) Randomly select a given number of provinces from the list of all provinces ii) Randomly select from within each chosen province schools from the list of all schools of the defined type iii) Randomly select from within each chosen school individuals from the list of all individuals of the defined type. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 41 e) Cluster sampling This is sampling in which groups, and not individuals, are randomly selected. Examples of clusters include schools, classes, towns etc. For example, instead of randomly selecting Form I students, one could randomly select Form I classrooms and use all students in each classroom. Non-Probability Sampling techniques This method is less strict and makes no claim for representativeness. It is generally left up to the researcher or the interviewer to decide which sample units should be chosen, and is employed in exploratory research, observational research, and qualitative research. In this procedures of sampling not everyone in the population has an equal chance of being selected. These techniques are therefore biased techniques. Types of non-probability sampling techniques a) Purposive sampling/ judgement sampling In this sampling, the researcher purposively chooses the subjects that have the required information with respect to the objectives of his/her study. It is sampling that involves selection of cases that manifest the phenomenon of interest to the researcher. b) Quota sampling This technique is similar to stratified random sampling and the objective is to include various groups or quotas of the population in the study based on some criteria e.g. untrained teachers, graduate teachers, PI teachers. The researcher purposively selects subjects to fit in the quotas identified. Here the strategy is to obtain representatives of the various elements of a population, usually in the relative proportions in which they occur in the population. Hence, if socio- economic status were considered of importance in a particular survey, then the categories 'professional/ managers and employers/intermediate and junior non-manual/skilled manual/semi-skilled manual/unskilled manual' might be used. Interviewers would be given a quota of each category (with examples to assist them in categorization). Within the category, convenience sampling (see below) is normally used. The interviewer will, for example, seek to interview a given number of unskilled manual workers, a given number of semi-skilled manual workers, etc. by, say, stopping passers-by, and will continue until his quota for the day is complete. The common use of the term 'representatives' in quota sampling has to be looked at with some care. They are representative only in number, not in terms of the type of persons actually selected. All such means of gathering quota samples are subject to biases. Careful planning, experience and persistence can go some way to addressing obvious biases. If, for example, home visits are involved, avoiding houses where there is a Rottweiler or other large dog, or sounds of a ghetto-blaster, or there are no curtains, or apartments where the lift is out of order, etc. may be understandable behaviour on the part of the sensitive interviewer, but mitigates against representativeness in RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 42 householders in the sense of all householders having an equal chance of appearing in the sample. Quota sampling is widely used by market researchers and political and other opinion poll-sters. It is in their interest, particularly in situations where the accuracy of their survey findings can be checked (as in election polling) to devote considerable resources to the training of their staff. For non-specialists in the field it is best avoided if possible. c) Dimensional sampling This is an extension of quota sampling. The various dimensions thought to be of importance in a survey (perhaps established by pilot work) are incorporated into the sampling procedure in such a way that at least one representative of every possible combination of these factors or dimensions is included. Thus a study of race relations might identify ethnic group and length of stay in this country as important dimensions. Hence the sampling plan could consist of a table or matrix with 'ethnic group' and 'length of stay' constituting the rows and columns. Refinements of this approach involve selection of particular combinations of the dimensions (e.g. 'Kenyan Asians' with '10-15 years' residence') either because of their particular importance, or because of an inability through lack of time and resources to cover all combinations. The critical comments made about quota sampling apply with equal force to dimensional sampling. d) Convenience sampling/accidental sampling It involves selecting units or subjects that are convenient to the researcher. In most cases, the units selected are those which are easy to reach and are willing to take part in the study. Convenience sampling is sometimes used as a cheap and dirty way of doing a sample survey. You do not know whether or not the findings are representative. It is probably one of the most widely used and least satisfactory methods of sampling. The term 'accidental sample' is sometimes used but is misleading as it carries some suggestion of randomness whereas all kinds of largely unspecifiable biases and influences are likely to influence who gets sampled. Appropriate uses of convenience sampling include gaining a feeling for the issues involved or initial piloting for a proper sample survey. e) Snowball sampling In this method, initial subjects with the desired characteristics are identified using purposeful sampling technique. Here the researcher identifies one or more individuals from the population of interest. After they have been interviewed, they are used as informants to identify other members of the population, who are themselves used as informants, and so on. This is a useful approach when there is difficulty in identifying members of the population, e.g. when this is a clandestine group. It can be seen as a particular type of purposive sample. As well as its value in identifying a sample it has also been used to shed light on social and other networks (Browne, 2005; Farquharson, 2005). RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 45 understand the main research question and sub questions and the study population well enough that they can cue prompts and probe for certain information. Probing In an in-depth interview the interviewer probes the respondent to obtain certain additional information. According to Nachmias and Nachmias (1992 p.230), “probes have two major functions: they motivate the respondent to elaborate the answer or to explain the reasons behind the answer and they help focus the conversation on the specific topic of the interview. An interview should last from 20 minutes to one hour and a half. Start the interview with a non-threatening open-ended question to put the interviewee at ease and get him/her talking. Seidman (1991) has proposed the following guidelines for conducting an interview: • Listen more, talk less; • Follow up on what the participant says. • Ask questions when you do not understand. • Ask to hear more about a subject. • Explore, do not probe. • Avoid leading questions. • Ask open-ended questions which do not presume an answer. • Follow up and do not interrupt. • Keep participants focused and ask for concrete details. • Ask participants to rephrase or reconstruct. • Do not reinforce the participant’s response. • Tolerate silence and allow the interviewee to be thoughtful. Interviewer Manual An interviewer manual for use during the survey should be developed during the preparation of an interview guide. The manual usually consists of the following items:- 1. Detailed instructions regarding selection of the subjects. 2. Special instructions regarding each question in the questionnaire. 3. Guidance on how to deal with unusual situations. 4. Guidance on when and how completed questionnaires should be submitted for data processing or analysis. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 46 Training of Interviewers The training of interviewers is normally carried out by one of the researchers. The interviewers and researcher should go over all questions in the interview guide. This should be followed by practical exercises. The practical activity usually consists of one interviewer administrating the interview guide to another interviewer while the researcher and other interviewers look on followed by an evaluation and discussion of the interview by the whole group. After the initial training, the interviewers should try out interviews with people from the target population in the presence of the researcher. This should be followed by detailed critical evaluation and discussion of each interview or a series of interviews. Procedure for Interviewing 1. Introduce yourself to the respondent(s) avoiding titles unless it is necessary. Tell him/her who you are and where you work. 2. Explain to the respondents the purpose of the study and how it will contribute to the improvement of educational practices in the country. Assure him that the information he provides will be treated confidentially. 3. Be friendly and make the respondent feel comfortable and relaxed in your presence. 4. Ask your questions; probe the respondent to obtain more data. 5. Avoid making negative comments and criticisms. 6. Avoid negative facial expressions like frowning. 7. Avoid interrupting the flow of information when the respondent is talking. 8. Interview the respondents in a place with minimum disruptions from the environment. 9. Ask your questions simply and in a friendly manner. 10. Listen carefully to the answers. 11. Never hint either y specific comments or non-verbal clues to suggest particular responses. 12. In case, you apprehend that the informant is not giving you correct information, cross examine him. If sensitive questions are to be asked, remind the respondent that answers will be held in strict confidence. 13. Write down the answers. Interact with the respondent as an equal. 14. At the end of the interview, thank the informant. Disadvantages a) It is more expensive than questionnaire and tests. b) It is time consuming. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 47 Observation Observation takes place while things are actually happening. In qualitative research, observation takes the form of participant observation. You should make field notes. PARTICIPANT OBSERVATION The researcher’s’ role varies from that of a passive viewer to an active participant. He/she can play the following roles depending on the objectives of the research and the social setting. The researcher can: a) Participate in the work and life of a group. b) Sit unobtrusively taking notes. During observation, the researcher should write down observations in this form of field notes and diagrams. He should also take photographs and pictures and collect documents and artifacts for analysis. The following are potential limitations of observational methods. 1) Observer bias. Observer bias or error occurs due to the background, expectations or perceptions of the observer. 2) Contamination. This occurs when the observer knows the expected outcomes and which groups are experimental or control. 3) Halo effect. This “occurs when the observer allows an initial impression about one aspect of a person or group to influence observations on other aspects” (McMillan, 1992, p.132). DOCUMENT ANALYSIS Analyse materials such as photographs, videos, films, memos, letters, diaries, clinical card records, official documents produced by …. DATA COLLECTION Go to where their subjects are and spend time with them in their territory; their school. These are the places where the subjects do what they normally do. Encourage the subjects to talk. The research should unobtrusively keep a written record of what happens as well as collect other forms of descriptive data. They may participate in their activities. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 50 Questionnaire can also be sent to participants through post office. The participant will fill in and send it back to the searcher. E-mail The researcher can also send the questionnaire using electronic mail. When the respondent finish filling it, he/she will send it back through the same E-mail Address. b) Interview An interview is a conversation in which one person, the interviewer, seeks responses for a particular purpose from another person, the interviewee. The interviewer asks questions intended to lead the respondent towards giving data to meet the study objectives. An interview guide (used in qualitative research) is a general plan that the interviewer follows. It is a list of questions or topics that the interviewer intends to explore during each interview. In quantitative research it is called interview schedule. Just like a questionnaire, an interview can be structured or unstructured. Structured Interview Guide Structured interview guide is an instrument used during data collection in which the researcher reads the items as they are without changing or expounding on the items. A questionnaire can be used as a structured interview guide when the researcher fills the information while the respondent gives the responses. In-depth Interview Guide An in-depth interview guide is a research instrument that is used to collect detail information by way of PROBING further and further until “no stone is left un turn”. It is flexible because the researcher can make clarification, expound on the items, change the items need arise. Administering an Interview Guides Face-to-face interview: This is a method where the interviewer sits with the interviewee and asked questions concerning the issues being studied following the items in the interview guides. Telephone interview: This is a method where the interviewer uses the telephone to interview the respondents without going physically to the field. Structured response formats for questionnaire and interviews The four categories of structured response items are: checklist, inventory ranking and rating/scaling types. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 51 Checklist- A checklist presents a number of options from which the respondent is expected to check or tick the most relevant response or all suitable responses. Inventory- An inventory is used when respondents are required to mark every option in a particular way, such as “True or False”, “Yes or No” Ranking- This format requires the respondent to arrange a number of statements or ideas in order of a given criterion. Scaling/rating- Here, a number of statements are made and the respondent is asked to indicate the extent to which s/he agrees with each statement. An example is the Likert-type attitude scales Procedure for Construction of Questionnaires and Interview Schedules 1. Define the objectives of the study 2. Identify the funds available for the project 3. Define the population and identify the sampling method 4. Ascertain the type of questions that will best meet the objectives 5. Decide which type of results are required 6. Develop the questionnaire and cover letter 7. Determine the type of analysis to be performed on each question in the questionnaire 8. Pilot the questionnaire, revise it if necessary and pilot-test it again Guidelines for developing questionnaire items 1. Eliminate double questions e.g. Are you an effective and efficient teacher? 2. Do not use leading or biased questions e.g. Are you shocked by the many subjects our children are learning? 3. Do not use presuming questions e.g. How do you like teaching in a remote part of the country? 4. Avoid or re-phrase sensitive or threatening questions e.g. How old are you? 5. Each question should deal with a single concept and be worded as clearly as possible e.g. Do you spend a lot of time on making lesson plans? The questions should be relevant • The questions should be vital to the respondent. They should reflect on things that are relevant to the respondent or what they are about. Failure to observe this leads to misleading results. Avoid biased items or terms • Check on the wording of the items to avoid any form of biases. For example: Do you sometimes disagree with the head teacher’s recent remarks on….? Avoid words like sometimes, often, always and instead use ‘number of times per week’. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 52 Keep the questionnaire short. • The instrument should be short. Respondents get bored and tired of very long questionnaires. The maximum number of items should be thirty. Always eliminate questions whose answers can be found in other sources. Some disadvantages of questionnaires:  Questionnaires, like many evaluation methods occur after the event, so participants may forget important issues.  Questionnaires are standardized so it is not possible to explain any points in the questions that participants might misinterpret. This could be partially solved by piloting the questions on a small group of students or at least friends and colleagues. It is advisable to do this anyway.  Open-ended questions can generate large amounts of data that can take a long time to process and analyse. One way of limiting this would be to limit the space available to students so their responses are concise or to sample the students and survey only a portion of them.  Respondents may answer superficially especially if the questionnaire takes a long time to complete. The common mistake of asking too many questions should be avoided.  Students may not be willing to answer the questions. They might not wish to reveal the information or they might think that they will not benefit from responding perhaps even be penalized by giving their real opinion. Students should be told why the information is being collected and how the results will be beneficial. They should be asked to reply honestly and told that if their response is negative this is just as useful as a more positive opinion. If possible the questionnaire should be anonymous. Some advantages of questionnaires:  The responses are gathered in a standardized way, so questionnaires are more objective, certainly more so than interviews.  Generally it is relatively quick to collect information using a questionnaire. However in some situations they can take a long time not only to design but also to apply and analyse (see disadvantages for more information).  Potentially information can be collected from a large portion of a group. This potential is not often realised, as returns from questionnaires are usually low. However return rates can be dramatically improved if the questionnaire is delivered and responded to in class time. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 55 1 = male, 2= female Age group 1 = 15 to 20 years, 2 = 21 to 25 years, 3 = 26 to 30 years The numbers used in here have no level of magnitude ii) Ordinal Scale Ordinal scale ranks objects or observation in order of magnitude but has no zero point. This means it does not show the distance between the values. For example when the class is graded as A, B, C, D and F, it is clear that the person who got grade A has done better than the one who got grade B but it does not show how much the person who got A has defeated the person who got B. Ordinal scale has the property of identity and magnitude only. iii) Interval Scale Interval scale ranks objects or observation in order of magnitude and has a zero point but the zero is arbitrary. The zero point means the distance between the numbers are equal. For example when measuring temperature, the starting point in a thermometer labelled zero does not mean absence of heat but rather it is arbitrarily fixed that at that point it is as good as zero heat. When a student gets zero in a test, it does not mean the students has no intelligence but zero at that point simply means the students ability is as good as Zero. iv) Ratio Scale Ratio scale ranks objects or observation in order of magnitude, has equal intervals between numbers and has a true zero point. For example in measuring distance, weight or height, when one is at the initial point he or she is at zero point. Ratio scale is the highest level of measurement because it has all the four properties of measurement. (Identity, magnitude, equal interval between numbers and true real point). RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 56 TYPES OF MEASUREMENT Frequency A frequency is the number of times an event occurs in given set of object. In other words frequency indicates the number of times similar scores or items appears in a given set of scores or items. For example: Set: 50, 50, 45, 49, 40, 50, 60, 65, 60, {40 Once} {45 Once} (49 Once} {50 Thrice} {60 Twice} {65 Once} Percentage This is the ratio of a score over 100 For example: Set: 50, 50, 45, 49, 41, (50÷235) x 100 = 21.3 (50÷235) x 100 = 21.3 (45÷235) x 100 = 19.1 (49÷235) x 100 = 20.9 (41÷235) x 100 = 17.4 Presentation of Data There are three ways of summarizing data: tabular, graphical and numerical. a) Tabular This involves the presentation of data using tables. The most used tables include the Simple frequency tables, cumulative frequency tables and grouped frequency distribution tables. They are used for nominal and ordinal data. i) Simple frequency distribution Table All possible values if the variables are listed in a column and the frequency of each value listed in the second column to the right. Example Data set: 49, 65, 98, 74, 86, 74, 65, 74, 86, 74 Simple Frequency Distribution RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 57 Marks Tally Frequency 98 I 1 86 II 2 74 IIII 3 65 II 2 49 I 2 372 10 ii) Cumulative frequency distribution This can take two forms: cumulative frequency above or cumulative frequency below. Marks Frequency Cumulative Frequency above Cumulative frequency below 98 1 1 10 86 2 3 9 74 3 6 7 65 2 8 4 49 2 10 2 372 10 iii) Grouped Frequency distribution Example Grouped Frequency Distribution Marks (Class Interval) Frequency 41-50 2 51-60 0 61-70 2 71-80 3 81-90 2 RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 60 iv) Scatter diagram It is used for a bi-variate frequency distribution i.e. that which has two variable. The purpose is to show how the variables are related. The values of the first variable are located on the x-axis and the values of the second variable on the y-axis. The intersection points for corresponding values are denoted by an asterisk. Example Y axis X axis v) Frequency Polygon When constructing a polygon, the same axes as the histogram are used. The midpoints of the class intervals are then used with the corresponding frequencies to plot the frequency polygon. The intersection points are then joined with a straight line. Example 5 10 15 20 25 30 Scores in test Numerical Representation This includes measures of central tendency, measures of variation and measures of association. * * * * * * * * * 12 10 8 6 4 2 RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 61  Measures of central tendency- mean, mode, median  Measures of variability- range, standard deviation, variance  Measures of association- correlation coefficient, regression analysis MEASURES OF CENTRAL TENDENCY There are three common measures of central tendency or typical scores. These are the mean, median and mode. THE MEAN The mean of a set of numerical observations is the sum of the set divided by the number of observations. The symbol µ (Greek lower case “mu”) and X (read x bar) are used to denote a population and sample mean respectively. The following formula is used to calculate the mean. Population mean, µ n X= Where: µ = population mean Σ = sum of observations Σ = total number of observations The following formula is used to calculate the mean. Sample mean, n X XM == Where X = sample mean Σ = sum of the X scores n = total number of scores RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 62 Example Obtain the mean of the following scores 20, 25, 30, 28, 17. Solution: n X XM == 5 1728302520 ++++= 5 120= = 24 ADVANTAGES AND DISADVANTAGES OF USING THE MEAN The mean is widely used in practice because of the following reasons: 1. It is the basis of parametric statistical tests. 2. It is the exact central position on an interval and continuous scale. 3. It is affected by the value of every item in the series. The main disadvantages of using the mean are: 1. Since the value of the mean depends upon each score in the series, a single extreme score in one direction (an ‘outrider’) can distort the value of the mean. 2. It does not follow that any of the scores in the set of the same value is the mean score. THE MEDIAN The median is the score corresponding to the 50th percentile. To find the median of a set, arrange the numbers in a set in numerical order, e.g. 32, 36, 38, 40, 45. For odd numbered lists only – select the central value. For even numbered lists only – calculate the mean of the two central values e.g. in the set 6, 7, 11, 12. The median is 9 2 117 =+ Example Find the median of the following scores, 8,4,5,3. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 65 Reasons for Measuring Variation The following are the main reasons for measuring variation: 1. Measures of variability, also known as measures of dispersion, make it possible for us to gauge how much scores vary from each other or how they spread around the central point of the data and across the distribution. This enables us to gauge the relativity of an average score. 2. The second reason for measuring dispersion is to determine the nature and cause of variation in order to control the variation itself. In education, variations in students’ scores are basic guides to determining students with learning difficulties and designing remedial action. 3. Measures of variability make it possible for us to compare two or more distributions. COMMONLY USED MEASURES OF VARIABILITY The following measures of variability are used most frequently by researchers: range, variance and standard deviation. THE RANGE The range is simply the difference between the highest and smallest score. It is used when we want to make a rough comparison of two or more groups for variability. Example: Find the range of the following set of scores. 2, 8, 9, 12, 15 Computation: The range = 15 – 2 = 13 The range is the interval between the largest and smallest value. Range = L – S Where: L = Largest value S = Smallest value Illustration: The following are the scores of 10 Form II pupils: 22, 35, 36, 29, 42, 56, 59, 60, 67, 72. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 66 Calculate the range. The range is 72 – 22 = 50 The main limitation of the range is that it is based on two values alone and therefore reflects only the dispersion in some defined section of the distribution. Variance considers each data value rather than only the two extremes. Variance averages the total distance between each observation and the mean and since distance is not negative, each difference must be squared, . The standard deviation restores the data to their original measurement unit. Estimation We assume that the population and its parameters are unknown. Ratio Scale Ratio scales have all properties of measurement identity, magnitude, equal intervals and an absolute zero. Examples are age, height, speed, time and class size. VARIANCE The variance is the average of squared differences of the values from the mean. A high variance indicates that most scores are far away from the man; a low variance shows that most scores cluster tightly about the mean. POPULATION VARIANCE σ2 = Σ(x - µ)2 (N) Where: Σ = adds the values in parentheses x = individual values (items) in the data set µ = population mean N = number of values (items) in the data set Illustration: The following marks were obtained by 5 Form IV pupils in an examination: 5, 10, 15, 20, 25. Calculate the variance. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 67 However, as a measure of variability, the variance is difficult to relate directly to the original data. This is mainly because it measures in squared units. This problem is bypassed by taking the square root of the variance, thereby transforming the variance into the standard deviation. THE STANDARD DEVIATION This is the statistic most commonly used to measure the distribution of scores around the mean. The standard deviation (SD) is an average of the degree to which a set of scores deviate from the mean. The following formula is used to calculate the standard deviation. SD = ( ) N XX − 2 or N d 2 Where : X = each individual score X = mean of all the scores Σ = sum of N = number of scores d = deviation of each score from the mean Calculation of the standard deviation 1. Find the mean. 2. Obtain the deviations (subtract each score from the mean score). 3. Square each deviation. 4. Find the sum of squared differences. 5. Divide the value obtained in step 4 by N. 6. Find the square root of the value obtained in step 5. Example 1: Find the standard deviation of the following set of scores: 13, 10, 16, 18, 11, 4. 1. Computation: X = 12 6 72 = RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 70 Converting each of the original scores to z-score yields Kiswahili English Maths x 80 65 78 x 85 55 60 SD 10 5 2 z - 0.5 2 9 It is now clear that the pupil’s worst subject was Kiswahili, 5 points below average and his best subject is Maths with 18 points above average. First, it is possible that the Kiswahili paper was easy and resulted in many high scores, while the English paper was difficult. Secondly, the English paper may have been based on a total of 80 points and the Kiswahili paper on a total of 100. The difficulty of comparing raw scores is overcome by transforming the score on each test to a common scale with a specified mean and standard deviation. These scores are Z or standard scores with a mean of 0 and a standard deviation of 1. Unlike raw scores, Z – scores give an accurate picture of the standing of each score relative to the reference group regardless of the scale that is used. MEASURES OF RELATIONSHIPS A researcher might be interested in finding out the degree to which two variables are related and the extent to which changes in one variable are accompanied by or are dependent upon changes in the second variable. For instance, a researcher may want to determine whether students who perform well on a statistical achievement test also score highly on a research methods achievement test. The correlation coefficient provides researchers with a means of expressing the intensity of the relationship between two or more variables. Two methods are normally used to compute the correlation coefficient. These are: 1. Pearson product moment correlation, and 2. Spearman’s rank correlation. RELATIONSHIPS BETWEEN INTERVAL-LEVEL DATA The Pearson’s Product Moment Coefficient represented by the symbol r is the most commonly used measure of correlation. It is used when the variables we wish to correlate are expressed in continuous scores. For example, a teacher may wish to know which scores RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 71 of Form IV students on a geography test are related to the scores of the same students on a history test. The formula for computing the Pearson’s coefficient correlation is as follows: r = NΣxy - (Σx) (Σy) [NΣx2 - (Σx)2] [NΣy2 - (Σy)2] Where: r = Pearson’s coefficient of correlation N = number of subjects Σ = the sums of x = scores on one test y = scores on the other test Pearson’s product moment correlation coefficient ranges from –1.0 to +1.0. Example Compute Pearson’s product moment correlation coefficient between pupils’ scores on writing and reading tests for the data in the Table. Computation Table: Computation of the Pearson’s – Production Moment Correlation Coefficient between Writing and Reading Scores Student Writing/reading scores x y x2 y2 xy Jack John Paul David Enid Francis Gatu Hellen Issa Jacob 6 5 6 9 8 9 8 7 8 6 8 5 4 7 8 8 6 7 9 8 36 25 36 81 64 81 64 49 64 36 64 25 16 49 64 64 36 49 81 64 48 25 24 63 64 72 48 49 72 48 RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 72 r = NΣxy - (Σx) (Σy) [NΣx2 - (Σx)2] [NΣy2 - (Σy)2] = 10 (513) - 5040 [10(536) - 5184] [10(512) - 4900] = 5130 - 5040 [5360 - 5184] [5120 - 4900] = 90 (176) (220) = 90 38720 = 90 196.77 = 0.457 INTERPRETING CORRELATION COEFFICIENT STRENGTH OF THE RELATIONSHIP The strength of relationship between two variables is the extent to which one value, y, is consistently associated with one and only one value of x. A correlation coefficient is expressed as a decimal value that ranges from +1.00 to –1.00. +1.00 is a perfect positive relationship; 0.00 means that two variables are unrelated; -1.00 is a perfect negative correlation. It means that as one variable increases, the other decreases. Interpreting the strength of a correlation coefficient Guilford (1956) proposes the following guidelines for interpreting the strength of a correlation coefficient when the sample size is fairly large. .20 - .40 Low correlation, definite but small relationship .41 - .70 Moderate correlation, marked relationship .71 - .90 High correlation, substantial relationship >.90 Very high correlation, very dependable relationship (p.145) RELATIONSHIP BETWEEN ORDINAL LEVEL DATA When two variables are measured on an ordinal scale, the correlation co-efficient is calculated by the Spearman’s correlation. Spearman’s rank correlation is used to determine the degree to which each student’s rank on the set of scores tends to be correlated with her rank on another set of scores. The formula for the Spearman’s rank correlation coefficient (rs) is: RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 75 Reporting Quantitative Data There are two types of statistical analysis used: descriptive statistics and inferential statistics. Descriptive statistics include the measures of central tendency and the measures of variation. Inferential statistics are used in testing hypothesis in order to generalize the findings of the study. They include the chi square, the t-test, the ANOVA (Analysis of Variance), regression analysis etc. Ethical Issues in research Ethics are the principles and guidelines that help us to uphold the things we value. It deals with one’s conduct and serves as a guide to one’s behaviour. Ethics in research can be discussed under the following categories: i) Ethical issues relating to individual researcher ii) Ethical issues concerning research subjects, and iii) Ethical issues concerning the research process Ethical issues relating to individual researcher These include plagiarism and fraud, and misuse of privileges by the researcher. • Plagiarism To use someone else’s exact words without quotation marks and appropriate credit, or to use the unique ideas of someone else without acknowledgment, is known as plagiarism. In addition, stealing ideas from another scholar is also plagiarism. In publishing, plagiarism is illegal; in other circumstances, it is, at the least, unethical. • Fraud Fraud refers to a situation where a researcher fakes data that has not actually been collected, or the false presentation of research methodology and results. Fraud is a punishable crime. • Misuse of privileges Research subjects normally participate in the research on trust. It would therefore be extremely unethical for researchers to abuse this trust by using their position power negatively. E.g. it would be unethical for a doctor researcher to undertake certain research tests on the pretext of providing treatment. Ethical issues concerning research subjects These include confidentiality and privacy, anonymity, physical and psychological harm, voluntary and informed consent, use of vulnerable and/or special populations, dissemination of findings and, financial issues in sponsored research. RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 76 • Confidentiality and privacy Keep the information provided by respondents confidential. Seek the consent of respondents before revealing any information. • Anonymity Here you identify the individuals being protected by using numbers or pseudo names. Refers to a situation where a respondent’s name is not disclosed. • Physical and psychological harm Never undertake research that may cause physical or psychological harm. Psychological harm can be caused by asking embarrassing questions, expressing shock or disgust while collecting data, using threatening statements or causing fear and anxiety among respondents. • Voluntary and informed consent Voluntary consent-respondents willingly participate in research. It is unethical when the researcher fails to disclose the purpose of research for fear that respondents may refuse to participate in the research. Informed consent- Request the respondents to participate in the research. A respondent must be told the truth and be given all the facts about the research in order to make informed decision about participating or not. • Use of vulnerable or special populations These include children, mentally disabled, sick people, the poor, street children etc. Permission from those who care for them must be sought and must be based on the principle of informed consent. Dissemination of findings It is unethical to conceal research findings after completion of research. Where a joint research was undertaken, it is unethical for a researcher to publish the findings as a personal effort without consulting the other team members. • Academic freedom Researchers should freely share their findings without fear of intimidation, loosing of jobs or being victimized. Ethical issues concerning research process • Ignoring pertinent issues in research It is unethical to purposely avoid a pertinent research issues/topic for fear of consequences or due to conflict of interest. • Experimental designs RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 77 The unequal treatment of the experimental and control groups in unfair and unethical. To avoid this, provide the same treatment to the control group after the study. • Use of tests in research Subjecting respondents to tests is unethical due to discomfort and anxiety experienced. Writing a Research Report The research project should contain all or some of the following materials arranged in the order shown below. Chapters 1-3 constitute the research proposal which is written in future tense but changed to past tense in the final research report. Chapters 4-5 are only found in the research report and not the proposal. 1. PRELIMINARY PAGES i) Title ii) Certification iii) Dedication iv) Abstract v) Acknowledgements vi) Table of Contents vii) List of Tables viii) List of Figures 2. MAIN BODY OF THE REPORT Chapter 1: INTRODUCTION 1.1 Background to the problem 1.2 Statement of the problem 1.3 Research Questions 1.4 Research Hypotheses (Optional) 1.5 Significance of the study 1.6 Scope and delimitations of the study 1.7 Assumptions of the study (Optional) 1.8 Limitations of the Study (Optional) 1.9 Theoretical Framework/Conceptual Framework 1.10 Operational definitions of key terms Chapter 2: REVIEW OF RELATED LITERATURE 2.0 Introduction 2.1 Review of Related Theories 2.2 Review of Related Empirical Studies 2.3 Restating the Research Problem Chapter 3: RESEARCH DESIGN AND METHODOLOGY 3.0 Introduction 3.1 Research Design RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 80 ii) Certification Page This page contains two statements. The first statement should be by the author to the effect that the study is the author’s original work. The second statement should be signed by the supervisor(s) to the effect that the thesis has been submitted for examination with his/her/their approval as university supervisor(s). Example iii) Dedication The research report may be dedicated to a person or persons who deem special in the author’s life e.g. spouse, children, parents, or very dear friends. This should be a brief statement. Examples: a) This study is dedicated to my daughter Faith Forever. b) To all my teachers iv) Abstract This is a brief summary of the report and should be no more than 500 words. It summarizes the entire thesis by presenting the reader with the research problem, research questions, the population studied, methods of data analysis, major findings and conclusions of the study, and recommendations. Since the abstract refers to work which has been done, most or all of it should be written in the past tense. It is written only after other sections of the thesis have been written. CERTIFICATION This research project is my original work and has not been presented for degree in any other university ……………………………… Furaha Mashaka 5/10/2017 This research project has been submitted for examination with our approval as university supervisors …………………………………… ……………………………… Dr. Busara Mwema Date …………………………………… ……………………………… Mr. Willy Diid Date RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 81 v) Acknowledgements The author expresses appreciation to people and institutions that have contributed significantly to the completion of the report e.g. research supervisor(s), organizational head(s), head teachers, teachers, parents, data coders and typists. vi) Table of Contents This section shows the chapter number, chapter headings and subheadings and their corresponding page numbers. Example. vii) List of Tables This shows the number and titles of tables and their corresponding pages. viii) List of figures This shows the number and headings of each figure and its corresponding page. 2. MAIN BODY OF THE REPORT This contains chapters one to five of the study. Table of Contents Content Page Title Page …………………………………………………………………………………i Declaration……………………………………………………………………………..…ii Acknowledgements............................................................................................................iii Dedication...........................................................................................................................v Table of Contents………………………………………..……………………………….vi List of Tables……………………………………………………………………………...ix List of Abbreviations…………………………………………………………………….xiv Abstract.............................................................................................................................xvi 1.0 CHAPTER ONE: INTRODUCTION………………………………………………1 1.1 Background to the Study………………………………...………………………...1 1.2 Statement of the Problem……………………………………………………..…....7 1.3 Research Questions/Hypotheses…………………………………..……….… .….9 RESEARCH METHODS Instructor: Herbert Makinda, 2018 Page 82 Chapter 1: Introduction This section provides a background to the problem, indicating how the need for the study arose. This is followed by statement of the problem, research questions and hypotheses, significance of the study, theoretical/conceptual framework, scope and delimitations of the study, assumptions of the study and operational definitions of key terms used in the study. Chapter 2: Literature Review This section focuses on prior research studies that are relevant to the current study. It is advisable to start with a review of the research in the broad area. The review should be a critical analysis of the selected works to reveal the done and the untouched, therefore revealing the gaps that require filling. It is similar to that used in the proposal except that the future tense used in the proposal changes to past tense. Chapter 3: Research Design and Methodology Describe exactly how the study was carried out, the sample and how it was selected. Give a description of the demographic characteristics of the subjects and the institutions from which the subjects were obtained. This should be followed by a description of the instruments used in the study. The data collection subsection describes the techniques used to obtain data. Finally, describe the statistical and qualitative techniques used to analyse data. Chapter 4: Findings In this section, after analysing data, the findings of the study are presented. First present the results involving descriptive statistics followed by the results of inferential statistical tests. The researcher should explain the main findings of the study and compare them with what has been previously written on the subject. Chapter 5: Summary, Conclusions and Recommendations In this section, a summary of the study and implication of the main findings are given. Conclusions and recommendations are also given. Areas that need further research are also suggested. 3. REFERENCING This section lists the references that have been cited in the thesis. Good referencing tells the reader which parts of the thesis are descriptions of previous knowledge and which