RESEARCHPROPOSALDEVELOPMENT-

RESEARCH METHODS BY DR. CHARLES SAANANE (DR.PHIL.NAT.). PREAMBLE: RESEARCH refers to search for knowledge. More appropriately, research is a scientific and systematic search for pertinent information on a specific topic. Actually, research is an art of scientific investigation. Research can be considered as a movement from the known to the unknown. It is a voyage of discovery. All of us possess the vital instinct of inquisitiveness because when the unknown confronts us, we wonder and our inquisitiveness makes us probe as well as get full and fuller understanding of the unknown. Such inquisitiveness is the mother of all knowledge and the method that a human employs for obtaining knowledge of whatever the unknown, can be termed as research. Indeed, research is an academic endeavour and as such, the term should be used in a technical sense. Research comprises defining and redefining problems, formulating hypothesis/hypotheses or suggested solutions; collecting, organizing and evaluating data, making deductions and reaching conclusions; and finally, carefully testing conclusions to determine whether or not they fit the formulated hypothesis. Thus, research is an original contribution to existing stock of knowledge making for its advancement. It is the pursuit of truth with the help of study, observation(s), comparison(s) and experiment(s). In short, it can be asserted that research is search for knowledge through objective and systematic method of finding solution/solutions to a research problem. Systematic approach concerning generalization and formulation of a theory is also research. Therefore, the term ‘research’ refers to systematic method consisting enunciating the problem, formulating a hypothesis, collecting facts or data, analyzing the facts/data and reaching certain conclusions either in form of solutions towards the 1 concerned problem or in certain generalizations for some theoretical formulation. Objectives of research The purpose of research is to discover or unravel answers to questions by applying scientific procedures. The main aim of research is to find out truth, which is hidden and that has not been discovered. Though each research has its own purpose, research objectives fall into several of the following broad groupings: - To gain familiarity with a phenomenon or to achieve new insights into it, such studies with this object are termed exploratory or formulative research studies; - To portray accurately characteristics of a particular individual, situation or a group, such studies with this object in view are known as descriptive research studies; - To determine frequency with which something occurs or with which it is associated with something else, such studies with this object in view are called diagnostic research studies; and - To test a hypothesis of a causal relationship between variables, such studies are known as hypothesis-testing research studies. Motivation in research There are possible motives for doing research. They may be either one or more of the following: - Desire to get a research degree along with its consequential benefits; - The need to face a challenge is solving unresolved problems, that is, concern over practical problems initiates research; - The need to get intellectual joy of doing some creative work; - The need to be of service to society; and - The need to get respectability. Note, there could be many more factors such as government directives, employment conditions, curiosity about new things, need to understand causal relationships, social thinking as well 2 awakening and so on. All may motivate or at times compel people to undertake research works. - Types of Research Basic types of research include the following: i) Descriptive Research versus Analytical Research; ii) Applied Research versus Fundamental Research; iii) Quantitative Research versus Qualitative Research; iv) Conceptual Research versus Empirical Research; and v) Some Other Types of Research. (i) Descriptive Research versus Analytical Research Descriptive research includes surveys and fact-finding enquiries of different kinds. The major purpose of descriptive research is description of state of art affairs that it exists at the moment. In social sciences and business research, proponents usually use the term Ex post facto research for descriptive research studies. In this type of research, the researcher has no control over variable. He/she can only report what has happened or what is happening. Most ex post facto researches are used for descriptive studies in which the researcher seeks to measure items such as frequency of shopping, preferences of people, say over a certain mobile phone provider like TIGO or similar data. Ex post facto researches also include attempts by researchers to discover causes even when they cannot control variables. Methods of research utilized in descriptive studies are survey methods of all kinds that include comparative and correlation methods. On the other hand, in analytical research, the researcher has to use facts or pieces of information already available and analyze them to make a critical evaluation of the material. (ii) Applied Research versus Fundamental Research Research can either be applied (or action) research or fundamental (or basic or pure) research. The former aims at finding a solution for an immediate problem facing society or an industrial/business 3 organization, whereas the latter is mainly concerned with generalizations and formulation of a theory. Thus, gathering knowledge for knowledge’s sake is known as ‘pure’ or ‘basic’ research. Research concerning some natural phenomenon or relating to pure mathematics are examples of fundamental research. In the same vein, research studies concerning human behaviour carried out with a view to make generalizations about human behaviour are also examples of fundamental research. But research aimed at certain conclusions, for example, a solution facing a concrete social or business problem is an example of applied research. In addition, research to identify social, economic or political trends that may affect a particular institution or copy research (research to find out whether or not certain communications will be read and understood) or marketing research or evaluation research are all examples of applied research. Therefore, the central aim of applied research is to discover a solution for some pressing practical problem(s), whereas basic research is directed towards finding information that has a broad base of application and thus, adds to already existing organized body of scientific knowledge. (iii) Quantitative Research versus Qualitative Research Quantitative research, as the name implies, is based on measurement of quantity or amount. Quantitative research is applicable to phenomena that can be expressed in terms of quantity/quantities. On the other hand, qualitative research is concerned with qualitative phenomena, phenomena relating to or involving quality or kind. For example, when scientists are interested in investigating reasons for human behaviour, that is, why people think or do certain things, scientists often think of ‘Motivation Research,’ an important type of qualitative research. Such type of research aims at discovering underlying motives and wants to use in-depth interviews for the purpose. Other data collection methods for such research are word association tests, sentence completion tests, story completion tests and similar other 4 projective data collection methods. Attitude or opinion research, that is, research designed to find out how people feel or what they think about a particular subject or institution is also qualitative research. Qualitative research is especially important in behavioural sciences where the aim is to discover underlying motives of human behaviour. Through such research, scientists can analyze various factors, which motivate people to behave in a particular manner or which make people like or dislike a particular thing. However, in practice, to undertake qualitative research is a relatively difficult job and thus, while doing such research, one should seek for guidance from experimental psychologists. (iv) Conceptual Research versus Empirical Research Conceptual research is research related to some abstract idea(s) or theory. Conceptual research is generally used by philosophers and thinkers to develop new concepts or to re-interpret existing concepts. On the other hand, empirical research relies on experience or observations alone, often without due regard for system and theory. Empirical research is data-based research, coming up with conclusions that are capable of being verified by observation or experiment. It can also be called experimental type of research. In such a research, it is necessary to get at facts firsthand, at their source and actively go about doing certain things to stimulate production of desired information. In such a research, the researcher must first provide himself/herself with a working hypothesis or guess as to probable results. He/she then works to get enough facts (data) to prove or disprove his/her hypothesis. He/she then sets up experimental designs, which he/she thinks will manipulate persons or materials concerned so as to bring up the desired information. Thus, such research is characterized by the experimenter’s control over variables under study and his/her deliberate manipulation of one of them to study its effects. Empirical research is appropriate when proof is sought that certain variables affect other variables in some way. Evidence gathered 5 through experiments or empirical studies is considered to be the most powerful support possible for a given hypothesis. (v) Other types of Research. All other types of research are variations of one or more of the presented research approaches based on either purpose of research or time required to accomplish research or environment in which research is done or on the basis of some other similar factor. Thus, research can be thought either as one-time research or longitudinal research. In the former, research is confined to a single-time period, whereas in the latter case, research is carried out over several time-periods. Research can be field-setting research or laboratory research or simulation research, depending on environment in which it is to be conducted. Also research can be understood as clinical or diagnostic research. Such research follows case-study methods or in-depth approaches to reach basic causal relations. Such researches are carried out deep into causes of things or events of interest by using very simple samples and very deep probing data collection instruments. The research may be exploratory or may be formalized. The objective of exploratory is development of hypotheses rather than their testing, whereas formalized researches are researches with substantial structure and are with specific hypotheses to be tested. Historical research utilizes historical sources such as documents, material object/remains, oral history and so n to study events or ideas of the past including philosophy of persons and groups at any remote point of time. Also there could be conclusion-oriented and decision-oriented research. While undertaking conclusion-oriented research, the research is free to pick up a problem, redesign the enquiry as he/she proceeds and prepared to conceptualize as he/she wishes. On the other hand, decision-oriented research is always for the need of a decision-maker and, in this case, the researcher is free to embark upon research according to his/her own inclination. 6 Operations research is an example of decision-oriented research because it is a scientific method of furnishing departments with a quantitative basis for decisions regarding operations under their control. Research Approaches There are three types of research approaches, namely, Quantitative Research Approach; Qualitative Research Approach; and Mixed Research Approach. The first involves generation of data in quantitative form that can be subjected to rigorous quantitative analysis in a formal and rigid fashion. The approach can be further classified into the following approaches: inferential, experimental and simulation approaches to research. The gist of inferential approach to research is to form a data base from which to infer characteristics or relationships of population. It usually means survey research where a sample of population is studied (questioned or observed) so as to determine its characteristics and then it is inferred that the population has the same characteristics. On the other hand, experimental approach is characterized by much greater control over the research environment. In this approach, some variables are manipulated to observe their effect(s) over other variables. Then, simulation approach involves construction of an artificial environment within which relevant information and data can be generated. That permits an observation of dynamic behaviour of a system (or its sub-system) under controlled conditions. Simulation approach can be useful in building models for understanding future conditions. Qualitative research approach is concerned with subjective assessment of attitudes, opinions and behavior. In such a situation, research is a function of the researcher’s insights and impressions. 7 Such an approach generates results either in non-quantitative form or in form, which are not subjected to rigorous quantitative analysis. Generally, techniques of focus group interviews, projective techniques and depth interviews are used. Mixed Research Approach is one that uses both Quantitative Research Approach and Qualitative Research Approach. RESEARCH METHODS VERSUS METHODOLOGY There is a difference between research methods and research methodology. Research methods are all methods/techniques that are used for conducting research. Research methods or techniques refer to methods researchers use in performing research operations. That means all methods that are used by the researcher during the course of studying his/her research problem are termed research methods. As long as the object of research, in particular, applied research, is to arrive at a solution for a given problem, available data and unknown aspects of the problem have to be related to each other to make solution possible. There are three groups of research methods: - First, methods concerned with data collection. Such methods will be used where already available data are insufficient to arrive at the required solution. - Second, statistical techniques that are used for establishing relationships between data and the unknown. - Finally, methods that are used to evaluate accuracy of obtained results. It has to be noted that research methods under the last two groups are generally taken as analytical tools of research. Research methodology is a way of systematically solving a research problem. It is a science of studying the manner research is done scientifically. In it, various steps are studied that are generally adopted by a researcher in studying his/her research problem along 8 with the logic behind them. Thus, it is necessary for the researcher to know not only research methods/techniques but also the research methodology. Researchers not only need to know the manner to develop certain indices or tests, the way to calculate the mean, the mode, the median or the standard deviation or chisquare, but also they need to know which of these methods or techniques are relevant and which are not, and what would they mean and indicate and get reasons. Also researchers need to understand assumptions underlying various techniques and they need to know criteria by which they can decide that certain techniques as well as procedures will be applicable to certain problems and others will not. It means that it is necessary for the researcher to design his/her methodology for his/her problem because they may differ from problem to problem. Thus, in doing research, the scientist has to expose the research decisions to evaluation before they are implemented. He/she has to specify very clearly as well as precisely what decisions he/she selects and why he/she selects them so that they can be evaluated by others. In due regard, research methodology has many dimensions, while research methods constitute part of the research methodology, the scope of research methodology is wider than that of research methods. That means by referring to research methodology we not only refer to research methods but also consider the logic behind the methods we use in the context of our research study and explain why we are using a particular method or technique and why we are not using others so that research results are capable of being evaluated by the researcher himself/herself or by others. Thus, why a research study has been undertaken, how the research problem has been defined, in what way and why the hypothesis has been formulated, what data have been collected and what particular method has been adopted, why particular technique/techniques of data analysis have been used and a lot more similar questions are usually answered when we talk of research methodology concerning a research problem or study. 9 TPOIC 1: SCIENTIFIC EPISTEMOLOGY 1.1 Science as process and product Scientific Epistemology Indeed, it is essential that in order to plan research effectively, the researcher should understand how his/her activities fit into the endeavour of science as whole. Some explanations of the “scientific method” confound epistemology – the manner we accumulate knowledge and understanding through science – with specific research activities of the investigator. Science can be conceived as a cycle of activities as well as results based on procedures that are often referred to as the “hypotheticodeductive” method. The method evaluates a hypothesis (more generally a model) of the manner the world might work and deduces consequences that must be true if the world actually works in the way posited. One can then check the real world to see if these predicted consequences are verifiably true. The process is cyclical because if consequences cannot be verified, a new hypothesis (model) must be tried. Even if predicted consequences can be verified, however, that result might be coincident. Therefore, one strives to deduce new predictions from the same hypothesis and test these as well. The cycle, known as the epistemological cycle, can be represented schematically as shown in Figure 1. 10 Figure 1: A schematic representation of the epistemological cycle MODEL induction deduction DESICION PREDICTION Comparison observation DATA Figure 1 uses some simple conventions of illustration. Boxes, which are labeled in FULL CAPS, represent statements that can be written down: products of processes. Arrows, which are labeled in lower case, represent processes that give statements in the boxes to which they point. Furthermore, each arrow originates at a statement and points to another statement so as to show the sequence of steps in the process of science. As the diagram represents a cycle, we can scrutinize its parts beginning at any arbitrary place and then return to that place by completing the cycle. The practicing scientist, however, usually enters the cycle at one of the two places: creating (or revising) a model from data at hand, or drawing testable predictions from an existing model. Let us begin our discussion at the latter point, assuming that a model of some natural phenomenon already exists. 1.2 Deduction and prediction Deduction is a type of reasoning that in logic leads from a set of premises to a conclusion. In science, logical premises constitute the model, which is a speculation of the manner things work in nature and the logical conclusion is a specific prediction that is testable by observation. In other words, a deduction is the result of 11 the process of deducing, but that use would give two meanings to the same word, thus, we restrict “deduction” to the process itself. Deductive process can be thought of as a rearrangement of knowledge. In the terminology of logic, deductive reasoning extracts from a set of premises (the MODEL of Figure 1) one or more conclusions (the PREDICTION of Figure 1). There is no new knowledge to the prediction; everything in the prediction is already inherent in the model. The deductive process simply isolates part of the model, or isolates several parts and combines them. For example: MODEL: the earth rotates on its axis, spinning counterclockwise when viewed from above the North pole. From this model, one can deduce the prediction that the sun (and the other stars, for that matter) should rise above the Eastern horizon, travel across the sky and set in the West. The model embeds some hidden assumptions, such as all these celestial bodies being fixed in space relative to the earth. Almost all models have implicit assumptions and failing to recognize them could lead to problems in reasoning. The prediction deduced can be written down and it can be checked empirically, which means that one uses the process of observation, including informal measurement, to see if the predictions match reality as we view it. Deduction is at heart a stipulated set of rules for assuring this relationship between model and prediction: if the model is true, the prediction deduced from it must also be true. The rules take many forms, the oldest of which is Aristotelian syllogism. In its commonest form, the syllogism produces a conclusion (prediction of scientific epistemology) from two premises (together making the model in epistemological terms). 12 For example: PREMISE 1 (part of model): If only Mazengo drives the four wheel drive vehicle PREMISE 2 (part of model): And if the four wheel drive vehicle is on a field trip CONCLUSION (prediction): Then Mazengo is on a field trip Of course science does not depend upon syllogistic reasoning, which is full of pitfalls because of the inherent impression of language. For example, consider the following syllogism: Goats all have horns (premise 1), and this animal has a horn (premise 2), therefore, this animal must be a goat (conclusion). The logical error in that syllogism is termed a fallacy of affirmation and is common in reasoning with language; just because A implies B (if it is a goat, it has horns) does not necessarily mean that B implies A (if it has horns, it is a goat). The error leads to apparent substantiation that goats exist. Anyone can see through the problem in a simple syllogism like this one, but much reasoning from scientific model is similarly, linguistic in nature and so entails all dangers of language itself. Linguistic reasoning does not have to be in syllogistic form. For example, it is possible to draw a conclusion from only one premise: PREMISE: If a horse is an animal CONCLUSION: Then a horse’s head is an animal’s head Most scientific deduction is basically mathematical in nature and, conversely, most mathematics taught in secondary schools is deductive in nature, for example, Euclidian geometry and algebra. Simple algebraic deduction can be written in syllogistic form: PREMISE 1: If x + y = 6 PREMISE 2: And if y = 4 CONCLUSION: Then x = 2 13 Note that the deductive logic underlying the earlier example concerning rotation of the earth on its axis was geometric. The link between linguistic reasoning and mathematics is symbolic logic. Many such logic systems have been devised from different starting points and, in most cases, different systems are easily shown to be equivalent. Thus, given the same set of premises (together, the model of Figure 1), deduction by rules of any given system leads validly to the same conclusion (the prediction of Figure 1). Boolean algebra is a form of symbolic logic that stands kind of midway between linguistic reasoning and traditional mathematics. The Boolean system uses linguistic-like operators AND, OR, IF, THE, and EXCEPT to relate variables such as propositions. Formal set theory is even more mathematical-like, using symbols for operators. All systems of deduction have in common the key property: if premises (model) are true, then the conclusion (prediction) deduced from them is true, assuming that the deductive process followed the rules of the system. Observation and Data Observation in Figure 1 is the process that leads to data. “Observation” might connote merely not visually what an animal is doing or some other aspect of the world. “Observation” is used as a general term to include all types of ways in which human senses are extended by instruments to record data. Ultimately, the investigator observes: for example, observes a digital display on an instrument, or observes sound spectrograms made from tape recordings. Thus, observation in Figure 1 means any sensing and recording process that leads to data that can be written down (or otherwise represented in hard copy). The term “observation” might also be applied to data produced by observing. In order to avoid confusion, “observation” is restricted 14 to the process and “data” should mean to describe results of that process. In Figure 1, the arrow originates at PREDICTION because prediction specifies the kinds of data that are needed to be observed. The prediction states what must be the case if the model is true and data show whether or not the prediction is upheld in the real world. Data that have nothing to do with the prediction might be an informative sidelight to a particular investigator’s activities, but they have no direct bearing on the empirical test of the model. Observation with no particular model in mind can produce data that ultimately lead to a hypothesis. Recall, it has been mentioned before that the investigator usually enters the epistemological cycle with a model to be tested or with data (often from literature) for generating or revising hypotheses. Thus, mentoring the scientific cycle through simple observation without starting predictions is an exception. The term “experiment” does not appear in Figure 1. It has been avoided because it tends to connote a carefully controlled laboratory environment in which every attempt is made to control extraneous variables that could influence data. There is no fundamental difference, in terms of epistemology, between a laboratory experiment and field observations. Each approach to data collection relevant to a prediction has its advantages and disadvantages. Laboratory experiments usually provide considerable control over extraneous variables that could influence results, but the laboratory environment may produce artifacts. Field studies may be greatly natural and realistic, but they generally exercise little control over extraneous variables that could influence results. Both laboratory experiments and field observations play a role in research and can provide a particularly powerful approach when used in concert. 15 Comparison and decision As with deduction and observation, the term “comparison” could be applied to both the process and results of that process. In order to avoid confusion, the term “comparison” is used to refer only to the process. Results of the process constitute the “decision” (whether or not results fit the prediction). The process of comparison commonly uses statistical methods for comparing data with predicted results. For example, a model might predict that older animals of some species tend to dominate younger ones. Data could show that the predicted relationship is an imperfect one. Thus, the question becomes whether dominance structure is unrelated to age or is influenced by age as predicted or by some other age-associated traits. The investigator would probably employ some appropriate statistical test to see if dominance relations were random or non-random with respect to age. Comparison between data observed and the prediction deduced from the model gives a decision. Ideally, the decision is simply whether or not data fit the prediction, but things do not always turn out so nicely. One may decide that it is impossible to tell whether or not there is a match. A common outcome of the process of comparison is that some expected difference is not statistically reliable and yet, there is a trend in the predicted direction. Therefore, the difference could be real but not established by data, either because extraneous variables unduly influenced the data (as commonly occurs in the field studies) or because the sample size was insufficient to provide statistical reliability. In such cases, the main recourse is to collect better data, either with improved control over extraneous variables or with larger samples. Yet, another way in which comparison between predictions and data can fail to yield an unambiguous decision about the model is 16 when some predictions are supported and others are not. Any validly deduced prediction that is rejected by empirical data falsifies the model, but when the deductive chain is not tight and other predictions are consistent with data, researchers sometimes refer to “partial confirmation” of the model. Such a situation usually suggests that the model is not quite right but has merit that could be improved by modifications based on careful scrutiny, without having to “go back to the drawing board.” 1.3 Induction and model The decision resulting from comparison between data observed and the prediction deduced from the model dictates the next step in doing science. As noted before, three kinds of decisions are possible. They include the following: first, data do not resolve the question because they are inadequate or conflicting; second, data do not match the prediction; or third, data do not confirm the prediction. In the first case, if data are insufficient there is nothing to do except to go back and collect data adequate to the task. On the other hand, if results testing different predictions conflict, the model is probably not quite right and needs to be modified. Suppose data show unambiguously that the prediction cannot be correct, the data are simply not as predicted if the model is true. The only explanation for that situation is that the model is false. It is said in science that data reject the model because they do not match the prediction deduced from the model. In this case, it is necessary to produce a new model, or at least revise the old one and then proceed with making and testing new predictions. The creative process of proposing how nature works involves induction. Induction is sometimes characterized as reasoning from particulars to generalities, but that seems vague and, in any case, 17 may not always apply. Induction is thought of a cluster of very complicated creative processes in which the thinker identifies possible patterns from available facts and proposes causal relationships that might explain the patterns. There are no rules for creative induction. Thus, induction involves many complex mental processes that find a pattern where none was previously recognized. Most of the famous scientific models have come from persons who reflected on disparate empirical data and somehow united them into a coherent framework. For example, Sir Charles Darwin realized that traits of parents and their offspring tend to be similar (note that genetic inheritance was not yet understood by then), that more offspring are born than can survive to reproduce themselves and that survival probably depends at least partly on traits of the individual. From such empirically verifiable facts, Darwin reasoned correctly that if the survival traits are heritable, evolution must occur. That was Darwin’s model of natural selection. The example from Darwin is what the historian of science, Thomas Kuhn, has called paradigm shifts or revolutions. These are whole reorganizations of thinking in a given area of science. Kuhn first believed that progress in scientific understanding was completely dependent upon such revolutions. However, he later realized that stepwise revision of models also moved science forward. Thus, an entire spectrum exists from minor polishing (or sharpening) models through substantial revisions and generalizations to major shifts in paradigms. All have their place in the progress of science. Uniqueness of models As long as induction is a type of creativity, scientific models that result from it are unique to their creators. Next are some examples that illuminate the nature of both the inductive process and models that result from it. 18 First, radar was invented nearly simultaneously and independently on both sides of the Atlantic at about the time of the Second World War. Radar was a marvelous technological innovation, but it was not a scientific model. Modern inventions are hardly the product of experimenting. They usually incorporate all kinds of science in their development, but they are not scientific models. Essentially, the same invention may be developed independently by two different parties. Second, there was the famous independent creation of the Calculus by Leibnitz and Newton. The two geniuses quarreled over priority, each believing that he had conceived the Calculus first. However, it has to be noted that Mathematical techniques are not scientific models. They cannot be rejected by empirical facts, but rather, they are logic systems that must be evaluated by their natural internal consistency. Third, the planet Neptune was discovered independently and almost simultaneously by different astronomers. That was a scientific discovery, but it was not the creation of a causal model of how nature works. In fact, discovery of Neptune was at DATA in the sense of our illustration and had been predicted on the basis of a model erected to account for perturbations in the orbit of Uranus. Of course, two scientists can make the same empirical discovery wholly independently of one another. Finally, the most celebrated proposal for independent creations of the same model is the idea of natural selection proposed simultaneously in 1858 by Alfred Wallace and Charles Darwin in separate essays published together. Darwin was convinced that they had said exactly the same thing and had even used some of the same examples such as domestication as a form of selection. There is no doubt that Wallace and Darwin had similar ideas and used related examples, but a close comparison of their original essays shows that the models were distinctly different. 19 In sum, two scientific models may appear similar and even be virtually identical in some respects. Nevertheless, it is unlikely that all underlying premises will be the same. Thus, each model is truly unique. 1.4 Models: hypothesis, theory and law Back to our epistemological cycle, there arises a question, what happens when data observed match the prediction deduced from the model? One may be first tempted to say that the empirical test has proven the model to be true. However, this cannot be the case simply because false models can make verifiably true predictions. Let us consider the model that the earth is fixed in space and the heavens rotate about it from East to West. The model predicts that the sun should rise from the East, move across the sky, and set in the West. As early risers know, such a prediction can be verified empirically on a clear day. Yet, science holds that the model from which it was validly deduced is FALSE. The reason we reject the model is that it makes certain other, more subtle and detailed, predictions that are not upheld by comparisons with other data. An inevitable consequence of scientific epistemology is that no model can ever be proven true. The inevitability is inherent in the method of science. The only certainty achievable is rejection of a false model, as true models generate only true predictions. False models can generate predictions that fail to match reality and thus, may be discovered to be false, but false models can also generate true predictions. Therefore, finding that observed data match a deduced prediction merely suggests that the model might be true. Data can confirm a prediction but they cannot prove as true the model from which the prediction was generated. The diagram of the epistemological cycle of science (Figure 1) might be somewhat misleading when it comes to a decision that 20 data match the prediction. Here, no new revised model is called for, but merely further testing of the model that survived the current test. The arrow from DECISION to MODEL is correct, but it indicates that the process of science cycles anew. In this case, the label “induction” may not be applicable. The only process represented by the arrow is that of beginning the cycle of epistemology for another round. Nevertheless, it has been argued that because few models correspond with reality exactly, practicing scientists really modify their underlying models at least a little with every pass through the epistemological cycle. Each pass of the cycle that fails to reject the model under scrutiny increases our confidence that the model might be true. Labels are commonly used to indicate the degree of confidence placed in a model. There is no special name for very preliminary models, especially those of which have not been checked against already available facts. Such models are informally referred to as ideas, suggestions, hunches, possibilities and so forth. “Hypothesis’ is the name given to a well-stated model that is already known to be in accordance with the existing body of evidence and may even have survived some formal, empirical tests. Thus, a hypothesis is a proposition or set of propositions set forth as an explanation for occurrence of some specified group of phenomena either asserted merely as a provisional conjecture to guide some investigation or accepted as highly probable in light of established facts. Quite often a research hypothesis is a predictive statement, capable of being tested by scientific methods, that relates an independent variable to some dependent variable. “Theory” designates a well-tested model, one in which the scientific community places a great deal of trust because it has survived repeated attempts to prove it false (for example, theories of dinosaur extinction – Decan traps in India, volcanism and so on). “Law” applies to only the best-tested models, those so robust 21 that they are unlikely ever to be rejected entirely, for example, Newton’s laws of motion in mechanics. Occasionally, laws fail to make true predictions in unexpected realms, so they have less than universal generality. In such cases, the laws might turn out to be special cases of a more general model – as classical Newtonian mechanics proved to be a special case of Einstein’s general relativity model. 1.5 Role of the individual researcher Every scientist does at least some of the things shown in Figure 1 – deduction, observation, comparison and induction – but not necessarily all of them, especially on a given research project. For example, Physicists commonly characterize themselves as either theoreticians or empiricists. The former spend most of their energies devising mathematical models, while the latter devote to experimental data collection in order to test the models. Similar divisions, usually not so clear-cut occur in biological disciplines such as community and population ecology. Even if such a division is not general for all research undertakings, Biologists and other scientists may restrict their activities to one or several parts of the epistemological cycle for a given research project. In addition, most biological research projects are multifaceted involving several related but different phenomena to which several different models may apply. The multifaceted approach is especially common in field studies such as of behavioural ecology. That fact means that the researcher may be involved with different parts of the epistemological cycle for different aspects of the study, as in data collection about one aspect while attempting to devise a model about another aspect. 22 2.0 Planning Research Preamble Planning is the essence of success in science because when things go in an unanticipated direction, it is easier to make viable adjustments in a well-conceived plan rather than when the approach was hazy to begin with. It is important to plan carefully as there may be no second opportunity. Consider the fact that funding for research is a golden opportunity and an unplanned field research may fail such that there may be consequences. Simple good fortune can play an important role in science, but the general saying that chance favours the prepared mind cannot be overlooked. The prepared mind in science ranges broadly in background reading and familiarity with other disciplines. Students frequently make the mistake of concentrating extremely narrowly on materials of their chosen discipline. It is strongly recommended that students should seek out interdisciplinary seminars, go to talks outside their immediate fields and read journals and books in other disciplines. A general background in the sciences and social sciences pays good dividends when trying to find a problem, to formulate a model and to devise testable predictions in your own area of science. For example, a sociologist seeking for data on behavioural aspects related to HIV/AIDS and Sexually Transmitted Illnesses (STIs) may need to have general background to Pathology, Epidemiology and so on such that he/she can read books, journals from such disciplines. 2.1 Finding a problem Students or novice researchers often experience difficulties in getting started on doing research because they do not yet know 23 how to find a problem worthy of investigation. They may be familiar with many theoretical issues in their field but do not yet know how to translate generalities into specific problems for investigation. Some comments that may help to identify a feasible problem. First, the student/novice researcher should go to her or his academic advisor/supervisor. One should not make the mistake of believing that he or she must have a well-articulated research problem ready for scrutiny. Instead, such a person must be prepared to explain things that seem to motivate her or him most: field versus laboratory work, observation versus experimentation and kinds of phenomena that grab one’s attention. In other words, the student/novice researcher has to try to circumscribe the sphere of possibilities within which she or he might work. Good advisors do not want to assign specific problems in concert with and tailored to the individual advisee. A first experience in science, such as an honours (seniour) thesis (one that is done in the third year for our case), usually requires the advisor to impose a lot of structure based on her or his own research experience. As one progresses to master or doctoral studies, the advisor will expect the student to show greater independence and to arrive at the office door with more specific problems in mind form discussion. The general strategy for successful scientists is to choose the most important problem that they may think they may be able to solve. There are many ways of finding a problem worthy of investigation. Sometimes a problem is more or less provided to an investigator, especially a thesis problem of a (senior) third/fourth year undergraduate or beginning postgraduate student. In other cases, particularly involving technological or other practical research, a funding source may encourage selection of one of the problems that the agency or foundation would like to have solved. Although having a funding source dictates a problem may sound queer, it is not necessarily so. For example, HELSB is funding academic 24 members of staff for their graduate schools at our higher learning institutions. L.S.B. Leakey Foundation on Baldwin fellowship grants students who can pursue master degrees on research based on human evolution (Physical Anthropology) in Africa. More commonly, problems arise from ideas, encouragements or interactions with other scientists. For example, many advisors may present their students with a list of possible problems or a problem area within which specific problems might be identified. Another source of ideas includes gurus of a discipline: seniour scientists who write review papers and books and so on. The other and probably, the richest source of ideas for problems come from the discipline in general. They include the fact that one can read technical journals, electronic mail communications, go to professional meetings where one can here oral papers and talk with other practitioners, listen to talks on campus and finally, discuss issues with peers and others on the home campus. No one is uninfluenced by what has transpired in a given field and what is going on currently. Thus, really there is no such thing as finding a research problem readily. It has to be noted that it should become the mind-set of every scientist to find as good a problem as she or he thinks solvable. In most cases, it has been asserted that the most creative aspect of science lies in asking a good question and identifying a good problem. One has to launch her or his career by asking good questions of nature right from the beginning. One has to try also to calibrate her or his own expectations and perceptions of a good research project with those of her or his advisor, committee members and other more experienced colleagues. An interesting research question might be solvable once a particular technique is mastered, validated and applied, but if it is going to take four or five years to develop the technique, it 25 may not be an appropriate problem for a student to tackle. Frequent consultations with the advisor and other teaching members of staff establish good communication skills that are necessary to the enterprise of science. Some few points that are helpful for selecting a research problem or a subject for research: a) The subject, which is overdone, should not be normally chosen because it will be a difficult task to throw any new light in such a case. b) A controversial subject should not become the choice of an average researcher. c) Too narrow or too vague problems should be avoided. d) The subject selected for research should be familiar and feasible so that related research materials or resources for research are within one’s reach. e) The importance of the subject, qualifications and training of a researcher, costs involved and time factor are few other criteria that must be considered in selecting a problem. Thus, one has to consider the following aspects: i) Whether or not the researcher is well equipped in terms of her or his background to carry out the research. ii) Whether or not the study falls within the budget she or he can afford. iii) Whether or not necessary cooperation can be obtained from those who must participate in research as subjects. iv) Selection of a problem must be preceded by a preliminary study. 2.2 Formulating a hypothesis/model The second major hurdle in planning research is to formulate potential explanations for phenomenon chosen for study. It is 26 usually as difficult to build a testable model as it is to identify a good problem. Yet, sometimes problems suggest potential solutions. There are many paths to developing a model for a chosen problem. The paths range from adopting a pre-existing model (for instance, from literature) to constructing one wholly from scratch. In most cases, some complicated path between these extremes is followed, amalgamating elements and aspects for various sources. One ordinarily adopts a model totally from literature for the explicit purpose of testing an idea that has interested researchers in the field but is rather neither well confirmed nor convincingly rejected. When the problem to be tackled is already well recognized but available models have proven inadequate, a new model is frequently developed by modification of a pre-existing one. In cases where a familiar problem has been broadened for revived study, pre-existing models may also be useful. One may be able to extrapolate from older, more limited models. On the other hand, when the problem to be solved is unfamiliar in literature, pre-existing models may still prove useful in developing a new model by analogy. Borrowing an idea from one field of study to use in another is quite common in science, in particular, where mathematical models are concerned. It has to be noted that it is not always possible to modify, extrapolate from or make an analogy with a pre-existing model. Thus, one must nurture other strategies in formulating a new model. One strategy is to discuss the problem to be solved with peers and more experienced scientists. Sometimes, other persons may have suggestions that can, with development, become the basis for a new model. Perhaps more frequently, questions of others can rekindle one’s thinking and elicit new ideas about 27 possible explanations for the phenomenon to be studied. Group discussions can be particularly useful in this regard. Another strategy for formulating a model is to generalize from a particular instance. For example, something observed in nature. It is suggested that in developing models, one should think broadly, formulate general classes of possible explanations and, wherever possible, develop two or more different models to account for the same phenomenon. 2.3 Devising testable predictions The last difficult aspect of planning research is to devise testable predictions from the model or models to be evaluated empirically. As with finding a problem and formulating models that might explain the chosen phenomenon, devising predictions can be aided by existing literature. The best straightforward way of devising a prediction is to adopt a standard testing method that has been used by many others to test similar models. In a similar vein, even if no standard methodology exists, it may be possible to adopt ideas in literature for applying to your particular model. For example, suppose the model states that younger monkeys are especially exploratory of new potential foods. One might begin to test this model through observations by recording for animals of different age on kinds of plants or animals they attempt to eat, predicting that the diversity of items will be larger in younger monkeys. Even if data are consistent with the prediction, they do not account for reasons younger monkeys have more diverse diets than older ones. Therefore, one may also devise an experiment based on previous works, for example, experiments conducted to Japanese macaques (monkeys), where primatologists provisioned macaque troops with unfamiliar foods. Here one would predict from the model that younger animals are most likely to be the first to try the new food provided. 28 In many cases, past literature will be of limited help and thus, one must devise testable predictions more or less freshly. The task is usually not as easy as it might seem at first trial. One approach is to scrutinize the model and follow any line of deductive reasoning that might lead to an observable result. For example, if the model states that butterflies meet at wet spots on the ground in order to obtain one or more critical nutrients, the prediction is that all such spots will contain at least one such nutrient – a prediction testable by chemical analyses of samples from the spots. Another approach to devising testable predictions is to make further observations while keeping in mind the model to be tested. New observations may suggest predictions that had not been evident through initial scrutiny of the model. Back to the example of butterflies, further observations in the environment might turn up wet spots at which butterflies do not gather. The model predicts that these spots should lack nutrients found in spots where insects gather. In this case, one could consider this a new prediction or, taken together with the first one, a stronger single prediction from the model. Whatever strategy used for devising predictions from the model to be tested, it is always good science to devise multiple predictions from the same model. Although results consistent with a model increase our confidence in that model, they cannot prove it to be true, so the more predictions that are tested, the more the confidence we can place in the model. There are inherent difficulties in testing predictions convincingly, but if several predictions are derived from the same model, chances of obtaining a good test of a model are obviously enhanced. 29 2.4 Implementing the plan The next Topic will offer a guide in preparing a proposal for funding the planned research. Of course after the research is completed, one must write up results (as an honour’s thesis, dissertation or for publication). 3. How to Write a Research Proposal Preamble Research proposals are of course required by funding agencies, but undergraduate and postgraduate students usually must submit a formal proposal of thesis/dissertation research. There are occasions upon which established scientists must write proposals. Research is a systematic collection, analysis and interpretation of data to answer a certain question or solve a problem. There are several characteristics of research that include the following: research demands a clear statement of the problem; research requires a plan (it is not aimlessly looking for something in the hopes that you will come across a solution); it builds on existing data using both positive and negative findings; and finally, new data should be collected as required and be organized in such a way that they answer the original research question(s). Explain what you want to do, how you will do it and why it is important. It is assumed that a disorganized or poorly written proposal reflects a disorganized or poorly conceptualized study. In due regard, organization, content and clear writing are essential aspects to success of a proposal, whatever its purpose. Even if no funding or prior approval is required, the thoroughness demanded by articulating your knowledge of a subject and your research intentions can help you to identify inconsistencies in logic and 30 inappropriate fits between questions you ask, data you intend to collect and your methods. Describing what you plan to study, how you will do so and why it is important is a vital, and in many cases mandatory, precursor to conducting scientific research. Successful scientific research proposals convey good salesmanship: you are selling a future product, your research idea and protocol, to critical audience that must select from among many such products. To sell your proposal effectively, you must know who your audience is, whether they are highly trained specialists in your field, established researchers in related fields, or board members seeking for research that is consistent with their foundation’s funding directives. Persuasive writing does not compensate for scientific competence, but many good ideas and research designs are undermined by explanations that are either too vague or too specific to the audience evaluating them. Format Funding agencies and university programmes may differ in their format and content requirements for scientific proposals. Before one begins to write, one should obtain all of available information and forms from all funding agencies and programmes to which one expects to submit the proposals. Books listing funding agencies can be found in reference sections of most university libraries. Also, some funding agencies have their web pages that give details for funding procedures including application for funding. Such guides to grants and foundations include addresses of funding sources as well as summaries about the programmes. You will need to write e-mail or call to request application materials, which usually take several weeks to arrive. Some other funding agencies, for example, National Geographic, requires applicants to apply electronically. Many professional societies publish guidelines and other ethical considerations. When you receive application materials from the 31 granting source, read the instructions carefully, paying particular attention to the following aspects: i) the number of copies you will need to submit; ii) page limits; iii) funding criteria and amounts available; iv) deadlines; and v) supporting documents. Supporting documents include things such as letters of recommendation, academic transcripts, evidence of collaboration and research permits. Whatever grant is being applied for, follow precisely all instructions including those for specific format. A very important set of instructions is that relating to length and format of sections, especially the heart of the proposal where you explain hypotheses to be tested and methods to be used. You have to note that reviewers are inevitably overworked persons who must read a number of proposals in a short period. Do everything possible to make your proposal easy to read. Criteria for evaluation Each funding agency has its own criteria for evaluating proposals. Scrutinizing such criteria before you write or adapt your proposal will help you to avoid off-target or inappropriate submissions. Some funding agencies provide written evaluations when they announce their results. Others will discuss reasons for a negative decision if they are asked. Still, others provide little or no feedback even on request. Criteria used by reviewers may change from submission to submission, so it is wise to verify what the current criteria will be. Many foundations and agencies make available to investigators printed information concerning how their proposals will be reviewed. 32 The topic will elucidate how to prepare and submit a research proposal. You have to note that there are differences between fieldand laboratory – based behavioural research, between observational and experimental research, and among kinds of questions and methods that can be applied to particular topics/problems. Be sure to take these differences into account when using these suggestions to guide your own proposal. 3.1 Significance of a title The title of your proposal should state precisely what your study is about. Many first-time proposal writers underestimate the importance of an informative title. If you cannot state concisely what you intend to study, you may need to clarify your thinking on the subject. One of the most common problems with titles is that they claim to cover more than the research will actually achieve. For example, a proposal titled “The Effects of Nutrition on Baboon Reproduction” implies that nutritional intake and reproductive condition will be measured and correlated. Suppose the study actually involved use of feeding observations without corresponding nutritional analyses and focused exclusively on female reproductive status. A more appropriate title to the research would then be “The Effects of Food (or Diet) on Reproduction in Female Baboons.” However, at the same time, a title should not be so narrowly defined that it fails the broader scientific context in which the research is situated. Correlations between variables other than diet and reproduction such as dominance rank or age might constitute equally important questions in the proposed research. Note the different emphasis and what each implies about research goals in the following possible titles: - The Effects of Diet on Age, Rank and Reproduction in Female Baboons 33 - The Effects of Reproduction on Diet, Age and Rank in Female Baboons - The Relationships Among Diet, Rank and Reproduction in Female Baboons Proposals rarely fail solely due to an inappropriate or misleading title. But like all first impressions, titles set up expectations for what will follow. An accurate and informative title will help ensure that your proposal is sent to appropriate reviewers and that they will not need to readjust their expectations once they begin to read your proposal. EQUALLY IMPORTANT ASPECTS A) Background Information In this section, one presents a general background to the problem. That should be well written such that it enlightens people about the problem that needs to be researched. It builds upon the foundation of the proposal for the subsequent sections. That means if not well presented, the next section, Statement of the Problem cannot be formulated. Likewise, Research Objectives, Literature Review, Methodology and Work Plan are built upon this section such that if improperly presented, the proposal will be bad. B) Statement of the Problem This is another major section in a research proposal. Why is it important to state and define the problem well? Because you will find out that a clear statement of the problem: - Is the foundation for further development of the research proposal (research objectives, methodology, work plan, budget and so on); 34 - Makes it easier to find information and reports of similar studies from which your own study design can benefit; and - Enables you to systematically point out why the proposed research on the problem should be undertaken and what you hope to achieve with the study results. This is important to highlight when you present your project to community members, relevant ministry and donor agencies that may need to support your study or give their consent. Information that should be included in the statement of the problem: 1. A brief description, depending on the study, say, of socioeconomic and cultural characteristics and an overview of the status in the country or district in as far as these are relevant to the problem. Include a few illustrative statistics, if available to help describe the context in which the problem occurs. 2. A concise description of the nature of the problem (the discrepancy between what is and what should be) and of its size, distribution, and severity (if it is a health problem, say, who is affected, where, since when and what are the consequences for those affected and for the services?)/ 3. An analysis of major factors that may influence the problem and a convincing argument that available knowledge is insufficient to solve it. 4. A brief description of any solutions that have been tried in the past, how well they have worked and why further research is needed. 5. A description of the type of information expected to result from the project and how this information will be used to help solve the problem. 6. If necessary, a short list of definitions of crucial concepts used in the statement of the problem. 35 A list of abbreviations may have to be annexed to the proposal, but each abbreviation also has to be written out in full when introduced in the text for the first time. 3.2 Identifying objectives The purpose of research is to discover answers to questions through application of scientific procedures. The main aim of research is to find out the truth, which is hidden and that has not been discovered. Though each research has its own specific purpose, research objectives fall into a number of the following broad groups: 1. to gain familiarity with a phenomenon or to achieve new insights into it (studies with this objective in view are termed exploratory or formulative research studies); 2. to portray accurately characteristics of a particular individual, situation or a group (studies with this objective in view are known as descriptive research studies); 3. to determine the frequency with which something occurs or with which it is associated with something else (studies with this objective in view are known as diagnostic research studies); and 4. to test a hypothesis of a causal relationship between variables (such studies are known as hypothesis-testing research studies). Thus, usually research objectives are meant to examine, expand, investigate, explore, develop or evaluate a set of data relevant to a set of questions that inform the overall research. Objectives should be concise statements that provide enough detail to communicate the scientific focus of the study. In defining your objectives, it is helpful to think in terms of three or four broad aims. The aims may be parallel or an organized gradient from specific to general. For example, a study on “Sex Differences in Territorial Behaviour in a 36 Polgynous Bird (Species)” may have three objectives. They include the following: i) to evaluate the presence and degree of sex differences in territorial defense; ii) to explore the reproductive correlates of territorial defense for males and females; and iii) to develop a model of dynamics of polygyny. Note that you have to elaborate on each of your objectives in the body of your proposal. Thus, when formulating your objectives, as in the case of title, be careful that you do not set yourself an objective that your research cannot address. Objectives, like titles, may require fine-tuning as you develop the body of your proposal. The following are further main points for consideration while formulating research objectives: OBJECTIVES OF A RESEARCH PROJECT SUMMARIZE WHAT IS TO BE ACHIEVED BY THE STUDY. In due regard, objectives should be closely related to statement of the problem. For example, if the problem is low utilization of child welfare clinics, the general objective of the study could be to identify reasons for the low utilization to find solutions. The general objective of a study states what is expected to be achieved by the study in general terms. An important aspect to note is that it is possible (and of course, advisable) to break down a general objective into smaller, logically connected parts. These are normally referred to as specific objectives. Specific objectives should systematically address various aspects of the problem as defined under “Statement of the Problem” and key factors that are assumed to influence or cause the problem. They should specify what you will do in your study, where and for what purpose. 37 The general objective “to identify reasons for low utilization of child welfare clinics in Manyoni district to find solutions” can be broken down into the following specific objectives: 1. To determine the level of utilization of child welfare clinics in Manyoni district over years 1999 and 2000 as compared with the target set; 2. to identify whether there are variations in utilization of child welfare clinics related to reason, type of clinic and type of children served; 3. to identify factors related to child welfare services offered that make them either attractive or not attractive to mothers (this objective may be divided into smaller sub-objectives focusing on distance between the home and clinic, acceptability of services to mothers, quality of services and so on); 4. to identify socio-economic and cultural factors that may influence the mothers’ utilization of services (Also this may be broken down into several sub-objectives); 5. to make recommendations to all parties concerned (managers, health staff and mothers) concerning what changes should be made and how to improve use of child welfare clinics; and 6. to work with all parties concerned to develop a plan for implementing the recommendations. Remarks: The first specific objective focuses on quantifying the problem. This is necessary in many studies. Often use can be made of available statistics or of health information system. Specific objective two further specifies the problem, looking at distribution. Specific objectives three and four examine possible factors that may influence the problem and objectives five and six indicate how results will be used. 38 An objective focusing on how results will be used should be included in every applied research study. Thus, formulation of objectives will help you to: - focus the study (narrowing it down to essentials); - avoid collection of data that are not strictly necessary for understanding and solving the problem you identified; and - organize the study in clearly defined parts or phases. Properly formulated specific objectives will facilitate development of your research methodology and will help to orient the collection, analysis, interpretation and utilization of data. How should you state your objectives? Take care in such a way that objectives of your study: - cover the different aspects of the problem and its contributing factors in a coherent way as well as in a logical sequence; - are clearly phrased in operational terms, specifying what you are going to do, where and for what purpose; - are realistic considering local conditions; and - use action verbs that are specific enough to be evaluated. Examples of action verbs are: to determine, to compare, to verify, to calculate, to describe and to establish. Avoid use of non-action verbs such as: to appreciate, to understand or to study. 39 Keep in mind that when the project is evaluated, results will be compared to objectives. If objectives have not been spelt out clearly, the project cannot be evaluated. Using the example on utilization of child welfare clinics, we may develop more specific objectives such as the following: . To compare the level of utilization of the child welfare clinic services among various socio-economic groups; . To establish the pattern of utilization of child welfare clinic services in various seasons of the year; . To verify whether increasing distance between the home and the health facility reduces the level of utilization of the child welfare clinic services; and . To describe mothers’ perceptions of the quality of services provided at the child welfare clinics. 3.3 Literature review The background to your proposal provides formal scientific context from which your study is derived and to which your research will ultimately contribute. It is a section in which you review what is already known and what the outstanding questions in your study are. In essence, it is a formal review that should outline, usually in a third person, what stimulated your interest in your proposed research. The section cannot be written until one has a thorough command of literature in his/her field and has identified existing questions or gaps in this literature. How you focus this section – what you choose to include or omit – will depend on what your study proposes to accomplish. You have to relate your stated objectives to existing knowledge. If one of your objectives is to evaluate an existing model with data, you will need to review the model and provide evidence 40 for why such a test of the model is important. Exceptions that challenge the model of your study subject are examples of how you may be able to situate your research contribution within a broader context. It is important to recognize that your preparatory reading is likely to be much more extensive than what you will have space to review in the background section of your proposal. It is also the case that not all of your prior reading will be equally relevant to your proposed research. Choose your examples and citations carefully being sure to indicate whether your list is inclusive or selective. Profiling an example as being “… the sole exception …” when there are others that you fail to acknowledge will raise doubts in reviewers’ minds about your ability to integrate your proposed work with existing knowledge. The background section should demonstrate that you have a clear idea of what is new about your proposed research. Inappropriate claims about your own originality may offend reviewers who have done similar work or know of other work in the area, undermining the credibility of your proposal and assessments of your ability to interpret your data if you are given the opportunity to collect them. Then, there is the following question, why is it important to review already available information when preparing a proposal? The following reasons help to give an understanding for literature review: i) It prevents you from duplicating work that has been done before. ii) It helps to find out what others have learned and reported on the problem you want to study. This may assist you in refining your statement of the problem. 41 iii) iv) It helps you to become more familiar with various types of methodology that might be used in your study. It should provide you with convincing arguments for reasons your particular research project is needed. Again, possible sources of information can be obtained from the following: i) Individuals, groups and organizations; ii) Published information (books, articles, indexes and abstract journals); and iii) Unpublished information (other research proposals in related fields, reports, records and computer data base). There are several places to find these such as the following: i) Community and district or regional levels; ii) National level; and iii) International level. The following are strategies that may help one to gain access to each source and obtain information in the most productive manner: i) Identifying a key person (a research or decision-maker) who is knowledgeable on the topic and asking if he/she can give you a few good references or names of other people whom you could contact for further information; ii) Looking up for names of speakers on your topic at conferences who may be useful to contact; iii) Contacting librarians, research institutions, responsible ministry and newspaper offices and requesting for relevant references; iv) Examining bibliographies and reference lists in key papers as well as books to identify relevant references; v) Looking up for references in indexes (for example, Index Medicus for Health Systems Research and so on) and abstract journals; and 42 vi) Requesting for a computerized literature search (for example, Medline, Google and the like). Note: You may get assistance from agencies for your literature search through telephone or in writing. However, the request should be specific. A computerized search should be used through key words that can be used in locating relevant references. What can you do after identifying the references? References that are identified: - should first be skimmed or read. - Then, summaries of important information in each of references should be recorded on separate index cards or as computer entries. These should then be classified so that information can easily be retrieved. - Finally, a literature review should be written. The following are several steps that you should take when preparing a review of available literature and information: - First, organize your index cards in groups of related statements according to which aspect of the problem they touch upon. - Second, decide in which order you want to discuss various issues. If you discover you have not yet found literature or information on some aspects of your problem that you suspect are important, make special efforts to find that literature. - Third, write a coherent discussion of one or two pages in your own words, using all relevant references. 43 Possible bias Bias in literature or in literature review is a distortion of available information in such a way that it reflects opinions or conclusions that do not represent the real situation. The following common types of bias in literature review have to be avoided: i) Playing down controversies and differences in one’s own study results; ii) Restricting references to those of which support the point of view of the author; and iii) Drawing far reaching conclusions from preliminary or shaky research results or making sweeping generalizations from just a case or small study. Ethical considerations The presented forms of bias would put the scientific integrity of the responsible researcher in question. Moreover, careless presentation and interpretation of data may put readers who want to use the study’s findings on the wrong track. That may have serious consequences in terms of time and money spent on the research. A similarly serious act, for which a researcher can be taken to court, is presentation of research results or scientific publications from other writers without quoting the author(s). Therefore, appropriate referencing procedures should always be followed in research proposals as well as in research reports. 44 3.4 Defining a theoretical framework Etymology of the word theory The word ‘theory’ derives from the Greek ‘theorein,’ which means ‘to look at.’ According to some sources, it was used frequently in terms of ‘looking at’ a theatre stage, which may explain why sometimes the word ‘theory’ is used as something provisional or not completely resembling real. The term ‘theoria’ (a noun) was already used by scholars of ancient Greece. To a scientist, theories are statements about relationships between abstract concepts or variables. A simple theory might specify how only two or three variables are related. Thus, a theory designates a well-tested model, one in which the scientific community places a great deal of trust because it has survived repeated attempts to prove it false. LOGICAL STRUCTURE, THEORETICAL FRAMEWORK Problems do not exist in nature but in the minds of people. This can be seen from an examination of the definition of problem: problems stem from the juxtaposition of factors, which results in a perplexing or enigmatic state of mind (a cognitive problem), an undesirable consequence (a psychological or value problem), or a conflict, which obscures the appropriate course of action (a practical problem). Cognitions, values and practices are attributes of persons, not the objective world (whatever that is). Problems cannot be articulated except within a conceptual system. No inquirer can investigate a problem from all perspectives simultaneously. Also that is what a logical structure or theoretical framework is all about. It establishes a vantage point, a perspective, a set of lenses through which the researcher views the 45 problem. In this sense, selection of a logical framework is both a clarifying and exclusionary step in the research process. While it sharpens focus and consequently increases clarity brought to the problem area, it excludes from the view of the inquirer other perspectives that might be brought to bear on the problem, but does so in explicit recognition of those perspectives and the rationale for their rejection. In fact, it is the choice of frameworks chosen by the researcher that has contributed to new understandings or problem solutions by some researchers or to inadequate inquiry or false conclusions by others. For example, decades of research on organizational management and behavior viewed organizations from the classic, rational model of hierarchical bureaucracy with tightly coupled substructures and linked and linear organizational processes (as posited by the German sociologist Max Weber in the early part of the last century). But that perspective never led to adequate understandings of how organizations, such as corporations and universities, actually work. Recent researchers, working from the vantage point of alternative perspectives, using metaphors derived from long-term observation of life in universities and other organizations, have broken away from the image of organizations as bureaucracies to study them as "organized anarchies" and "loosely coupled systems." They have created a powerful new line of inquiry that has greatly enhanced our understanding of the structures and processes of work life in public institutions. Thus, there are usually multiple frameworks from which to view the same problem, the more viable often being obscured by dominance of a worn-out paradigm that blinds the observer to alternative views of the world. The framework used by the researcher is not always explicit (as in the example of "organized anarchy" when first used as a perspective for studying organizations); but the burden of the 46 argument is that to the extent possible the framework should be explicated for several reasons: 1. Since the problem is a function of its framework, the problem can be better articulated and understood if its basic system is well understood and articulated. Additional facets of the problem may be generated as a result, and the known facets will take on greater clarity and form. 2. When the framework is well articulated, it is possible to conceive and consider alternative frameworks. The explication of behaviorist theory in early psychology made it possible to see what its strengths and weaknesses were and to develop alternative theories that ultimately had high payoff (for example, the advent of cognitive psychology and one of its offspring, Rational Emotive Therapy). Given several possible frameworks, the researcher chooses from among them on the basis of criteria such as heuristic value, inclusiveness, efficiency, and the like. The power of a proposed solution to the problem may thus be considerably enhanced. 3. The explication of a theoretical framework or logical structure provides focus to all the subsequent steps in planning and carrying out the proposed inquiry, for example, charting variables and their relationships. It makes it possible to generate a relatively complex set of objectives and questions; it provides a basis for including and excluding literature and research that is actually related to the inquiry by identifying the variables of greatest interest and concern; and it provides focus to the inquirer's procedural planning and choices from initial design selection, through instrument development or adoption, to the organization, analysis and interpretation of data, for example, research design, statistical tests, making sense of empirical findings. 4. Perhaps most important is the impact of the explicit theoretical structure on subsequent inquiry in the same area. The investigation 47 no longer hangs loose but becomes part of a line or tradition of inquiry, which other researchers can check, replicate or build upon. Knowledge growth in a field becomes an additive phenomenon of increasingly useful structures or concepts with which inquirers can work. 5. Without a clear explication of the problem and a workable perspective with which to view it, it is likely that the research project will be flawed by uncontrolled extraneous variables, overlooked variables, faulty instruments, haphazard procedures and the like. "You cannot get there from here" without taking this step. A failure in this regard is why so many graduate students end up with a procedural plan that runs them in circles. Defining a Logical Structure or Theoretical Framework A logical structure or theoretical framework is the set of terms and relationships within which the problem is formulated and solved. Such frameworks may vary greatly in format and sophistication. In its simplest form, a conceptual framework may be no more than a set of descriptive categories. For example, one may decide to investigate teacher behavior by noting whether a teacher's verbal statements are questions, informational comments, supportive comments or disciplinary comments. Such a set of terms would be quite useful in categorizing behavior even though there is no pretense that all behavior could be categorized this way, or that the terms were pre-selected to conform to some particular point of view. Maslow's Hierarchy of Need is another example of a conceptual framework that has been heavily used in social science research, including efforts to refute its utility as a classification scheme for human behavior. When such a set of categories meets the additional criteria that all categories are independent of each other and are (together) necessary and sufficient to encompass all relevant phenomenon, 48 they may be said to comprise a taxonomy, as, for example, the biological taxonomy of life forms. A theory interconnects categories (whether or not they form a taxonomy) through a set of relationships. Both categories and/or relationships may be derived from a basic set of postulates. Hypotheses may be derived by deduction from the theory for testing. In short, a conceptual framework is a concise description (often accompanied by a graphic or visual depiction) of the major variables operating within the arena of the problem to be pursued together with the researcher's overarching view of how variables interact (or could be made to interact under manipulated conditions) to produce a more powerful or comprehensive "model" of relevant phenomena than has heretofore been available for shedding light on the problem. Think of it as a MAP with conceptual directions. The framework, in fact, either anticipates or directly presents the basic design of the study. The figures at the end of this section attempt to depict example structures for two hypothetical studies. Figure 1 is a "lick the world" structure examining the effectiveness of an in-service training program for teachers that would take a well funded staff of researchers to complete if left as it is. Figure 2 is a model for investigating factors that shed light on the problem of antisocial behavior among youth, and is considerably less complicated than Figure 1, although it poses some formidable measurement challenges. A little narrative has been put with those figures to aid comprehension. It might be helpful to examine these figures and then re-read this section a few times. 49 Functions of a Logical Structure or Theoretical Framework 1. Expounding - To explain the structure or framework within which the situation will be investigated, that is: (1) in case of the logical structure, to provide a rationale for the perspective from which the investigator will examine the problem; or (2) in case of the theoretical framework, to conceptualize or state the theory in which terms the investigator will examine the problem. 2. Validating - to validate application of the particular logical structure or theoretical framework in the investigation of the problem in terms of its anticipated advantages and consequences. The process of structure building is the researcher's creative step in research design that minimizes irrelevancies, tightens focus on constructs that comprise the relevant substance of the inquiry, and maximizes the "real world" utility of the inquiry; that is, it is the means by which validity in its various forms is achieved. Therefore, structure or framework embodies the ontological and epistemological character of the study and anchors the methodological phases of inquiry (sampling procedures, choice of research questions, statistical design for each question, hypothesis and so on). The following are common deficiencies in proposal structures -- Failure to offer any framework - raw empiricism -- The inappropriate framework -- The overly complex framework -- The framework unrelated to other competing structures -- The imprecise framework 50 Generating a Framework or Structure Many researchers find themselves perplexed by the notion that they are responsible for positing a theoretical structure on which their inquiry is to be based, or they are just perplexed. Consequently, no framework is offered at all or they end up as overly complex, sophisticated structures and statements, which the researcher finds dysfunctional to the conduct of the inquiry. The framework, in effect, becomes an independent step in the inquiry process carried to a successful conclusion as an academic exercise. But it misses the central point of the activity. The theoretical or conceptual grounding of a study is designed to help the inquirer -not boggle his/her mind. It is undertaken not simply for the advantage of the reader of a proposal, but for the researcher as a conceptual map to the investigation. As was noted, a conceptual framework is the necessary concomitant of any problem, for the problem could not be stated except within at least an implicit framework. Thus, the task is simply to make explicit what is already there at the implicit level in the statement of the problem. The researcher may begin by simply noting key terms and basic assumptions underlying the inquiry (variables and their interactions). Several structures, classification systems, taxonomies, and theories may already have been explicated precisely in the field in which the researcher is working; or, in other cases, in fields that could be applied. For example, use of the social psychological structure of interaction analysis in observation of classroom teacher behavior (that is, Flanders' interaction analysis); or the notion of political cultures borrowed from political science, applied to a study of state level social service policy development (that is, the legislative or regulatory decision process). 51 Sources to support the inquirer in theory development are no different from those will be turned to for support in other areas of planning the research, that is: -- Extant structures of varying levels of sophistication in the literature of the field. -- Structures from related fields that could be adopted for, or adapted to, the inquiry. -- Previous research studies that have employed either implicit or explicit structures pertinent to the inquiry. -- De Novo explication of structures -- these usually occur at simple rather than complex levels of structure building or the research itself must be geared first to a theory development project. Investigations frequently employ not one but several structures to clarify dimensions of the inquiry. For example, a "futures" study may posit one structure to support the substance of the inquiry, a second to organize the futures view, and perhaps a third to clarify a methodological orientation toward data collection (probably detailed in the procedures section). The critical point is that multiple structures should be truly complementary and exclusive in their orientation. If they overlap, an effort should be made to synthesize the structures and postulate a new, more comprehensive structure. It is virtually always beneficial to create a drawing or figure at early stages of conceptualization that depicts the major variables and categories, connected by lines and arrows to show relationships and interactions, in much the same way that an architect or designer might make preliminary sketches of a building or landscape. "How can I know what I think until I see what I say?" The narrative for the conceptual framework describes 52 and justifies the elements of the figure. The figure serves two important purposes. First, it helps the researcher think about his/her own thinking but also aides in the preparation of the narrative for the conceptual framework. Second, when included in the proposal, it can be helpful, even essential, to the reader who seeks to understand what the researcher is trying to do. Although structures sometimes get needlessly complicated or overly sophisticated (thus, obligating the researcher to an unrealistic project), an essential step in structure building is to PURPOSELY complicate and make more comprehensive the initial structure so that the scope of the inquiry can be examined for missing categories or inappropriate causal constructs; that is, to make sure the bases are covered with regard to issues of validity and utility. The researcher is then in a better position to make appropriate limiting and delimiting choices that shrink the endeavor back down to manageable parameters and at the same time ensure his/her efforts will be fruitful. Once a framework has been prepared, it is important to ask what advantages and disadvantages may accrue as a result of using it. In the event that there seems to be available only a single alternative framework, its use is mandated even though it may have some obvious drawbacks. In other cases where multiple frameworks may be available, as for example, in learning research (behaviorism against constructivism), the choice of the particular framework should certainly be made to maximize those advantages that are most salient for the investigation or development project and to minimize those disadvantages that are most inimical to it. In general, when one who proposes is able to demonstrate that first, the proposed framework has relevance to the study, and second, the particular framework has more advantages and/or fewer disadvantages than some other framework that might have been used, then the validating function of this section of the proposal has been met. 53 *Additional Readings for the Section Brody, H.S., Ennis, R. H. & Krimerman, L.I. (1973). Philosophy of Educational Research. New York: John Wiley & Sons. Chap. 5, pp. 271-79; Chapter 8. Kaplan, A. (1964). The Conduct of Inquiry. San Francisco: Chandler. Chapters 7-8. Kerlinger, Fred (1986). Foundations of Behavioral Research (3rd. ed.). New York: Holt, Rinehart & Winston. Chapters 2-3. Rudestam. K. & Newton, R. (1992). Surviving Your Dissertation: A Comprehensive Guide to Content and Process. Newberryi Park, CA: Sage. Chapter 2. Remarks: In case of qualitative studies, a theoretical framework may not be explicitly articulated since qualitative inquiry typically is often oriented toward grounded theory development in the first place; the exception is a study that tests or makes more comprehensive an existing theory. If the study is about theory development, the researcher is still not off the hook. Some kind of overarching conceptual framework is still necessary and will test the writer's creative use of existing, related theory, intuition and tacit knowledge. Figure 1 Effectiveness of in-service training for teachers 54 Figure 2. A model of antisocial behavior 55 Examples of Theoretical Frameworks Note: These are not meant to be models to which you should necessarily aspire. They are just examples. Some are overly complex and needlessly verbose. Nevertheless, the collection should provide you with a good idea of how structures are presented and what they accomplish. The last one is from the first dissertation completed at Arkansas State University. Example 1 School or Library Staff Reactions to Challenges to School or Library Materials Hall and Fagen (1956) define a social system as a bounded set of elements (sub-systems) and activities in mutual interaction that constitute a single social entity. A given social system must have distinct boundaries, which separate it from the environment, defined as anything outside the system that can affect and is affected by the social system. Social systems may be open or closed, depending on their interaction with the environment. Social systems strive to maintain homeostasis and equilibrium, using internal and external feedback loops to identify imbalances. Public schools can be defined as open social systems. The GetzelsGuba Model (Hoy and Miskel, 1978) identifies their structural elements as: Institutional -- the organization's formal bureaucratic structure, encompassing hierarchy or authority; policies, rules and procedures; and specialization. Individual -- the varying behavior of individuals in similar situations. Informal Group -- the formation of groups by individuals to balance bureaucratic expectations and individual needs. Groups 56 also establish an organizational climate that will affect individual behavior. Internal Feedback Loop -- tells individuals how their behavior is viewed by the organization's formal and informal structure and can also influence individual behavior. External Feedback Loop -- provides an opportunity for input from the environment which may influence bureaucratic expectations and group intentions. As pointed out by Tumin (l977), the heterogeneous nature of social organizations and their environments lend themselves to conflict -to situations where values, goals, interests, and ideologies are in disagreement, result in confrontation, and must be resolved. The conflict under consideration in this study is the challenge, made by some person or persons from within the environment, to materials used in or acquired by a school or school library with a written collection development policy. Thus, bounded systems to be studied are schools or school libraries, which have written collection development policies and which have had at least one challenge to their materials from someone within their environments, (that is, communities, which they serve). Adapting the Getzels-Guba Model, the independent variables in the study can be grouped into Formal, Informal, and External groups. Staff is defined as teachers, librarians, administrators, and school boards. Formal (Institutional) Factors As previously stated, there have been guidelines established by the national professional associations for writing of policy statements and procedures for handling complaints. Serebnick (l979), in A Review of Research Related to Censorship in Libraries, found that research studies accepted that there was a distinction between selection and censorship and that the Library Bill of Rights was 57 accepted as the librarians' official statement on professional guidelines for selecting materials. The roles that public library and school boards of education can play in initiating or supporting censorial action in libraries and schools are critical. "Decision in Tulsa: An Issue of Censorship" and "Innocence Lost: One Librarian's Experience," two journal articles recapping the successful stand taken by two public libraries against a challenge, report that the library boards stood behind the rights of freedom of expression guaranteed by the First Amendment. Conversely, the heated and sometimes violent censorship fights in Warsaw, Indiana, and Kanawha County, West Virginia, were precipitated by members of the school boards (Jenkinson, l979). The two public library case reports also indicated that the libraries had written collection policies, complete with procedures to handle complaints. All members of staff were aware of the policy and knew exactly what to do when the complaint was received. In an opinionnaire survey of 900 Midwestern librarians' attitudes toward intellectual freedom, censorship, and certain authoritarian beliefs, Busha (l971) found a significant relationship between the position held by a librarian and his/her attitude toward censorship. In this survey, which had a 69.3% final response rate, library directors obtained the highest mean censorship score. A study by Pope (l973), also an opinionnaire survey, of public, school, and college librarians' attitudes toward sexually oriented literature, with a 59.8% final response rate, substantiated Busha's conclusion that there appears to be a relationship between administrative responsibility of the librarians and their tendency to restrict materials. Jenkinson (l979) indicates two possible reasons for the vulnerability of schools to the attacks of censors -- 1) the lack of effective communication between schools, students, parents, and 58 other members of the community, and 2) quick acceptance of complaints by school personnel who ignore the established procedures for handling complaints. Generalizing from the above, the independent variables grouped as Formal will be:       the content of the policy and procedures the extent to which current staff participated in the development of the policy the extent to which current staff were made aware of the policy and procedures the extent to which established procedures were followed the formal approval of the policy at all hierarchical levels the extent to which there is a parent/community information sharing policy Informal (Individual, Informal Group, Internal Feedback Loop) Factors One of the most consistent research findings in library censorship studies is the willingness of the librarian to perform voluntary censorship of materials, also known as self-censorship. The Fiske (l959) study of book selection and censorship in California public and school libraries in the l950's is considered the most influential research on censorship in United States libraries. Twenty-six California communities were chosen based on their size, rate of growth, ethnic composition, geographic location, and type of library service. 204 municipal, county, and school librarians and administrators in 48 municipal and county units and 46 senior high schools in the 26 communities were interviewed using a structured questionnaire. Among other things, Fiske found that: 59 --although a majority of librarians stated that they believed in the freedom to read, large numbers of those same librarians restricted or condoned restricting materials. --the qualities of the institution in which the librarian worked had more impact on the librarians' attitudes and behavior toward controversial issues than did any characteristics of the community in which the institution was located. --although librarians felt few direct pressures for censorship and they gave verbal allegiance to the idea of intellectual freedom "because they sense a kind of free-floating anxiety in the air about them, they are reluctant to put their concepts to the test of practice" (pp. 73-74). Busha (l972, p.300) also found that although librarians "did not hesitate to agree with cliches of intellectual freedom, many of them apparently did not feel strong enough as professionals to assert those principles in the face of real or anticipated censorship pressures." In 1979 Woods and Salvatore studied the practice of selfcensorship in the nation's secondary school library media centers. A one-page questionnaire was sent to a select sample of high school librarians in the United States asking them to identify, from a list of controversial titles, the status of those titles held by their library media centers. One of their conclusions, based on the presence of large numbers of materials which had some sort of restricted access, was that many librarians are reluctant to face censorship battles. The survey had an initial response rate of 76.6%, but only 66.7% were usable returns. No attempt was made to follow up on non-respondents. Even so, the finding is consistent with Fiske and Busha such that it indicates the continuing relevance of the issue. 60 Generalizing from the above, the independent variables grouped as Informal will be the non-official behaviors of current staff at all hierarchical levels in terms of:  awareness of the policy and procedures  informal approval of the policy (i.e., attitude toward intellectual freedom)  peer pressure toward accepting or rejecting the policy  willingness to follow established procedures  willingness to be involved in a controversial situation  willingness to implement and support a parent/community information sharing program External Factors In a comparison of the censorship controversies in Warsaw, Indiana, and Kanawha County, West Virginia, Jenkinson (l979) indicates that both were fanned by outside agencies that supplied reviews of the school materials, had proponents and opponents, and formed local groups of concerned citizens against textbooks. Similarly, Serebnick's (l979) review includes references to a report by Wallis about serious censorship cases which were initiated and prolonged by restrictive mayors who sought and achieved removal and restriction of controversial publications. Serebnick also references a report by Good, whose analysis of additional censorship cases indicated that ministers, chiefs of police, district attorneys, newspaper editors, and members of parent-teacher associations had also exerted noticeable influence to provoke or combat censorship actions. Another important external factor is publicity. As Jenkinson (l979, p. 161) indicates: In far too many incidents of censorship, school officials have attempted to `keep the matter quiet.' That strategy may well backfire. Censors are willing to exert a great deal of pressure 61 because they know that some school officials will respond favorably in fear that they might otherwise receive negative publicity. On the other hand, few censors are willing to be exposed . . . And in many instances the general public would oppose the censors if they only knew what the censors were after. Fiske (l959), likewise, found that respondents believed the press could play a decisive role in the outcome of a censorship incident, although the role rarely could be predicted. A fast growing area of external influence are court rulings regarding censorship incidents. David Alexander (l980), writing for the National Organization on Legal Problems of Education three of seven federal court decisions reached during l972-l979 in the Second Circuit upheld school board decisions to remove materials from classrooms, libraries, or the curriculum. In three other federal cases, however, the First and Sixth Circuit courts held that First Amendment rights to freedom of speech necessarily implied a "freedom to receive" information and that removal of books once available on the grounds that their content was unacceptable abridged that freedom. Serebnick (l979) indicated that only minimal attention had been given to many community demographic variables with the exception of size of population served. Generalizing from the above, the independent variables in the External group will be the influence of:       publicity generated about the incident pressure groups support groups community leaders judicial rulings/legislation prior censorship incidents in the community 62 Since little research has been done on community demographic variables, the political/social climate of the community, also an external variable, will be used as a descriptor variable. Communities will be categorized as either conservative, moderate, or liberal. The dependent variable is the school or library's reaction to the challenge. Reactions may range from immediate support of the challenge at the initial point it was made to overruling of the challenge at the highest hierarchical level within the system. The study will consider only schools or libraries which were successful in having the challenge overruled. Example 2 Adult Learners Koffka (l953) defined the environment not as it is, but as it is perceived and experienced. This phenomenological approach was expanded by Lewin's (l936) field theory of life space. Lewin defined behavior not as a function of the objective physical properties of the stimulus environment, but as environment transformed into an "inner world" by a cognizing organism. Thus it is the psychological environment rather than the physical environment that determines the way an individual will respond. Of the existing theories of person-environment interaction, some focus primarily on the subjective frame of reference. Others are less phenomenologically oriented. Barker's (l968) behavior settings theory and the subculture approach suggest that environments select and shape the behavior of people who inhabit them. Both are primarily concerned with describing the environment and fail to include a theoretical concept or operational definition of the individual. Astin and Holland (l96l) suggest that the dominant features of an environment are dependent upon the typical characteristics of its members. Although they view behavior as a 63 function of the person and the environment, their theory emphasizes the person and neglects the perceptions of the environment. The need/press/culture theory (Pace & Stern, l958) analyzes the person and the environment in terms of personality needs and perceptions of the environment. This theory is based on that of Lewin (l936) who contended that scientific psychology must take into account the "whole situation," defined as the state of both the person and the environment. Stern (l970) also has defined the important concepts of Murray's (l938) need/press model. In the broadest sense, the term "need" refers to denotable characteristics of individuals, including drives, motives, goals, etc. A need state is characterized by the tendency to perform actions of a certain kind. Environment is defined in terms of "presses." Press is defined as the characteristic demands or features of the environment as perceived by those who live in it. Thus the environment is defined as it is collectively perceived and reported by its participants. The frame of reference of this theory will support the procedures used in this study. In order to measure student perceptions of college and university environments, Pace and Stern initially developed the College Characteristics Index (CCI). The CCI identifies environmental perceptions according to 30 scales (300 items). The content of the items is composed of the descriptions of activities, policies, procedures, attitudes, and impressions that may characterize college environments. In his research efforts, Pace became increasingly concerned about the absence of crucial, personal information which would truly distinguish one college form another. After several years of work, he developed the College and University Environment Scales (CUES). The CUES is a direct outgrowth of the CCI. Using CCI scores from students in a "norm group" of 50 institutions, the 300 64 items were reduced by restructuring and selective elimination to 150 items and reorganized into five scales. The CUES measures perception along five (5) dimensions: practicality; community; awareness; propriety; and scholarship. Not only is the CUES less cumbersome for reporting purposes, it also provides a more parsimonious presentation of differences in college environments (Michael & Boyer, l965). This instrument, therefore, will be used in the present study. Adult Learners Existing studies concerning life-long learning offer various definitions of the "adult learner." Most studies deal with students enrolled in continuing education programs. In this study, adult learners are defined as those persons 23 years of age or older enrolled in postsecondary education and registered for seven (7) or more credit hours. Across the institutions and within them, adult learners compose a heterogeneous mixture of individuals. Different kinds of adults seek out different types of educational experiences. In consolidating data from four (4) reports for College Board, Arbeiter (l977) used six (6) demographic variables to indicate trends among these adult populations: age; sex; race; socioeconomic background; educational level; and rural/urban setting. In this study, these six (6) demographic variables will be used. One additional factor -- academic major -- will be considered. Institutional Type Colleges and universities differ greatly from one another in many measurable characteristics. In assessing factors related to the impact of higher education on student development, institutional type continues to be an important variable. Various conventional classification systems have been used. In a study of 935 colleges and universities, Rogoff (l957) first divided them by curricular organization and then by type of control. Other classifications have 65 been made in terms of geographic location and the highest degree conferred by the institution. Although such nominal classifications are useful and at times unavoidable, they are not completely satisfactory as devices either to categorize colleges or to measure their environments (Barton, l96l). Other investigators have sought to identify categories of colleges and universities that would be relatively homogeneous with respect to the functions of the institution, as well as with respect to characteristics of students and faculty members. In l970, the Carnegie Commission on Higher Education developed such a classification system. The system divides institutions into five (5) main categories: (l) doctoral-granting institutions; (2) comprehensive universities and colleges; (3) liberal arts colleges; (4) two-year colleges and institutes; (5) professional schools and other specialized institutions. The present study will use this classification system with certain modifications. Specifically, environmental perceptions of adult learners enrolled in a research university II (subcategory of doctoral-granting institutions), a comprehensive university, and two (2) community colleges will be studied. Literature concerning community college students indicates that important differences exist between students attending urban institutions and those in rural institutions. For this reason, both types of community colleges will be included in this investigation. Example 3 Investigation of Innovativeness Among Nurses There have been many research studies which have investigated the characteristic of innovativeness within groups. There have not been any which explore the distribution of this characteristic in populations primarily composed of women. The precedent for studying innovativeness has been to use self-report data superimposing categories such as Rogers' within a given time 66 frame for the adoption of an innovation to occur within a social system. This methodology may not give accurate information simply because of the inability of respondents to remember details as they actually happened. It may also be true that, depending upon the innovation in question, the respondent may wish to appear either more or less committed to the notion than he/she actually was. For this purpose, of identifying innovative persons, Hurt, Joseph, and Cook devised an instrument which will predict the innovativeness of persons within a social group. Their instrument, though fairly new, and not well utilized on a variety of populations, has demonstrated high reliability where it has been used. (Nunnally's split half analysis yielded a reliability of .94). Based on a definition of innovativeness as "willingness to change" the authors recommend its use for the prediction of innovativeness as a single characteristic (Hurt, Joseph & Cook). The tool has not been used, however, in conjunction with the more Rogerian self-report of actual practice changes and would be more useful if such a cross-validation were done. Similarly, other studies, geared at a predictive model, have established that certain characteristics are held in common by those who lean toward an innovative nature. Rogers has referred to these as "generalizations". They are attributes which seem highly associated with innovativeness and which might suggest a predictive model. Most of them were however established on a post-deductive inquiry methodology. Therefore, it is desirable to combine the work of Hurt, Joseph and Cook, using their twenty-item index, and validate this index with a self-report measure of actual adoption of a given practice. Further it would serve useful purposes to select certain generalizations suggested by Rogers and examine the data produced to determine whether they are genuinely predictive of the innovative 67 respondents. Similarly, it is advantageous to determine what socioeconomic determinants may also interact with the respondent's degree of innovativeness. The finer breakdown of this structure lends itself to a correlative study. By utilizing demographic information as in the Hollingshead two-factor index, and by incorporating the Pancratz and Pancratz index for Autonomy in Nursing Practice, it would be possible to identify specific characteristics of the individual respondents who are both innovative by self-report, and who are "willing to change" according to the Hurt, Joseph and Cook index. This would allow for three kinds of variables to be superimposed upon the initial two-by-two. Demographic variables including the occupation, income, education, and age of both the respondent and the head-of-household would be available. Personal characteristics will include willingness to break with tradition, empathy, belief in practice autonomy, and cosmopolitanism. Finally, environmental characteristics identified as the degree of specialization in the work environment, size of agency or facility where employed, hours of work per week, and level of responsibility would be included. Example 4 Response to Scarcity in Institutions of Higher Education The literature on organizational change indicates that to be effective an organization must not only maintain itself internally, but it must also adapt to its environment. An effective organization is stable yet able to change appropriately. Organizational health is a concept used to describe the degree to which an organization is capable of managing these two, somewhat at opposing forces. (Owens, 1981, p. 249) observed that in time organizations tend to become obsessed with maintaining themselves, increasing bureaucratic rigidity and seeking to shore up traditional practices. 68 He labeled this an unhealthy organizational climate which tends to emphasize maintenance of the organization at the expense of its need for constant adaptability to keep pace with changing demands and expectations of the external environment. This description of organizational climate fits closely the dynamics Whetten (l98l) listed as responsible for institutions' passive responses to scarcity. Those institutions which respond to scarcity with innovative changes can be viewed as having healthy organizational climates which enable them to adapt to changes in their environment. One way to produce organizational change is through organizational self-renewal. An organization capable of selfrenewal is characterized by: (l) a climate that values adaptability and responsiveness to change, open and upward communications, and problem-solving; (2) a set of clear-cut, explicit and well known procedures through which participants can engage in orderly, systematic, collaborative problem-solving; and (3) an ability to reach out appropriately for ideas and resources to help solve problems (Owens, l98l, p. 250). Rensis Likert (l96l) described ways of managing the interaction-influence system of the organization in ways that would stimulate creativity, promote growth of people in the organization, and facilitate a solution of the organization's problems. He labeled this "organizational climate" which could be measured through seven organizational processes: leadership, motivation, communication, interaction, decisionmaking, goal setting and control feedback (Likert and Likert, l976). Each of these processes can be evaluated in terms related to one of four "systems" of management ranging from System l, exploitive-authoritative to System 4, participative. The systems range from McGregor's (l960) Theory X to Theory Y views of motivation. Likert (l977) found that institutions of higher education having administrative systems closer to System 4 than l experienced more 69 favorable outcomes including innovativeness, which is the focus of this study. Although Smith (l977) did not document his findings, he reported that liberal arts colleges which were able to effect renewal possessed three necessary characteristics: (1) effective leadership, (2) enlistment of widespread support and (3) the unification of purpose and people. These findings suggest that differences in organizational climate and leader behavior should be related to institutional response to scarcity. Owens' conditions for renewal address five of Whetten's dynamics which impede innovation: conservative efforts stress, trained incapacity of administrators to deal with decline, innovation resistant organizational structure, problem-solving based on efficiency data and crises leading to espousal of traditional values. Therefore, measures of organizational climate will produce information which will deal with the speculations raised in response to these factors. The other dynamic, the most innovative members of the organization leave first, is not a plausible explanation of differences in response to scarcity given today's limited opportunities for academics to change institutions. One of the independent variables in this study will be organizational climate as measured by the Profile of Organizational Characteristics (Likert and Likert, l976, pp. 91-93). A second independent variable will be leader behavior as measured by the Profile of Leadership Behavior (Likert and Likert l976, pp.111-113). Because adapting institutions must continually assess and obtain feedback from their external environments, a third independent variable will be institutional external relations measured in three ways: (1) institution's public relations budget as a percentage of total budget, (2) percent of alumni participation and average gift in annual giving, and (3) percent of trustees who attend board meetings regularly and contribute annually. 70 The dependent variable will be response to scarcity at Liberal Arts Category II colleges. Measures of changes in six institutional characteristics during a four-year period from 1976-77 to l980-81 will be used to characterize colleges' responses to scarcity on a scale from passive to intermediate to innovative. The six characteristics are (1) size of student body, (2) addition and deletion of programs, (3) size of faculty, (4) total expenditures, (5) mission and goals and (6) inter-organizational cooperative arrangements. These measures are similar to those used by Mayhew (l979, p. 87) in determining when an institution had ended a creases. Jonsen (l978) included several of these characteristics in defining criteria for success of liberal arts colleges in the eighties. Several variables will be controlled. Institutions which have benefited from unusual and unanticipated changes in their external environments, such as the closing of a competing institution or the receipt of a major gift or grant will be excluded. Institutions in which the response was precipitated by near insolvency will be excluded. Institutions which have undergone frequent and drastic changes in leadership will be treated separately. Although only Liberal Arts II institutions will be included in the study, several differing characteristics of these institutions will be observed to determine their relationship to the dependent variable. These characteristics are rural or urban location, geographic region, institutional size, and single-sex or coeducational student body. The findings of this study will be compared with the findings of two studies of innovation at liberal arts colleges, namely, a study by Smith (l977) and a study of colleges, which had received funds for innovation from the Hill Family Foundation. Example 5 71 Field Independence/Dependence and Classroom Behavior of Teachers In investigating teacher characteristics, one may focus on presage variables, process variables, or product variables (after Mitzel, cited in Soar, l964). Presage variables include personality variables or other characteristics of the teachers themselves. Fieldindependence/field-dependence can be considered a presage variable. Process variables describe the behavior of teachers in action. Teacher-student interaction and classroom climate are examples of process variables. Product variables refer to those changes which occur in students as a result of the teaching process. School achievement can be considered a product variable. Presage variables can be thought of as having some effect on process variables, both of which in turn affect product variables. As one moves from presage to process to product variables, a greater number of variables outside the teacher per se come into play. Process variables are a function of teacher characteristics, characteristics of the teaching environment, and characteristics of the students. Product variables are a function of all these clusters of variables as they interact over time, allowing for the effects of history and maturation to become important. The problem posed here involves one presage variable, fieldinterdependence/field dependence, as it is related to the classroom behavior of teachers, and observed in process variables. Product variables lie outside the scope of this study. As implied by the foregoing, it is unknown at present specifically what process variables one might observe which might differentiate field-independent and field-dependent teachers. The absence of clear conclusions from previous research suggests that a wide investigative net is appropriate. Of use here is the distinction made by Barker (l966) between the role of the investigator as operator and the role of the investigator as transducer. In the 72 former role, the investigator controls the conditions of input to the subject, the conditions under which the subject performs, and the conditions of output from the subject. In the latter role, the investigator translates the phenomena of interest into data with relatively little control over input, context, and output. The investigator as operator contrives a closed system which specifies, among other things, that only certain types of output from the subject will become data. As transducer, the investigator strives to leave the system open which mans, among other things, that virtually any output from the subject has potential to become data. It seems clear that the nature of the problem posed is such that the role of the investigator as transducer is appropriate. Subsequent considerations of method will reflect this decision. A significant dimension of the problem posed is the manner in which the environment is conceived. Willems (l973), Bronfenbrenner (l976), Moos (l973; l976) and Scott (l977) are among those who have grappled with this issue. Scott (l977) noted an absence of an agreed-upon theoretical framework but suggested that three clusters of work could be identified -- environmental psychology, ecological psychology,and ecosystem psychology. Environmental psychology has tended to investigate physical and social variables as well as their effect on behavior. Here the concern is with such environmental characteristics as light, temperature, barometric pressure, noise, crowding, and population density as they affect human performance. Environmental psychology deals with behavior at a molecular level and employs traditional experimental methods. Ecological psychology views behavior and the environment as interdependent, and is the behavior setting which consists of the time, place, and object props and their associated rules and roles of behavior. A classroom with its teacher, students, books, blackboards, chairs, syllabus, lecture notes, and explicit and 73 implicit rules of behavior is a behavior setting. Ecological psychology studies molar behavior and uses naturalistic methods of inquiry. Ecosystem psychology may be defined as the study of behavior as it is influenced by large social systems. Ecosystem psychology studies global, presumably group or organizational behavior. Organization development and process consultation are examples of ecosystem approaches. These three clusters of work can be conceptually related within a number of common themes. One significant theme is the conception of the environment as a series of nested entities among which, although each obeys its own laws, mutual influences occur. Teachers and students, for example, are nested in classrooms, which are nested in departments or schools, which are nested in universities, and so on. Because of this conception of the environment, molecular, molar, and global views each have value, depending on the nature of the problem. I contend that the view afforded by ecological psychology, and specifically by behavior setting theory, has the greatest potential for exploration of the problems posed here. Since the approach is in the middle range, further research can move in either the direction of molecular or global behavior. Furthermore, the central concept of the theory that of the behavior setting can be applied to the analysis of the classroom environments within which the teacher functions. The utility of this concept can best be demonstrated after having explored the concept in greater detail. A behavior setting is a naturally occurring unit which has two major components, the milieu or the time, space, and object props, and the standing pattern of behavior which describes the behavior of the participants en masse. An optional example of a behavior setting was provided by Barker (l963) in the Kurt Lewin Memorial Address he delivered in Philadelphia: 74 It is not often that a lecturer can present to his audience an example of his phenomena, whole and functioning in situ -- not merely with a demonstration, a description, a preserved specimen, a picture, or a diagram of it. I am in the fortunate position of being able to give you, so to speak, a real behavior setting. If you change your attention from me to the next most inclusive, bounded unit, to the assembly of people, behavior episodes, and objects before you, you will see a behavior setting. It has the following structural attributes which you can observe directly: 1. It has a space-time locus: 3:00-3:50 p.m., September 2, l963, Clover Room, Bellevue-Stratford Hotel, Philadelphia, Pennsylvania. 2. It is composed of a variety of interior entities and events: of people, objects (chairs, walls, a microphone, paper), behavior (lecturing, listening, sitting) and other processes (air circulation, sound amplification). 3. Its widely different components form a bounded pattern that is easily discriminated from the pattern on the outside of the boundary. 4. Its component parts are obviously not a random arrangement of independent classes of entities; if they were, how surprising that all the chairs are in the same position with respect to the podium, that all members of the audience happen to come to rest on the chairs, and that the lights are not helter-skelter from floor to ceiling, for example. 5. The entity before you is part of a nesting structure; its components (e.g., the chairs and people) have parts; and the setting, itself, is contained within a more comprehensive unit, the Bellevue-Stratford Hotel. 75 6. This unit is objective in the sense that it exists independently of anyone's perception of it, qua unit (p. 26). Ecological psychology attempts to describe how the congruence between behavior and environment arises, in spite of what are considerable individual differences, people within a behavior setting behave in ways that are remarkably homogeneous. Barker proposed that behavior settings coerce behavior through linkages within behavior settings called circuits. For example, program circuits specify the behaviors to be enacted by the behavior setting inhabitants. In the case of a classroom behavior setting, the program circuit, which is under the control of the principal performer, the teacher, might consist of the list of class meetings, reading assignments, lecture outlines, course requirements, and grading criteria. By using Barker's behavior setting theory and its associated methods, a behavior setting can be described along a number of dimensions. Behavior settings can then be compared to one another. In addition, however, an important aspect of behavior settings is the way people who inhabit the setting are conceived. Barker (l978) explained the fundamental difference between behavior settings and their inhabitants: People have two positions in behavior settings: they are components, and as such they behave in accordance with behavior setting requirements, but they are also inhabitants with unique needs and abilities, and as such they behave in accordance with the requirements of their own motives and perceptions (p. 200). This perspective implies that the teacher plays a dual role. The teacher, like all the components of a behavior setting, is coerced and constrained by the requirements of the classroom behavior setting he or she inhabits. At the same time, the teacher acts in such a way that his or her plan of instruction, needs and motives are achieved. This perspective urges one, in considering 76 differences between teachers who are relatively field-independent or field-dependent, to attend not only to the behavior of the teacher as actor, but also to the characteristics of the behavior settings within which the teachers' actions are enmeshed. One may find, for example, that field-independent and field-dependent teachers differ in terms of how they respond to the forces within the behavior settings they inhabit. To do this, the behavior settings must first by described. At the same time, one may find that these teachers differ in their plans for controlling the classroom behavior settings in which they act. Behavior setting theory and its central concept allows one to investigate simultaneously the behavior of the individual teacher and the environment within which that behavior occurs. Since the problem posed suggests that the behavioral concomitants of fieldindependence/field-dependence are partly a function of the context within which the individual acts, then a framework which is applicable to both of these elements is necessary. In summary, the present study will investigate one presage variable, that of the field-independence/field-dependence of teachers, as it takes form in the process variables, which may be observed in the classroom. Since it is unknown what process variables might differentiate field-independent and field-dependent teachers, the investigator will take the role of the transducer. Further, since process variables are influenced by the context within which teaching occurs, the ecological perspective afforded by Barker's (l968; l978) behavior setting theory will be employed as a way of guiding the analysis of behavior and of the environment. This perspective appears to provide the greatest promise for exploration of the problem posed. 77 Example 6 Arkansas Direct Writing Assessment Local educators like to feel that they are primarily responsible for the curriculum taught in public schools and to claim that education is a local function responsive to the unique needs of children in a community. In reality, though, local educational decisions, practices, and performances are influenced by many external pressures and decisions, most of which are political in nature. The goals of America 2000, the recommendations in the SCANS report for workplace "know-how," and the demand by some for a set of national educational standards demonstrate the increased attention given to restructuring the nation's educational system. The discipline of language arts currently serves as a good example of these and other forces of external influences and pressures on local curricular decisions as well as instructional practices. In 1992, the National Council on Education Standards and Testing (NCEST) concluded that "there do not already exist accepted standards in English/Language Arts for what should be taught, how it should be taught, or what students should know and be able to do" (p. H-1). However, this report also notes that several national organizations are currently working on such standards. NAEP recently devised a set of reading and writing frameworks for use with the 1994 assessments, administered in Spring 1994. The National Council of Teachers of English (NCTE), the International Reading Association (IRA), and the Center for the Study of Reading at the University of Illinois have been cooperatively awarded funds to develop standards such as those recently developed for math. Individual states are also developing state frameworks and attempting to restructure both curriculum and assessment. Congruency of content standards, performance standards, teaching practices, and assessment is an issue deserving serious time and effort on the part of educational leaders. 78 Arkansas language arts educators, curriculum coordinators, and assessment specialists face these national agendas as well as statewide attempts to influence local educational practices. The legislative mandates to restructure education found in Act 236 of 1991 and the transformation of course content guides into learner goals/curriculum frameworks "fleshed out"with local objectives and strategies all signal changes for the language arts classroom. However, of more significance is the state-mandated change in "high stakes" assessment to include a direct writing assessment for students in grades 5, 9, and 11. As local teacher committees across Arkansas work to develop objectives and indicators for the curriculum frameworks, they are caught in the middle of two forces--the external pressures noted above juxtaposed against the local philosophies and resources available to implement the developed curriculum. For example, if language arts teachers (as well as those in other disciplines) are apprehensive about teaching or evaluating student compositions, the curriculum decisions will be different than if they value writing and feel comfortable about evaluating student work. Likewise, language arts indicators for learner outcomes will vary from district to district depending upon the philosophy, training, and skills of the local staff. Teachers operating from a traditional base will probably continue to emphasize basic skills exercises and/or a product-oriented approach to composition. Those trained through the Arkansas Writing Project, integrated language arts classes, or other current methodology instruction will likely favor the newer process-oriented approach measured by authentic assessments, such as portfolios. Even after a local language arts curriculum is in place, its use in classrooms throughout the district depends, to a large extent, upon each individual teacher's attitude toward writing and knowledge or skill necessary for its practical implementation. Based upon an international study reported by Takala and Degenhart (1988), one would expect classroom practices in 79 teaching writing to fall within four broad categories, three of which have traditionally been present in classrooms and only one of which is consistently congruent with current Arkansas restructuring efforts and state-mandated assessment decisions. Upon closing their classroom doors, teachers have always had the option of ignoring or de-emphasizing teaching writing skills in favor of some other subject or topic (that is, literature or reading) they deem more important. A second grouping includes those teachers who decide to stress only the writing "pre-requisite" skills of grammar, mechanics, and sentence structure. Whether their rationale is a lingering pre-occupation with basic skills, a perceived lack of time to "grade" compositions, or a professed lack of knowledge about how to teach writing, some language arts teachers opt for drill and practice exercises in recognizing correct editing conventions of writing rather than for practice in authentic writing tasks. A third grouping of classroom practices includes those teachers who do assign compositions, but who focus on evaluating the finished product rather than teaching the process of pre-writing, composing, revising, and editing. Their students do have practice in writing, but are often uncertain about how to accomplish the tasks or about the criteria upon which the teacher assigns the grade. The only group of teachers in sync with the current external influences impacting language arts instruction and assessment in Arkansas focus on the process of composing, affording students the opportunity to practice realistic tasks involving writing and to learn to self-assess their own writing by becoming familiar with established criteria and techniques. Each of these categories of classroom practices obviously impacts the nature and quality of students' writing experiences and opportunity to learn the skills assessed on the Arkansas Direct Writing Assessment. In turn, the quality of those experiences--i.e., whether students participate in collaborative activities, write in contexts outside the English classroom, practice self-assessment, view the tasks as authentic and meaningful--is very likely related 80 to the students' attitude toward writing, their confidence in their ability to write well, and their competence in performing writing tasks. Various classroom practices may produce different levels of student competence in writing across topics, for diverse audiences, and for different purposes. It is significant that students participating in the 1984 NAEP writing assessment indicated that they valued the importance of writing, but revealed that their enthusiasm for writing gradually decreased as they progressed from elementary school to high school (Applebee, et. al., 1986). All of these variables--from the external influences impacting local decisions to teachers' implementation of those decisions in the classrooms to students' response to those instructional strategies-are linked to student performance on a large-scale assessment such as the Arkansas Direct Writing Assessment. The part each plays, however, is unclear. Guskey (1994) points out the dichotomy of standardized achievement measures and performance assessment as they influence instructional practice. He says the former tend to narrow instruction so that teachers teach to the test. The latter, however, broaden the scope of instruction and may not impact teaching practices to the same degree. Viewing the impact from another perspective, Baker (1994) warns that "some thought must be given to the relationship between assessment and instruction, and the validity criterion of instructional sensitivity" (p. 60). She argues that, unless instruction affects performance on alternative assessments, educators may risk, at the least, misguided inferences about the reasons for differences in performance and, at the most, inequitable consequences for individual students in high-stakes assessments. Thus, once student outcomes are documented, the kinds of influences these bring to bear are two-fold. They may have the effect of altering classroom practices--the goal of those proponents of MDI--Measurement Driven Instruction. The may also have an impact on the external influences by confirming or refuting the 81 assumptions and/or needs underlying the decisions of policymakers. An overview of this conceptual framework is illustrated in Figure 1, and the logic is as follows. External influences on instruction and assessment are listed in the oval on the far left; these include national goals and standards, legislative mandates, state-adopted learner goals and curriculum frameworks, and assessment decisions made above the district level. All of these are sometimes influenced by student outcomes on the assessment tools over a period of time. As local educators define curriculum objectives for each district, they must consider both the external influences and the staff who will implement these objectives in the classrooms. Teachers' knowledge and skills as well as their attitudes can affect the instructional practices used in the classroom to afford students the opportunity to learn the skills necessary to master the objectives. Over time, the students' performance on the assessment measure will impact classroom practices by revealing strengths and weaknesses to teachers and administrators. What happens in the classroom has a major impact on students' opportunity to learn, both in the experiences they incur with the skills being assessed and in the attitudes they develop toward the skills as a result of the instructional practices and evaluations being used by teachers. Student performance on an assessment such as the Arkansas Direct Writing Assessment is a result of all of these factors. However, like many of the variables, it is both influenced by various decisions, practices, and attitudes and, in turn, influences decisions and practices. Unless congruency exists throughout this model, the assessment cannot be considered a valid measure of what students know and can do. The design of this study, as depicted by the model in Figure 1, places the two major variables of classroom practices and student outcomes within the context of the other 82 intervening variables and posits the overarching theoretical perspective that will guide the collection, analysis, and interpretation of data. 3.5 Hypothesis and prediction Recall, a hypothesis is a prediction of a relationship between one or more factors and the problem under study, which can be tested. Questions that you propose to address in your research will be clearer if they are framed in terms of specific hypotheses (models) and predictions. Carefully conceived hypotheses demonstrate that you are aware of how your research fits into prior theoretical or empirical work in your research area and carefully deduced predictions indicate whether or not your reasoning is logically sound. Examples of research hypotheses were given in Topic 1. Usually, there are at least two alternative hypotheses that could be made for any research question raised. The basis for each set of alternatives should be provided and properly referenced. Alternative hypotheses should encompass all possible outcomes of the inquiry. When possible, they should be mutually exclusive, making different predictions. The proposal should make explicit reference to how your data will enable you to distinguish among alternative outcomes. Basic concepts concerning hypothesis testing i) Null hypothesis and alternative hypothesis: In the context of statistical analysis, people talk about null hypothesis 83 and alternative hypothesis. If we are to compare method A with method B about its superiority and if we proceed on the assumption that both methods are equally good, then this assumption is termed as the null hypothesis. As against this, we may think that method A is superior or method B is inferior, we are then stating what is termed as alternative hypothesis. The null hypothesis is generally symbolized as Ho and the alternative hypothesis as Ha. The null hypothesis and the alternative hypothesis are chosen before the sample is drawn (the researcher must avoid the error of deriving hypotheses from data he/she collect and testing the hypotheses from the same data). In the choice of a null hypothesis, the following considerations must be kept in view: a) An alternative hypothesis is usually one, which one wishes to prove and the null hypothesis is the one which one wishes to disprove. Thus, a null hypothesis represents the hypothesis we are trying to reject and the alternative hypothesis represents all other possibilities. b) If rejection of a certain hypothesis when it is actually true involves great risk, it is taken as null hypothesis because then the probability of rejecting it when it is true is (the significance level, which is chosen very small. 3.6 Methods 3.6.1 Research design What is a research design? A research design is the arrangement of conditions for collection and data analysis in a manner that aims to combine relevance to research purpose with economy in procedure. Thus, it is a conceptual structure within which research is conducted. It constitutes the blueprint for collection, measurement and data 84 analysis. In due regard, the research design includes an outline of what the researcher will do from writing hypothesis and its operational implications to final data analysis. The overall research design may be split into four parts: i) the sampling design, which deals with the method of selecting items to be observed of a given study; ii) the observational design, which relates to conditions under which observations are to be made; iii) statistical design, which concerns with the question of how many items are to be observed as well as how information including data collected are to be analyzed; and iv) the operational design, which deals with techniques by which procedures specified in sampling, statistical and observational designs can be carried out. Important features of research design: a) It is a plan that specifies sources and types of information relevant to the research problem. b) It is a strategy specifying, which approach will be used for data collection and data analysis. c) It includes time and cost budgets since most studies are done under these constraints. Thus, research design must, at least contain: i) a clear statement of the research problem; ii) procedures as well as techniques to be used for gathering information; iii) the population to be studied; and iv) methods to be used in processing and data analysis. An efficient and appropriate research design must be prepared before starting the operations. It helps the researcher to organize his/her ideas in form whereby it will be possible for him/her to look for flaws and inadequacies. 85 Study area Study area is the place/places where the study will be carried out. Usually supplying with maps indicating location of the study area will be helpful. Depending on the nature of study, some other aspects may be included such as demographic data, physical features (with a map/maps), weather patterns and other useful pieces of information pertaining to the study area. The Sample and Sample Size SAMPLING involves selection of a number of study units from a defined study population Some studies involve only small numbers of people and thus, all of them can be included. However, often research focuses on such a large population such that it is only possible to include some of its members in the investigation. Thus, we have to draw a SAMPLE from the total population. In due regard, we will be confronted with the following questions: i) What is the group of people (STUDY POPULATION) from which we want to draw a sample? ii) How many people do we need in our sample? iii) How will these people be selected? The study population has to be clearly defined, for example, according to sex, age, residence and so on. Apart from persons, a study population consists of villages, institutions, records and so forth. Each study population consists of STUDY UNITS. The manner we define our study population and our study unit depends on the problem we want to investigate. See Table below. Problem Malnutrition related to weaning in Study population All children 6 to 24 months of age in 86 Study unit One child between 6 and 24 months in Manyoni District High drop-out rates in primary schools in Mpwapwa District Inappropriate recordkeeping for hypertensive patients registered in Dodoma Regional Hospital Manyoni District Manyoni District All primary schools One primary school in Mpwapwa District in Mpwapwa District All records on hypertensive patients in Dodoma Regional Hospital One record on a hypertensive patient registered in Dodoma Regional Hospital Representativeness If researchers want to draw conclusions that are valid for the whole study population, they should take care to draw a sample in such a way that it is representative of that population. A REPRESENTATIVE SAMPLE has all important characteristics of the population from which it is drawn. For example, if you want to interview 100 mothers to obtain a complete picture of weaning practices in Mpwapwa District, you would have to select these mothers from a representative sample of villages. It would be unwise to select them from only one or two villages because that might give you a distorted or biased picture. It would also be unwise to interview only mothers who attend the under-5s clinic because those who do not attend the clinic may wean their children differently. However, sometimes representativeness of the sample is not a major concern. For example, in exploratory studies, where the main aim is to get a rough impression of how certain variables are distributed in the population or to identify and explore new variables, you may deliberately choose to include study units that 87 are extremes in the study population, with respect to certain characteristics. Sampling Procedures An important issue influencing choice of the most appropriate sampling method is whether or not a sampling frame is available, that is, a listing of all units that compose the study population. If a sampling frame is not available, it is not possible to sample the study units in such a way that the probability for the different units to be selected in the sample is known. There are two nonprobability sampling methods, namely, convenience sampling and quota sampling. If a sampling frame exists or can be compiled, probability sampling methods can be used. With the methods, each study unit has an equal or at least a known probability of being selected in the sample. Note the following probability sampling methods that will be discussed; - Simple random sampling, - Systematic sampling, - Stratified sampling, - Cluster sampling and - Multi-stage sampling. Non-probability sampling methods 1. Convenience sampling CONVIENCE SAMPLING is a method in which for convenience sake, the study units that happen to be available at the time of data collection are selected in the sample. 88 Many clinic-based studies use convenience sample. For example, a researcher wants to study villagers’ attitudes toward familyplanning services provided by the Maternal Child Health clinic. He/she decides to interview all adult patients who visit the outpatient clinic during one particular day. This is more convenient than taking a random sample of people in the village and it gives a useful first impression. However, a drawback of convenience sampling is that the sample may be quite unrepresentative of the population you want to study. Some units may be over-selected, others under-selected or missed altogether. It is impossible to adjust for such a distortion. If you need to be representative, you have to use another sampling method. 2. Quota sampling QUOTA SAMPLING is a method that ensures that a certain number of sample units from different categories with specific characteristics appear in the sample so that all such characteristics are represented. In this method, the investigator interviews as many people in each category of study unit as he/she can find until he/she has filled his/her quota. For example, the researcher of family-planning study just mentioned suspects that religion might have a strong effect on patients’ attitudes toward family-planning services. He/she is afraid to miss Catholics who are a minority in the area (for example, in Unguja). Therefore, he/she decides to include in the study 60 patients from each of the different religious groups 89 (Hindu, Moslems, Assemblies of God, Catholics and so on) and to extend the study over 3 or 4 days to obtain the desired sample. Remarks: Quota sampling is useful when researchers feel that a convenience sample would not provide the desired balance of the study units. However, like a convenience sample, it does not claim to be representative of the entire population. PROBABILITY SAMPLING METHODS Non-probability sampling methods are inappropriate if the aim is to measure variables and generalize findings from a sample to the total study population. Non-probability sampling, for example, would not be appropriate in a study that aims to determine prevalence of malnutrition in a whole region. For that type of study, a probability sampling method should be used. PROBAILITY SAMPLING involves random selection procedures to ensure that each unit of the sample is chosen on the basis of chance. All units of the study population should have an equal or at least, a known chance of being included in the sample. Probability sampling requires that a listing of all study units exists or can be compiled. The listing is called the sampling frame. 1. Simple random sampling This is the simplest form of probability sampling. To select a simple random sample, you need to: - Make a numbered list of the units in the population from which you want to draw a sample; - Decide on the size of the sample; and - Select the required number of sampling units, using a “lottery” method or a table of random numbers. 90 For example, a simple random sample of 50 students is to be selected from a school of 250 students. Using a list of all 250 students, each student is given a number (1 to 250), and these numbers are written on small pieces of paper. All the 250 papers are put in a box, after which the box is shaken vigorously to ensure randomization. Then, 50 papers are taken out of the box and the numbers are recorded. The students belonging to these numbers will constitute the sample. 2. Systematic sampling In SYSTEMATIC SAMPLING, individuals are chosen at regular intervals (for example, every fifth) from the sampling frame. Ideally, we randomly select a number to tell us where to start selecting individuals from the list. For example, a systematic sample is to be selected from 1,200 students of a school. The sample size selected is 100. The sampling fraction is: 100 (sample size) 1 1,200 (study population) 12 Therefore, the sampling interval is 12. The number of the first student to be included in the sample is chosen randomly, for example, by blindly picking one out of twelve pieces of paper numbered 1 to 12. If number 6 is picked, then every twelfth student will be included in the sample, starting with student number 6, until 100 students are selected: the numbers selected would be 6, 18, 30, 42, and so on. Systematic sampling is usually less time consuming and easier to perform than simple random sampling. However, there is a risk of bias because the sampling interval may coincide with a systematic variation in the sampling frame. For instance, if we want to select a 91 random sample of days on which to count clinic attendance, systematic sampling with a sampling interval of 7 days would be inappropriate because all study days would fall on the same day of the week, which might, for example, be a market day. 3. Stratified sampling The described simple random sampling method does not ensure that the proportions of individuals with certain characteristics in the sample will be the same as those in the whole study population. It is important that the sample includes representative groups of study units with specific characteristics (for example, residents from urban and rural areas or different age groups). Then, the sampling frame must be divided into groups or STRATA, according to these characteristics. Random or systematic samples of a predetermined size will have to be obtained from each group (stratum). This is known as STRATIFIED Stratified sampling is only possible when we know what proportion of the study population belongs to each group we are interested. An advantage of stratified sampling is that we can take a relatively large sample from a small group in our study population. This allows us to get a sample that is big enough to enable us draw valid conclusions about a relatively small group without having to collect an unnecessarily large (and hence, expensive) sample of the other, larger groups. However, in so doing, we are using unequal sampling fractions and it is important to correct for this when generalizing our findings to the whole study population. For example, a survey is conducted on household water supply in a district comprising 20,000 households of which 20% are urban and 92 80% are rural. It is suspected that in urban areas, access to safe water sources is much more satisfactory. A decision is made to include 100 urban households (out of 4,000, which gives a 1 in 40 sample) and 200 rural households (out of 16,000, which gives 1 in 80 sample). Because we know the sampling fraction for both strata, access to safe water for all district households can be calculated. 4. Cluster sampling It may be difficult or impossible to take a simple random of the units of the study population, either because a complete sampling frame does not exist or because of other logistical difficulties (for example, visiting people who are scattered over a large area may be too time consuming). However, when a list of groupings of study units is available (for example, villages or schools) or can easily be compiled, a number of these groupings can be randomly selected. Selection of groups of study units (clusters) instead of selection of study units individually is known as CLUSTER SAMPLING. Clusters are often geographic units (for example, districts, villages) or organizational units (for example, clinics, training groups). For example, in a study of knowledge, attitudes and practices related to family planning in rural communities of Singida region, a list is made of all villages. Using the list, a random sample of villages is chosen and all adults in the selected villages are interviewed. 4. Multi-stage sampling In very large and diverse populations, sampling may be done in two or more stages. This is often the case in community-based 93 studies in which people to be interviewed are from different villages and the villages have to be chosen from different areas. For example, in a study of utilization of pit latrines in Dodoma district, 150 homesteads are to be visited for interviews with family members as well as for observations on types and cleanliness of latrines. Dodoma district is composed of six wards and each ward has between six and nine villages. The following four-stage sampling procedure could be performed: 1. select three wards out of six by simple random sampling. 2. For each ward, select five villages by simple random sampling (15 villages in total). 3. For each village, select ten households. Because simply choosing households in the centre of the village would produce a biased sample, the following systematic sampling procedure is proposed: o Go to the centre of the village. o Choose a direction in a random way: spin a bottle on the ground and choose the direction the bottleneck indicates. o Walk in the chosen direction and select every third or every fifth household (depending on the size of the village) until you have ten you need. If you reach the boundary of the village and you still do not have ten households, return to the centre of the village, walk in the opposite direction and continue to select your sample in the same way until you have ten. If there is nobody in a chosen household, take the next nearest one. 4. Decide beforehand whom to interview), for example, the household head, if present or the oldest adult who lives there and who is available). 94 A MULTI-STAGE SAMPLING procedure is carried out in phases and usually involves more than one sampling method. The main advantages of cluster and multistage sampling are the following: i) ii) A sampling frame of individual units is not required for the whole population. Initially, a sampling frame of clusters is sufficient. Only within clusters that are finally selected we need to list and sample the individual units. The sample is easier to select than a simple random sample of similar size because the individual units in the sample are physically together in groups, instead of scattered all over the study population. The main disadvantage of cluster and multistage sampling is that: Compared to simple sampling, there is a larger probability that the final sample will not be representative of the total population. The likelihood of the sample not being representative depends mainly on the number of clusters selected in the first stage. The larger the number of clusters, the greater the likelihood that the sample will be representative. BIAS IN SAMPLING BIAS in sampling is a systematic error in sampling procedures that leads to a distortion in results of the study. 95 Bias can also be introduced as a consequence of improper sampling procedures that result in the sample not being representative of the study population. For example, a study was conducted to determine health needs of a rural population to plan primary health care activities. However, a nomadic ethnic group, which represented one third of the total population, was left out of the study. As a result, the study did not give a picture of the health needs of the total population. Data collection methods Data collection methods allow us to systematically collect information about our objects of our study (people, objects and phenomena) and about the settings in which they occur. In data collection, we have to be systematic. If data are collected haphazardly, it will be difficult to answer our research questions in a conclusive way. Various data collection methods can be used such as: i) Using available information, ii) Observing, iii) Interviewing (face-to-face), iv) Administering written questionnaires, v) Focus group discussions (FGD) and vi) Other data collection methods There are two types of data: Secondary data Information/data from written sources, published and Unpublished 96 Primary data Data collected from the field by the researcher himself/herself Such data are collected using the above mentioned data collection methods. Observing OBSERVATION is a data collection method that involves selecting, watching and recording behaviour and characteristics of living beings, objects or phenomena. Observation of human behaviour can be undertaken in different ways: Participant observation whereby the observer takes part in the situation he or she observes Non-participant observation whereby the observer watches the situation, openly or concealed but does not take part. Observations may serve different purposes. They can give additional more accurate information on behaviour of people than interviews or questionnaires. Questionnaires may be incomplete because we forget to ask certain questions and informants may forget or be unwilling to mention certain things. Therefore, observations can check on information collected (especially on sensitive topics such as alcohol or drug use or stigmatization of leprosy, tuberculosis, epilepsy or AIDS patients). Or they may be a primary source of information (observations of children’s play, for instance, collected in a systematic way). Observations of human behaviour can form part of any type of study but because they are time consuming, they are most often 97 used in small-scale studies. Observations can also be made on objects. For instance, the presence or absence of a latrine and its state of cleanliness may be observed. If observations are made using a defined scale, they may be called measurements. Normally, they require additional tools. For example, in nutritional surveillance, we measure weight and height by using weighing scales and a measuring board. We use thermometers for measuring body temperature. Interviewing AN INTERVIEW is a data collection method that involves oral questioning of respondents, either individually or as a group. Structured or semi-structured interviews Focus group discussions Administering written questions A WRITTEN QUESTIONNAIRE (also referred to as selfadministered questionnaire) is a data collection method in which written questions are presented that are answered by respondents in written form. Can be administered in several ways, for example: - Sending questionnaires by mail - Gathering all or part of respondents in one place at one time, giving oral or written instructions and letting respondents fill out the questionnaires or - Hand-delivered questionnaires to respondents and collecting them later Questions can be either open-ended or close-ended. 98 Focus group discussions A FOCUS GROUP DICUSSION (FGD) is a group discussion composed of between 6 and 12 persons guided by a facilitator, during which members talk freely and spontaneously about a certain topic The purpose of an FGD is to obtain in-depth information on concepts, ideas, perceptions and ideas of the group. The FGD aims to be more than a question-answer interaction. The idea is that group members discuss the topic among themselves. FGD methods can be used to: i) Focus research and develop relevant research hypotheses by exploring in great depth the problem to be investigated and its possible causes. For example: The Bahi District Health Officer had noticed that there were an unusually large number of malnutrition cases for under 5 children reported from one large village in the district. Because he had an idea of why there might be more information in that village, he decided to organize three focus groups (one of leaders, one of mothers and one of health staff assigned to do home visits in that village). He hoped to identify potential causes of the problem through focus groups and then develop a more intensive study, if necessary. ii) Formulate appropriate questions for more structured, larger scale surveys. For example: In planning a study on incidence of childhood diarrhea and feeding practices, an FGD showed that in that community under study, children below the age of 1 year were not perceived as having “bouts of diarrhea,” but were merely “having loose stools” that were associated with milestones such as 99 sitting up, crawling and teething. In the questionnaire that was being developed, the concept diarrhea was thus, carefully circumscribed using community’s notions. iii) Supplement information on community knowledge, beliefs, attitudes and behaviours already available but incomplete or unclear. For example: There is a high drop-out rate in child welfare clinics among children over the age of six months. A previous study indicates that mothers give reasons such as “too busy,” “have other domestic commitments,” or “experience transport problems.” Because such same mothers have previously brought their infants under six months of age regularly, you suspect that there are other factors. A focus group discussion, with a few groups of mothers could provide in-depth pieces of information on reasons for the changes in their perceptions and behaviour regarding use of the clinic for children over six months old. iv) v) Develop appropriate messages for health-education Programmes. For example: A rural health clinic wanted to develop a health-education programme focused on weaning problems most often encountered by mothers in the surrounding villages and sought on how to do about them. An FGD could be used for exploring relevant concepts as well as for testing drafts when developing the messages. Explore controversial topics For example: In a household survey, it appeared that male informants most frequently said that their wives kept household money, whereas female informants 100 maintained their husbands kept the money. An FGD with a group of females and a separate one with a group of males may bring forward the complicated patterns and variations of financial responsibility in the domestic group. It may be interesting to have a third session of males and females together to discuss the differences in perception. It has to be noted that FGDs are not used to test hypotheses or produce research findings that can be generalized. How to conduct a focus group discussion Preparation Recruitment of participants: Participants should be roughly of the same socio-economic group or have a similar background in relation to the issue under investigation. The age and sexual composition of the group should facilitate free discussion. If you need to obtain information on a topic from several different categories of informants who are likely to discuss the issue from different perspectives, you should organize a focus group for each major category. For example: first, a group for men and a group for women or second, a group for older women and a group for younger women. It may be interesting to have an additional discussion with groups that are confronted with each other’s opinions. Participants should be invited at least one or two days in advance and the general purpose of the FGD should be explained. 101 Physical arrangements Communication and interaction during the FGD should be encouraged in every way possible. Arrange chairs in a circle. Make sure the area will be quiet, adequately lighted and so on and that there will be no disturbances. Try to hold the FGD in a neutral setting that encourages participants to freely express their views. For example, a health centre is not an ideal place to discuss about traditional medical beliefs or preferences for other types of treatment. Preparation of a discussion guide There should be a written list of topics to be covered. It can be formulated as a series of open-ended questions. Guides for different groups gathered to discuss the same subject may vary slightly, depending on their knowledge or attitudes and the manner the subject can first be explored with them. Conducting the session One of the members of the research team should act as “facilitator” for the focus group. One should serve as “recorder.” Functions of the facilitator: The facilitator should not act as an expert on the topic. His or her role is to stimulate and support discussion. Thus, the facilitator should: i) Introduce the session Introduce yourself as facilitator and introduce the recorder. Introduce participants by name or ask them to introduce themselves. Put the participants at ease and 102 ii) explain the purpose of the FGD, the kind of information needed and how the information will be used (for example, for planning of a health programme, an education programme, a water sanitation programme, an environment conservation and protection programme and so on). Encourage discussion Be enthusiastic and humorous and show interest in the group’s ideas. Formulate questions and encourage as many participants as possible to express their views. Remember, there are no “right” or “wrong” answers. React neutrally to both verbal and non-verbal responses. iii) Encourage involvement Avoid a question-and-answer session. Some useful techniques include the following: o Ask for clarification: “Can you tell me more about…? o Reorient the discussion when it goes off the track: Say, “wait, how does this relate to ..?” o Say, “Interesting point, but how about ..? o Use one participant’s remark to direct a question to another, for instance, “Mrs. Q said …, but now how about you Mrs. R?” o When dealing with a dominant participant, avoid eye contact or turning slightly away to discourage the person from speaking, or thanking the person and changing the subject. o When dealing with a reluctant participant, using the person’s name, requesting his/her opinion, making more frequent eye contact to encourage his/her participation. 103 iii) Build rapport, emphasize Observe non-verbal communication. Ask yourself, “What are they saying? What does it mean to them? Be aware of your own tone of voice, facial expressions, body language and those of participants. iv) Avoid being placed in the role of expert When you are asked for your ideas or views by a respondent, remember that you are not there to educate or inform. Direct the questions back to the group by saying: “What do you think?” “What would you do?” Set aside time, if necessary, after the session to give participants the information they have asked for. Do not try to comment on everything that is being said. Do not feel you have to say something during every pause in the discussion. Wait a little and see what happens. v) Control the rhythm of the meeting, but in an authoritative Way. Listen carefully and move the discussion from topic to topic. Subtly control the time allocated to various topics so as to maintain interest. If participants spontaneously jump from one topic to the other, let the discussion continue for a while because useful additional information may surface and then summarize the points brought up and reorient the discussion. vi) Take time at the end of the meeting to summarize, check for agreement and thank the participants. Summarize the main issues brought up, check whether or not all agree and ask for additional comments. Thank 104 the participants and let them know that their ideas have been a valuable contribution and will be used for planning the proposed research/intervention/health education materials and so on. Listen for additional comments made after the meting has been closed. Functions of the recorder The recorder should keep a record of the content of the discussion as well as emotional reactions and important aspects of group interaction. Assessment of the emotional tone of the meeting and the group process will enable you to judge the validity of information collected during the FGD. Items to be recorded include the following: - Date, time, and place; - Names and characteristics of participants; - General description of the group dynamics (level of participation, presence of a dominant participant, level of interest); - Participants’ opinion, recorded as much as possible in their own words, especially for key statements; - Emotional aspects (for example, reluctance, strong feelings attached to certain opinions); and - Vocabulary used – particularly in FGDs that are intended to assist in developing questionnaire or health-education materials. It is highly recommended that a tape recorder be used to assist in capturing information. Even if a tape recorder is used, notes should be taken as well, in case the machine malfunctions and so that information will be available immediately after session. 105 A supplementary role of the recorder could be to assist the facilitator (if necessary) by drawing his/her attention to: - missed comments from participants and - missed topics (the recorder should have a copy of the discussion guide during the FGD). If necessary, the recorder could also help resolve conflict situations that the facilitator is having difficulty handling. Number and duration of sessions Number of sessions The number of focus group sessions to be conducted depends upon project needs, resources and whether or not new information is still coming from the sessions (that is, whether contrasting views from various groups in the community are still emerging). One should plan to conduct at least two different FGDs for each sub-group (for instance, two males and two females). Duration A focus group session typically lasts up to an hour and a half. Generally, the first session with a particular type of group is longer than the following ones because all of the information is new. Thereafter, if it becomes clear that the groups have the same opinion on particular topics, the facilitator may be able to move the discussion along more quickly to other topics that still elicit new points of view. ANALYSIS OF RESULTS - After each focus group session, the facilitator and recorder should meet to review and complete notes taken during the meeting. This is also the right moment to evaluate how the focus group went and what changes might be made when facilitating future groups. 106 - Then, a full report of the discussion should be prepared that reflects the discussion as completely as possible, using the participants’ own words. List the key statements, ideas and attitudes expressed for each topic of discussion. - After the transcript of the discussion is prepared, code the - statements right away, using the left margin. Write comments in the right margin. Formulate additional questions if certain issues are still unclear or controversial and include them in the next FGD. - Further categorize statements for each topic, if required. Compare answers of different sub-groups (for example, answers of young mothers and answers of mothers above child-bearing age in the FGD on changes in weaning practices). The findings should be coherent. For example, if young women in all FGDs state that they start weaning some 3 to 6 months earlier than their mothers did and women above child-bearing age confirm this statement, one likely has a solid finding. If findings contradict each other, one may need to conduct some more FGDs or bring together representatives from two different sub-groups to discuss and clarify the differences. - Summarize the data in a matrix, diagram, flowchart or narrative, if appropriate and interpret the findings. - Select the most useful quotations that emerged from the discussions to illustrate the main ideas. 107 REPORT WRITING Start with a description of the selection and composition of the groups of participants and a commentary on the group process so the reader can assess the validity of the reported findings. Present your findings following your list of topics and guided by the objective(s) of your FGD. Include quotations whenever possible, particularly for key statements. Plan for data processing and data analysis Why is it necessary to prepare a plan for data processing and analysis? Such a plan helps the researcher assure that at the end of the study: - All information he or she needs has indeed been collected and in a standardized way; and - He or she has not collected unnecessary data that will never be analyzed. In due regard, the plan for data processing and analysis must be made after careful consideration of objectives of the study and the list of variables. It has to be known that procedures for analysis of data collected through qualitative and quantitative research approaches are quite different. Thus, one must consider the type(s) of study and the different data collection methods used when making a plan for data processing and analysis. For quantitative data, the starting point in analysis is usually a description of data for each variable for all the study units included in the sample. For qualitative data, it is more a matter of describing, summarizing and interpreting data obtained for each study unit (or for each group of study units). The researcher starts analyzing while collecting the data so that questions that remain unanswered (or new questions that come up) can be addressed before data collection is over. 108 Preparation of a plan for data processing and analysis will provide you with better insight into the feasibility of the analysis to be performed as well as resources that are required. It also provides an important review of the appropriateness of your data collection tools. Note: the plan for data processing and analysis must be prepared before data are collected in the field so that it is still possible to make changes in the list of variables or the data collection tools. What should the plan include? When making a plan for data processing and analysis, the following aspects should be considered: - sorting data, - Performing quality-control checks, - Data processing and - Data analysis Sorting data An appropriate system for sorting data is important for facilitating subsequent processing and analysis. If you have different study populations, for example, village health workers, village health committee and the general public, you would number the questionnaires separately. In a comparative study, it is best to sort the data right after collection into two or three groups that you will be comparing during data analysis. For example, in a study concerning reasons for low acceptance of family-planning services, users and non-users would be basic categories. In a study of reasons nurses object to being posted to rural area, rural and urban nurses would be basic categories. In a case-control study, the cases are to be compared with the controls. 109 It is useful to number the questionnaires belonging to each of these categories separately right after they are sorted. For example, the questionnaires administered to users of family-planning services could be numbered U1, U2, U3 and so on and those for non-users N1, N2, N3 and so on. In a cross-sectional survey, it may be useful to sort the data into two or more groups, depending on objectives of the study. Performing quality control checks Normally, data have already been checked in the field to ensure that all information has been properly collected and recorded. However, before and during data processing, the information should be checked again for completeness and internal consistency. - If a questionnaire has not been filled in completely, you will have MISSING DATA for some of your variables. If there are many missing items in a particular questionnaire, you may decide to exclude the whole questionnaire from further analysis! - If an inconsistency is clearly due to a mistake made by the researcher or assistant, for example, if a person in an earlier question is recorded as being a non-smoker, whereas all other questions reveal that he is smoking, it may still be possible to check with the person who conducted the interview and to correct the answer. - If the inconsistency is less clearly a mistake in recording, it may be possible 9in a small-scale study) to return to the respondent and ask for clarification. - If it is not possible to correct information that is clearly inconsistent, you may consider excluding this particular part of data from further processing and analysis. If a certain question produces ambiguous or vague answers throughout, the whole question should be excluded from further analysis. 110 Note: A decision to exclude data should be considered carefully because it may affect validity of the study. Such a decision is ethically correct and it testifies to the scientific integrity of the researcher. You should keep an accurate account of how many answers or questionnaires you had to exclude because of incompleteness or inconsistency and discuss it in your final report. If you process your data by computer, quality-control checks must also include a verification of how the data have been transformed into codes and subsequently entered into the computer. Data processing As you begin planning for data processing, you must make a decision concerning whether to process and analyze data: - manually using data master sheets or manual compilation of the questionnaire, or - by computer, for example, using a micro-computer and existing software or self-written programmes for data analysis. Data processing involves: - categorizing the data, - coding and - summarizing the data on a master sheet. Categorizing Decisions have to be made concerning how to categorize responses. 111 Categorical variables that are investigated through closed questions or observation, for example, observation of the presence or absence of latrines in homesteads, categories have been decided upon beforehand. In interviews, the answers to unstructured questions, for example, “Why do you smoke?” can be categorized to a certain extent, depending on knowledge of possible answers. However, there should always be a category called “others, specify ..,” which can only be categorized afterwards. These responses should be listed and placed in categories that are a logical continuation of categories you already have. Answers that are difficult or impossible to categorize may be put into a separate residual category called “others,” but this category should not contain more than 5% of answers obtained. For numerical variables, data are usually collected without any precategorization. Because you are often still discovering the range and dispersion of the different values of these variables when you collect your sample, for example, home-clinic distance for outpatient, decisions concerning how to categorize numerical data (and how to code them) are usually made after they have been collected. Coding If data are entered into a computer for subsequent processing and analysis, it is essential to develop a CODING SYSTEM. CODING is a method used to convert (translate) data gathered during the study into symbols appropriate for analysis. For computer analysis, each category of a variable is usually given a number, for example, the answer “yes” may be coded as 1, “no” as 2 and “no response” as 9. The codes should be entered on the 112 questionnaire (or checklists) themselves. When finalizing your questionnaire, for each question, you should insert in a box for the code in the right margin of the page. These boxes should not be used by the interviewer. They are only filled in afterwards during data processing. Take care that you have as many boxes as the number of digits in each code. Note: If you intend to process your data by computer, always consult an experienced person before you finalize your questionnaire. Also, if analysis is done by hand using data master sheets, it is useful to code your data. Coding conventions Common responses should have some code in each question because it minimizes mistakes by coders. For example: Yes (or positive response) code 1 No (or negative response) code 2 Do not know code 9 Codes for open-ended questions This can be done only after examining a sample of questionnaires. The most frequently occurring responses should be coded. It may be necessary to group similar types of responses into single categories in order to limit their number. If there are very many categories it is difficult to make meaningful summaries during analysis. Finally, it advised personnel responsible for computer analysis should be consulted very early in the study as soon as the questionnaire and dummy tables are finalized. 113 Data master sheet If data are processed by hand, it is often most efficient to summarize the raw research data in a so-called DATA MASTER SHEET in order to facilitate analysis. On a master sheet, all answers of individual respondents are tallied by hand. An example of a master sheet Respondent Ques. Q2: Sex number 1: Age M F (years) 1. 31 √ 2. 45 √ 3. 25 √ Etc Total Question 4: Smoking Yes No No response √ √ √ Q5: No. of cigarettes 10 --10-15 Data are easier to tally from master sheets than from original questionnaires. Straight counts can be performed for background variables (such as sex, residence) and for all independent variables under study (such as smoke/no-smoker in the exercise). Questionnaire data may be compiled by hand instead of using master sheets if it is difficult or impossible to put the information (such as answers to open-ended questions) in a master sheet. Hand compilation is also necessary if you want to go back to the raw data to make additional tabulations in which different variables are related to each other. Note: In a comparative (analytical) study, you should use different master sheets for the two or three groups you are comparing (for example, users and non-users of family planning services). 114 In a cross-sectional survey, it may be useful to have master sheets depending on the nature and objectives of the study and whether you want to compare two or more groups. Great care should be taken when filling in master sheets. You should verify that all totals correspond to the total number of study units (respondents). If not, all subsequent analytical work will be based on erroneous figures. There should be special columns for “no response” or missing data to arrive at accurate total figures. Hand compilation is used when the sample size is small. The following procedures will help to ensure accuracy and speed: 1. If only one person is doing compilation, use manual sorting. If a team of two persons work together, use either manual sorting or tally counting. 2. Manual sorting can be used only if data on each subject are on a different sheet(s) of paper. 3. To do manual sorting, the basic procedure is to; - take one question at a time, - Sort the questionnaires into different piles representing the various responses to the questions (examples: male/female; used hospital/health centre/traditional practitioners), - Count the number in each pile. When you need to sort out subjects who have a certain combination of variables (for example, females who used each type of health facility) sort the questionnaires into piles according to the first question, then subdivide the piles according to the response to the other question. 4. To do tally counting, the basic procedure is: - One member of the compiling team reads out the information while the other records it in the form of a tally (for example, “III” representing three subjects or IIII representing five subjects who have a particular variable). - Tally count for no more than two variables at one time (for example, sex plus type of facility used). 115 If it is necessary to obtain information on three variables (for example, sex, time of attendance at health centre and diagnosis), do a manual sorting for the first question, then tally count for the other two variables. After tally counting, add the tallies and record the number of subjects in each group. 5. After doing either manual or tally counting, check the total number of subjects/responses in each question to make sure that there has been no omission or double count. Note: Researchers often assume that hand compilation is merely “common sense” and do not train their staff in the correct procedure. In due regard, many hours of work are wasted in trying to detect the source of error due to double counts, wrong categorization and omissions. Computer compilation It has to be noted that the larger the sample, the more beneficial the use of a computer will be. You must ensure that necessary equipment is available including necessary expertise. Computer compilation has the following steps: 1. Choosing an appropriate computer programme; 2. Data entry; 3. Verification or validation; 4. Programming (if necessary); and 5. Computer outputs. 1. Choosing an appropriate computer programme There are several computer programmes that can be sued to process and analyze research data. The most widely computer programmes include the following: 116 - LOTUS spreadsheet programme (from the Lotus Development Corporation); - dBase (Version IV, etc), a data management programme (from Ahston-Tate). - Epi Info (Version xy), a very consumer friendly programme for data entry and analysis, which also has a word processing function for creating questionnaires (developed by Centers for Disease Control, Atlanta and the World Health Organization, Geneva). - SPSS, which is a quite advanced Statistical Package for Software Systems (by SPSS Inc.). 2. Data entry To enter data into the computer, you have to develop a data entry format, depending on the programme you are using. However, it is possible to enter using dBase (which is relatively good for data entry) and undertake analysis in LOTUS 1-2-3 or SPSS. *** Validation of instruments Significance Budget and budget justification Reasons for having a budget - A detailed budget will help to identify resources that are already locally available and additional resources that may be required. - The process of budget design will encourage you to consider aspects of the work plan you have not thought about and will serve as a useful reminder of activities planned, as your research gets underway. Time to prepare a budget A complete budget is normally not prepared until the final stage of project planning. However, cost is usually a major limiting factor. 117 Thus, costs must always be kept in mind during planning so that your proposals will not have an unrealistically high budget. Remember that donor agencies and the like usually set limits for research project budgets. How should a budget be prepared? It is convenient to use the work plan as a starting point. Specify, for each activity in the work plan, what resources are required. Determine for each resource needed the unit cost and the total cost. For example, in the work plan of a study to determine utilization of family planning methods in a certain district, it is specified that 5 interviewers will each visit 20 households in clusters of 4 over a time period of 5 working days. A supervisor will accompany one of the interviewers each day using a car. The other 4 interviewers will use motor cycles. The clusters of households are scattered over the district but are on average 50 kilometres from the district hospital from where the study is conducted. The budget for the field work component of the work plan will include funds for personnel, transport and supplies. NOTE: Unit cost (for example, per diem or cost of petrol per km), the MULTIPLYING FACTOR (number of days) and TOTAL COST should be clearly indicated for all budget categories. Costs involved in fieldwork for a family-planning study Budget category Unit cost Multiplying factor Total cost 1. Personnel Daily wage Number of staff- Total (including days (no. staff x per diem) no. of working Interviews $ 10 days) Supervisor 5 x 5 = 25 $250 $ 20 1x5=5 $ 100 118 2. Transport Motorcycles Car 3. Supplies Pens Questionnaires Personnel TOTAL Cost per Number of km km (no. vehicles x no. days x no. $ 0.10 km/day) 4 x 5 x 100 = $ 0.40 2000 1 x 5 x 100 = 500 Transport TOTAL Cost per Number item 12 $ 1.00 120 $ 0.20 Supplies TOTAL Grand TOTAL $ 350 Total $ 200 $ 200 $ 400 Total $ 12 $ 24 $ 36 $ 786 You have to convert from foreign currency to local currency at the end, for example, US$ 786 each dollar 1,200 T. Sh. equivalent to 943,200 T. Sh. Note: there are several formats for budgets. Citation Consider the following example. Brain is among many possible phenotypic windows on evolutionary changes, which occurred in the course of hominid evolution (Holloway, 1996). It is part of the central nervous system within the cranial vault (Seeley, et. al., 1999). Its major subdivisions are the forebrain consisting of the cerebrum and diencephalons, the midbrain and the hindbrain consisting the pons, medulla oblongata and cerebellum (Zuckerman, 1988; Van De Graaf and Stuart, 1999). Cranial nerves from all subdivisions 119 supplying other parts of the body emerge from the base of the brain (Zuckerman, 1988). Blood arteries supply these brain areas, while veins evacuate metabolic wastes from them (Zuckerman, 1988). Both the cerebrum and cerebellum have cortices including nuclei of grey matter, whereas the rest have only nuclei but no cortices (Zuckerman, 1988). However, it has to be noted that, in the course of brain evolution, substantial changes have been observed in the cerebrum and cerebellum, in particular, their corytices (Blinkov and Glezer, 1968; Marilyn, et. al., 1999; Kaas and Preus, 1999; Allman and Andrea, 1999). The cerebrum is the largest of the subdivisions accounting for about 80% by mass (Van De Graaf and Stuart, 1999). References The following format is suggested: Allman, J. and Andrea, H. 1999 “Brains, Maturation Time and Parenting.” Neurobiology of Aging 20: 447-454. Blinkov, S. M. and Glezer, I. I. 1968 “The Cerebral Cortex: Volume, Surface Area and Width of Its Layers.” In S. M. Blinkov, and I. I. Glezer (Eds.). The Human Brain in Figures and Tables: A Quantitative Handbook, pp 173-185. New York: Plenum Press. Kapama, Ndili n.d. “Extinct Elephants from Makuyuni.” Preliminary Field Report to COSTECH Kothari, C.R. 2003 Research Methodology: Methods and Techniques. New Delhi: Wishwa Prakashan 120 4.0 HOW TO WRITE A RESEARCH REPORT Research results are of little use unless reported so that others can access them. Scientific reports have a standard format that should always be followed. A usual sequence can be as follows: Title page *Certification *Declaration and Copyright *Acknowledgements *Dedication *Table of Contents *List of Tables *List of Figures *List of Abbreviations and Acronyms Abstract CHAPTER ONE 1.0 Introduction 1.1 Background Information 1.2 Statement of the Problem 1.3 Research Objectives 1.3.1 Main Objective 1.3.2 Specific Objectives 1.3.3 Hypothesis/hypotheses 1.3.4 And/or Research Tasks Research Questions 1.4 Significance of the Study 1.5 Definition of Basic Terms and Concepts 1.6 Limitation of the Study 1.7 Organization of the Study 2.0 Literature Review 2.1 General Overview 2.2 **Theoretical Grounding 2.3 Matter related to the Problem 2.4 Synthesis (including highlights to the gap) 3.0 Research Methodology 121 3.1 Research Design 3.2 Study Area 3.3 The Sample and Sample Size 3.4 Sampling Procedures 3.5 Data Collection Methods 3.6 Validity and Reliability 3.7 Data Analysis Plan 4.0 Results and Discussion ** Use Specific Objectives as Sections (Sub-titles) 5.0 Summary of the Study, Conclusion and Recommendations 5.1 Summary of the Study 5.2 Conclusion 5.3 Recommendations 5.3.1 Recommendations for Action 5.3.2 Recommendations for Further Research Bibliography Appendices: Appendix I: Interview Schedule Appendix II: Questionnaire Appendix III: Focus Group Discussion Guide Appendix IV: Observation Guide 4.1 Front matter (title page and abstract) Front Matter (in case of journal manuscripts) A lot of journal manuscripts begin with a title page followed by an abstract on a separate page. Then, body of the paper begins on the third page. The title and abstract need to be crafted carefully, for each need to be concise and informative for inducing the reader to read further. Title Page Many journals require a manuscript to have a separate title page, although others allow the abstract to appear on the same page. For manuscripts of monographic length, a table of contents may be required on or after the title page. In general, a title page consists 122 of the work’s title, the name(s) of the author(s) and other pieces of information appropriate to the type of report. The other pieces of information on the title page vary with the situation. Student papers often need to bear the course name and number, the name of the supervisor and date. Journal manuscripts generally require institutional affiliation(s) of the author(s) and mailing address of the corresponding author. In addition, some journals now ask the corresponding author to provide an e-mail address of fax number. Some require a suggested short title to be used as a running head in the journal. The title itself like in research proposal is a trade-off between conciseness and description of content. Considerations discussed in the Third Topic apply here, but there are some minor differences between titles of proposals and reports. It is frequently the case that one changes the title after drafting the report and may change it again after revision. Journals, like funding agencies, may have specific requirements of titles such restriction of their length, but there are at least two types of titles found in journal articles that many people would consider inappropriate for proposal: full questions and questions. The author’s name: In your very first manuscript for publication decide upon the name you will use and stick to it forever. For example, one used Janet Mabuga and later on, after being married changed to Janet Limo. After divorce, the author switched back to Janet Mabuga. Listing of publication from the same author will differ some being in M’s and others in L’s. Thus, do not contribute to the confusion: pick a publishing name and stay with it. Abstract As with summary of the proposal, write the abstract of a report last, even though it comes first. Just insert the word “Abstract” on 123 a page and leave the page blank as a reminder when first drafting your manuscript. Granting agencies may provide a form to which the summary is constrained, the equivalent constraint for journal abstracts being a word limit. Many journals require that the abstract not exceed 5 percent of the word count of the text (Introduction through discussion). The UDSM requires a one page abstract for Master degree dissertations/theses. Aspects to be included in the Abstract An abstract should state the problem under investigation and sketch methods used, but it must go on to concentrate mainly on results. Generally, restrict the discussion to principal implication of findings. Why is the Abstract necessary? The Abstract is the most important part of the report because far more people will read it than the whole of the paper (or any other part of it). Thus, save the Abstract to write last and work very hard to make it clear as well as informative. 4.2 Body of the report (introduction, methods, results, discussion) The corpus of a scientific report consists of an introduction, methods used, results obtained and discussion of those results. Decide on the best sequence of your report. For example, a hierarchically structured study is ordinarily sequenced with the general issue first and then more specific ones. Others may be reported by starting from the simplest issue to the most complex one. Once the best sequence has been decided, carry it through all sections of the report in parallel: identifying points in the 124 introduction, stipulating methodologies, presenting data in results section and considering outcomes in the discussion section. When a major section is divided - like it is often true for Methods, Results and Discussion - preserve the sequence of issues in the order and titles of subheads whenever possible. Introduction The introduction is the first major section of a scientific paper: the guide to what you did and why you did it. The first paragraph of the introduction is the most important one because it either succeeds or fails to entice the reader to look at the entire paper. A common error is attempting to report too much in the opening paragraph. Keep it short and interesting. ALWAYS STATE NO LATER THAN THE LAST SENTENCE OF THE OPENING PARAGRAPH WHAT THE STUDY IS ABOUT. The introduction usually contains the general statement of the problem and some indication of why the problem is interesting and important. Such justification often requires citation of what is already known about the problem, establishing the niche for the work that will be reported. When appropriate, the introduction should conclude specific predictions that were tested by the study. Methods The term “Methods” is generally taken to include both procedures and physical equipment used in the study. If “methods” is equated with procedures per se, the section is sometimes entitled Methods and Materials. Methods sections are usually subdivided, but the subdivisions are not standard because they must be appropriate to each individual report. Consider the following example: - Field site or study area, with ecological description if pertinent 125 - Housing and husbandry of captive animals Study population, with marking schemes and so on Division of subjects into experimental and control groups Field times, including dates and number of hours spent in observational studies Experimental design and protocol Equipment used to take and analyze data Criteria used in identifying and separating recognized entities Procedures and protocols followed in taking data Statistical methods used in data analysis The methods section needs to be complete and clear such that another investigator could replicate your study from the information provided. Inadequately described methods may cast doubt on the quality and validity of your results. If you followed a standard, well established protocol or your study area has been described in detail in a readily accessible source, you may find that you can summarize the most salient features or cite the more elaborate reference. If you modify established methodology or develop wholly new approaches, be certain to explain them in adequate detail. Results Findings are the heart of your report. Thus, they should be presented clearly and concisely. Where appropriate, provide results in figures and tables, which are easier to absorb at a glance than are long explanations. The accompanying text needs to point out what can be seen in figures and tables, but should not repeat the information there. Journals are always strapped for space and cannot permit redundancy among text, figures and tables. Some journals may also limit number of tables and figures. In general, results section should contain no discussion or interpretation of findings. 126 Results section may be subdivided appropriately for the study. Subdividing results will also help the reader follow logic of analyses as in moving from general results (for example, overall diet) to specific topics (such as seasonal variation). Discussion Discussion sections are nearly always too long. A common mistake of inexperienced authors is to attempt to wring from their data by discussion more than data can really show. In a good study, data speak for themselves if the introduction and results sections have been well crafted. Discussion sections deal with three matters: first, an analysis of sources of error in the data; second, integration with what was previously known; and finally, implications for future study. The first matter is too often overlooked, yet it becomes End matter (acknowledgement and appendices) References Tables and Figures Submission and review HOW TO PRESENT RESEARCH Research seminars (content and organization, practice, style and delivery, ending the talk and question and answer period) Audio visual aids Talks at scientific meetings Posters at scientific meetings 127