Conducting Scientific Research
Science and scientific research
Scientific research is systematic, controlled, empirical, and critical investigations of natural phenomena guided by theory and hypotheses about the presumed relations among such phenomena (Nachmias & Nachmias, 1987). Thus, science is a systematic process of producing knowledge. It develops theoretical structures, tests their internal consistencies, and tests the hypotheses by controlling variables. A theory is a set of interrelated constructs (concepts), definitions, and propositions that present a systematic view of phenomena by specifying relations among variables, with the purpose of explaining and predicting the phenomena. There are some major characteristics when science is viewed this way. The first characteristic is its cyclic nature. They typically start with a problem and ends in a tentative empirical generalization. And the generalization ending this cycle is the beginning of the next cycle. Second, scientific process is self-correcting as. Tentative generalizations to research problems are tested logically and empirically by built-in-checks.
The scientific approach is grounded on a set of fundamental assumptions that are unproved and unprovable. There are, (1) nature is orderly, (2) we can know nature, (3) knowledge is superior to ignorance, (4) all natural phenomena have natural causes, (5) nothing is self-evident, and (6) knowledge is derived from the acquisition of experience. The ultimate goal of the social science is to produce an accumulating body of reliable knowledge. Such knowledge would enable us to explain, predict, and understand empirical phenomena that interest us. The scientific methodology is first and foremost self-correcting. A major function of methodology is to facilitate a communication between scientists who either shared or want to share a common experience (methodology as rules for communication). The scientific methodology demands competence in logical reasoning and analysis (methodology as roles for reasoning) The term intersubjectivity is more appropriate than objectivity (methodology as rules for intersubjectivity). The scientific methodology establishes the logic of justification. Methodology is indifferent to how scientists arrive at their insights, but asks only whether they are justified in reaching claims for knowledge. The activities of scientist within the context of discovery are not limited by methodology.
Science and other alternative way of knowing
Science and common sense differ in five ways (Kerlinger, 1973). First, scientists systematically build theoretical structures, test them for internal consistency, and subject aspects of them to empirical test. Second, scientists systematically and empirically test their theories and hypothesis. Nonscientists test hypothesis, too, but they test them in a selective fashion. Third, scientists try systematically to rule out variables that are possible causes of the effects under study other than the variables hypothesized to be the causes. Fourth, the scientist is constantly preoccupied with relations among phenomena. The scientist consciously and systematically pursues relations. Fifth, the scientist carefully rules out metaphysical explanations. Science is concerned with things that can be publicly observed and tested.
According to Nachmias & Nachmias (1987), there are three other general modes that have served the purpose of acquiring knowledge: the authoritarian mode, the mythical mode, and the rationalistic mode. In the authoritarian mode, knowledge is looked for by referring to those who are socially or politically defined as qualified producers of knowledge. In the mythical mode, knowledge is obtained from supernaturally knowledgeable authorities such as prophets, divines, gods, and mediums. It depends on using ritualistic and ceremonial procedures. Rationalism is a school of philosophy that holds that the totality of knowledge can be acquired by strict adherence of the forms and rules of logic. The underlying assumptions of rationalism are that (1) the human mind can understand the world independently of observable phenomena, and (2) that forms of knowledge exist that are prior to our experiences. A major distinction among these modes is the way in which each vests credibility in the producer of knowledge (Who says so?), the procedure by which knowledge is produced (How do you know?), and in the effect of the knowledge produced (What difference does it make?). Similarly, Kerlinger (1973) says that there are four general ways of knowing, (1) the method of tenacity, (2) the method of authority, (3) the a priori method, and (4) the method of science. Only scientific approach has a characteristic of self-correction. There are built-in checks all along the way to scientific knowledge. Objectivity is agreement among expert judges on what is observed or what is to be done or has been done in research. Scientists systematically and consistently use the self-corrective aspect of the scientific approach.
Static and dynamic view of science
There are two broad views of science. The static view is that science is an activity that contributes systematized information to the world. The scientist's job is to discover new facts and to add them to the already existing body of information. The emphasis is on the present state of knowledge and adding to it and present set of laws, theories, hypotheses, and principles. The dynamic view or heuristic view regards science more as an activity. It emphasizes theory and interconnected conceptual schemata that are fruitful for further research. The function of science is to make discoveries, to learn facts, to advance knowledge in order to improve things. A different view of the function of science is that the function of science is to establish general laws covering the behaviors of the empirical events or objects with which science in question is concerned, and thereby to enable us to connect together our knowledge of the separately know events, and to make reliable predictions of events as yet unknown.
Normal science is viewed as the routine verification of the dominant theory in any historical period. Paradigms are necessary; without them scientific research could not take place as a collective enterprise, for science needs an organizing principle. Kuhn views revolutionary science as the abrupt development of a rival paradigm that can be accepted only gradually by a scientific community. Paradigm transformation is what is revolutionary science.
Science and paradigm development
Thomas Kuhn (1970) differentiates among the sciences by the extent to which they have a developed paradigm or shared theoretical structures and methodological approaches about which there is a high level of consensus. There are differences in the level of paradigm development across scientific fields and these differences have significant consequences for a number of important outcomes (Pfeffer, 1993). Paradigm development refers to the technological uncertainty associated with the production of knowledge in a given scientific field or subspecialty. Fields with highly developed paradigms, in which there was more consensus, should be characterized by more efficient communication -- less time needed to be spent defining terms or explaining concepts. Perhaps the most important effect of paradigm development and the consensus implied by that construct is on the subsequent development of knowledge in a field. As Stephen Cole (1983) argued: Accumulation of knowledge can occur only during periods of normal science which are characterized by the adherence of the scientific community to a paradigm. Without agreement on fundamentals, scientists will not be able to build on the work of others and will spend all their time debating assumptions and first principles.
When scientists agree among themselves to explain phenomena in terms of base-line theories, they project their findings into shared perceptual framworks that reinforce the collective nature of research by facilitating communication and comparison and by defining what is important or irrelevant. Consensus is enforced when numbers of a field develop a set of methodological standards and ensure that these are consistently maintained. A substantial amount of the variation in the level of paradigm development is a consequence of the social structure, culture, and power relations that characterize the discipline.
Building blocks of theory development
A theory is a statement of relations among concepts within a set of boundary assumptions and constructs. The function of a theory is that of preventing the observe from being dazzled by the full-blown complexity of natural or concrete events (Bacharach, 1989). It is a system of constructs and variables in which the constructs are related to each other by propositions and the variables are related to each other by hypotheses. According to Wetten (1989), a complete theory must contain four essential elements. (1) What: which factors (variables, constructs, concepts) logically should be considered as part of the explanation of the social or individual phenomena of interest? Comprehensiveness (i.e., are the relevant factors included?) and persimony (i.e., should some factors be deleted because they add little additional value to the understanding?) are exist for judging the extent to which we have included the right factors. (2) How: how are they related? Operationally this involves using arrows to connect the boxes and it typically introduces causality. (3) Why: what are the underlying psychological, economic, or social dynamics that justify the selection of factors and the proposed causal relationships? What and How describe; only Why explains. What is passing as good theory includes a plausible, cogent explanation for why we should expect certain relationships in our data. (4) Who, Where, When: These conditions place limitations on the propositions generated from a theoretical model (propositions involve concepts, whereas hypotheses require measures).
Legitimate, value-added contribution to theory development
What and How: one way to demonstrate the value of a proposed change in a list of factors is to identify how this change affects the accepted relationships between the variables. That is, how the addition of a new variables significantly alters our understanding of the phenomena by reorganizing our causal maps. Why: it commonly involves borrowing a perspective from other fields, which encourages altering our metaphors and gestalts in ways that challenge the underlying rationales supporting accepted theories. Who, When, Where: it is insufficient to point out limitations in current conceptions of a theory's range of application. Proposed improvement addressing only a single element of an existing theory are seldom judged to be sufficient. Theoretical critiques should marshal compelling evidence and should propose remedies or alternatives.
There are such factors to judge conceptual papers. (1) What's new? scope (how much of the field is impacted) is less important in determining the merits of a contribution than is degree (how different is this from current thinking.) (2) So what? will the theory likely change the practice of organizational science in this area? (3) Why so? Are the underlying logic and supporting evidence compelling? Are the author's assumptions explicit? (4) Well done? Does the paper reflect seasoned thinking? (5) Done well? It the paper well written? (6) Why now? Is this topic of contemporary interest to scholars in this area? (7) Who cares? What percentage of academic readers are interested in this topic?
Usefulness is a theory's prescriptive value in terms of the degree to which it contains actionable solutions to "real world" problems (Brief & Dukerich, 1991). Usefulness as a criterion for evaluating organizational behavior theories would appear to be threatened by such failure to expect generalizability or by the inability to specify a priori the likelihood that a prescription will hold in a given context. Usefulness, or a theory's prescriptive value, may not a good criterion for judging a theory. The first reason is that it may be illogical to expect a theory to be useful according to the definition. Second reason is that a theory could not logically contain authoritative rules of action for practitioners since all theories are necessarily fallible.
Gioia & Pitre (1990) propose multiparadigm perspective in theory building. The use of single research paradigm produces too narrow a view to reflect the multifaceted nature of organizational reality. Because different paradigms are grounded in fundamentally different assumptions, they produce markedly different ways of approaching the building of theory. A paradigm is a general perspective or way of thinking that reflects fundamental beliefs and assumptions about the nature of organizational phenomena (ontology), the nature of knowledge about these phenomena (epistemology), and the nature of ways of studying those phenomena (methodology). Burrell and Morgan (1979) have organized these differences along objective-subjective and regulation-radical change dimensions. The functionalist paradigm is characterized by an objectivist view of the organizational world with an orientation toward stability or maintenance of the status quo. The interpretive paradigm is characterized by a more subjectivist view, also with an apparent concern with regulation. The radical humanist paradigm also is typified by a subjectivist view, but with an ideological orientation toward radically changing constructed realities. The radical structurist paradigm is typified by an objectivist stance, with an ideological concern for the radical change of structural realities. The modern study of organizations has been driven mainly by social science variations of natural science models.
Theory building refers to the process or cycle by which such representations are generated, tested, and refined. The interpretive paradigm is based on the view that people socially and symbolically construct and sustain their own organizational realities. Therefore, the goal is to generate descriptions, insights, and explanations of events so that system of interpretations and meaning, and the structuring and organizing processes, are revealed. This way of theory building is inductive in nature and the process is iterative, cyclical, and nonlinear. The goal of radical humanist paradigm is to free organization members from sources of domination, alienation, exploitation, and repression by critiquing the existing social structures with the intent of changing it. In this paradigm, theory building is best viewed as having a political agenda. While interpretive theory building focus on how a particular social reality is constructed and maintained, radical humanists focus on why it is so constructed and ask whose interests are served by the construction and sublimation to the deep-structure level. Hypothesis testing is rare, and even literature reviews are not a central characteristic of theory-building efforts. The theory building in radical structuralist paradigm is related to that of radical humanism by virtue of the shared ideology for transformation Societal class or industry structures are seen as objectively real and are taken as instruments of domination. Historical, dialectical, and critical modes of inquiry are used in theory generation. The goal is to understand, explain, criticize, and act on the structural mechanisms that exist in the organizational world. Theory building involves the rethinking of data in light of refinements of viewpoints. The theory building process is a pronounced exercise in argumentation and marshalling of historical evidence. Functionalist paradigm seeks to examine regularities and relationships that lead to generalizations and universal principles. Organizational structures is taken as an objective phenomenon that is external to, and independent of, organization members. In functionalism, theory refinement is the watchword. Theory building typically takes place in a deductive manner.
Logic of Theoretical Networks
Mitchell & Bigram (1971) explain the logic of theoretical networks. A logic of theoretical network is based on at least four ideas. First, propositions mentioning unobservables are too weak in isolation to generate verifiable consequences by deduction. Second, confirmation should be substituted for verification as the criterion for the truth of the statements of science. A statement is said to be confirmed when its deductive implications are verified. Third, a network is said to be exist when at least one construct, which is linked to multiple reductions, is linked to another construct, which is also linked to multiple reductions. Fourth, the adequacy of a theory is evaluated in terms of its competitive support. A theory is competitively supported to the extent that data follow naturally from it, and competing theories require augmenting assumptions or modifications to explain the same data. Dulany's logic of theoretical networks may be summarized in five principles: (1) theoretical terms are introduced postulationally by a network of theoretical sentences, (2) the network is always open to the addition of new terms and to the specification of new linkages between terms that are already in the network, (3) terms get their meaning through connections to other variables in the network, which provide the competitive support of the sentences in which they appear, (4) the specification of the meaning of any term is always partial, and (5) confidence in the referential meaning of a term is as strong as the competitive support of the network or theory in which it appears.
Constructs and variables
Kerlinger (1973) explains constructs and variables. A concept expresses an abstraction formed by generalization from particulars. A construct is a concept. But it has the added meaning of having been deliberately and consciously invented or adopted for a special scientific purpose. It enters into theoretical schemes and is related in various ways to other constructs. A construct is defined and specified that it can be observed and measured. Variables are property that takes on different values. A variable is something that varies. A variable is a symbol to which numerals or values are assigned. Words or constructs can be defined in two general ways. (1) We can define a word by using other words. (2) We can define a word by telling what actions or behaviors it expresses or implies A constitutive definition defines a construct with other constructs. An operational definition assigns meaning to a construct or variable by specifying the activities or operations necessary to measure it. There are (1) measured and (2) experimental operational definitions. A measured operational definition describes how a variable will be measured. An experimental operational definition spells out the details (operations) of the investigator's manipulation of a variable.
Problems and hypotheses
Kerlinger (1973) also explains the problems and hypotheses. A problem is an interrogative sentence or statement that asks: What relation exists between two or more variables? The problem should express a relation between two or more variables. The problem should be stated clearly and unambiguously in question form. The problem and the problem statement should be such as to imply possibilities of empirical testing. A hypothesis is a conjectural statement of the relation between two or more variables. Hypotheses are always in declarative sentence form, and they relate variables to variables. Hypotheses are statements about the relations between variables. Hypotheses carry clear implications for testing the stated relations. Hypotheses are important and indispensable tool of scientific research because (1) they are working instruments of theory, (2) hypotheses can be tested and shown to be probably true or probably false, and (3) hypotheses are powerful tools for the advancement of knowledge because they enable scientists to get outside themselves. Problems and hypotheses direct investigation, enable the researcher to deduce specific empirical manifestations implied by them. The important difference between problems and hypotheses are that hypotheses can be tested and a problem cannot be scientifically solved unless it is reduced to hypothesis form because a problem is a question. If a problem is too general, it is usually to vague to be tested. Too great specificity is perhaps a worse danger than too great generality. Thus problems and hypotheses have to reflect the multivariate complexity of psychological, sociological, and educational reality.
According to Lundberg, there are several ways to develop hypotheses. (1) Four prerequisites: Acquiring a "knowledge of acquaintance" of the phenomena, really knowing the subject, possessing an ingrained paradigm (think unconsciously in accord with fundamental model), and the ability to "galumph" (voluntarily place obstacles in one's own path). (2) Exploratory approaches: it concerns accidents. The paradoxical incident, intensive case study, and analyzing of a practitioner's or craftsman's rule of thumb are source of hypotheses. (3) Intentional search approaches: intentional use of analogy, hypothetico-deductive, contextual twist. (4) Extending-coupling approaches: build on previous research, forming a well-defined line of investigation. Making findings interactive with the findings, observations or capabilities of others. New idea must be examined in light of critical requirement. Mental experiment.
In initiating research, one is interested in the context of discovery, not justification; and one's intuition (logic-in-use) is more prominent than one's rational logic. Intuition here is something preconscious and outside the preferred inference structure. Research depends on ideas, and valuable research comes from ideas for really new questions and hence new hypotheses. The process of getting and developing ideas is undoubtedly a confused mixture of observation, thinking, asking why, cherishing little unformed notions, etc. New research questions seem to result from examining our assumptions and a combination of passive observation, putting questions to nature and active observation. What is vital for stimulating research is the transformation of our assumptions into questions.
Platt (1964) insists on using strong inference in doing research. Strong inference consists of applying the following steps to every problem in science: (1) devising alternative hypotheses, (2) devising a crucial experiment, with alternative possible outcomes, each of which will, as nearly as possible, exclude one of more of the hypotheses, (3) carrying out the experiment so as to get a clean result, and (4) recycling the procedure, making subhypotheses or sequential hypotheses to refine the possibilities that remain. Francis Bacon showed the fruitfulness of interconnecting theory and experiment so that the one checked the other. The most important was the conditional inductive tree, which proceeded from alternative hypotheses, through crucial experiments, to exclusion of some alternatives and adoption of what is left. When the method of multiple hypotheses become coupled to strong inference, the scientific search becomes an emotional powerhouse as well as an intellectual one.
Choice of methodology
The choice of methodology should be made on the basis of the possibilities and limitations of that methodology vis a vis the research problem to which it is to be applied (McGrath, 1964). We should choose the methodology that we will use in a given case on the basis of the kinds of information we are seeking and to maximize the amount of information which we will gain about that problem. Then we must compare alternative approaches in terms of their relative effectiveness in providing the desired information. When we do research, we ask (a) what are the important or relevant variables, (b) how does each of them vary, and (c) how do they covary. The variables include (a) properties of the class of object or entity being studied, (b) properties of the environment, situation, or setting within which that class of objects exists, and (c) properties of the action or behavior of the objects in relation to the environment. Any one variable can be treated in one of four mutually exclusive ways: (W) to control a certain variable, (X) to manipulate a certain variable, (Y) to deal with a given variable by permitting it to vary freely and measuring the values of it, and (Z) to ignore a variable by permitting it to vary freely, but failing to determine what values do occur. The amount of information which can be gained about any given situation is a function of the amount of "uncertainty" or potential information, which is inherent in that situation. The potential information contained in a situation depends on the number of variables and the number of values which each variable can assume. We gain positive research information to the extent that we reduce the number of possible combinations of values by ascertaining that two or more variables vary concomitantly. The reduction of potential information represents a restriction of the scope of our study and a limitation on the generality of our findings. We should view the over-all effectiveness of a study by comparing its information yield to the total potential information of the referent situation. In this view, we lose "comprehensiveness" when we control variables; we lose "efficiency" when we ignore variables; and we lose "effectiveness" when we do either of those.
To control a variable rather than permitting it to vary freely without measuring it prevents a loss of efficiency in the study by reducing the noise. On the other hand, to control a variable at a single value rather than making it occur at each of a series of specific values or letting it vary but measuring freely reduces the information potential of the study below the information potential of the referent situation and thus reduce the scope or comprehensiveness of the study. Manipulations in the laboratory are often relatively weak, both for ethical reasons and because of the inherent artificiality of the motivational conditions under which participants are operating. The uncontrolled and unmeasured operation of a variable generates "noise" within a study design. Field studies are relatively comprehensive but insufficient. Laboratory experiments are relatively efficient but low in comprehensiveness. Experimental simulations lie between field studies and laboratory experiments in both comprehensiveness and efficiency. Computer simulations are often relatively low in comprehensiveness, while the concept of efficiency does not apply since they yield no research information in the present use of that term.
Internal and external validity
Campbell and Stanley (1963) mention the internal validity and external validity in research design. Internal validity is the basic minimal without which any experiment is uninterpretable: Did in fact the experimental treatments make a difference in this specific experimental instance? External validity asks the question of generalizability: To what populations, settings, treatment, variables, and measurement variables can this effect be generalized? In terms of internal validity, the eight different classes of extraneous variables are (1) history, (2) maturation, (3) testing, (4) instrumentation, (5) statistical regression, (6) selection, (7) experimental mortality, and (8) selection-maturation interaction. The factors jeopardizing external validity or representativeness are (9) the reactive or interaction effect of testing, (10) the interaction effects of selection biases and the experimental variable, (11) reactive effects of experimental arrangement, and (12) multiple-treatment interference.
According to Campbell and Stanley (1963), there are three pre-experimental designs. (1) The one-shot case study has such a total absence of control. (2) The one-group pretest-posttest design has such extraneous variables that can jeopardize internal validity as history, maturation, testing, instrumentation, statistical regression and so on. (3) The static group comparison has extraneous variables such as selection and mortality. Next, there are three true experimental designs. (4) The pretest-posttest control group design can rule out many of extraneous variables that influence internal validity. However, there are factors that jeopardizing external validity such as interaction of testing and X, interaction of selection and X, other interaction with X, and reactive arrangement. (5) The solomon four-group design deservedly has higher prestige and represents the first explicit consideration of external validity factors. (6) The posttest-only control group design controls for testing as main effect and interaction, but unlike Design 5, it does not measure them. The statistical tests available for Design 4 are more powerful than those available in Design 6. The availability of pretest scores makes possible examination of the interaction of X and pretest ability level, thus exploring the generalizability of the finding more thoroughly.
The definition of laboratory experimentation include that (1) experimental events occur at the discretion of the experimenter, (2) they use controls to identify sources of variation, and (3) most descriptions refer to all situations where precise measurement of variables are possible (Fromkin & Sternfert, 1976). They are most effectively used in laboratory environment. A laboratory is defined as any settings specifically created for the purpose of conducting research. The aim of laboratory experimentation is to identify cause-effect relationships.
The major advantage of laboratory experimentation is the wide variety of manipulative and statistical controls associated with experimental strategies, for the purpose of inferring cause-and-effect relationship. Extraneous variables can be (a) eliminated from the research settings, (b) held constant for all participants, or (c) allowed to vary from one participant to the next, but in such a way that their average effect across treatment conditions remains constant. Another advantage is that experiments identify several issues and problems which were previously unrecognized in the world of organization theory, research and practice. Weaknesses in laboratory experiment are, in addition to the problem of generalizability, there are demand characteristics and social desirability. Also, partly due to the ethical limitations and partly because of practical limitations, laboratory manipulations are often much weaker than their "real life" counterparts.
The artificiality of the laboratory imposes severe restraints on the external validity of findings. One approach to improve the external validity of research in the laboratory and field would be to determine the variables distinguishing the setting, subjects, and behaviors in a particular research study from those settings, subjects, and behaviors to which generalization is desired. Another approach would be to broaden our samples of actors, settings, and behaviors. A final approach would be to employ coordinated strategies of research in which theories in industrial-organizational psychology are tested in both laboratory and field setting. Other issues related to the generalizability are, whether the events we create in the laboratory of field follow the same lows as those that govern similar events in the target setting and whether or not our treatment conditions are restricted in the range of effects that exist in the target setting (the extent to which the independent variables are representative of stimuli as they exist in the environment.)
Maximum precision and generality (and internal and external validity) are not simultaneously obtainable with any single research strategy or by any single data collection. The limitations of experimental strategies with respect to breadth of information are offset by the two factors: (1) relatively lower cost of laboratory experimentation is conductive to the procedures of partial or complete replication, and (2) potential information loss due to the experimental control of variables is somewhat nullified by the existence of prior knowledge about most phenomena being studied on organizational psychology.
Three techniques of quasi-controls to delineate the effects of demand characteristics are (1) Post-experimental inquiry (the experimenter may conduct judicious inquiry of the subjects' perceptions during a post-experimental session), (2) Noon-experiment (inquiry prior to the experiment -- pilot research), and (3) Simulator groups (subjects are asked to pretend that they have been affected by the experimental treatment).
There may be cues in the experimental situation related to the arousal or heightening of evaluation apprehension, and related to hints about the appropriate responses for obtaining favorable evaluation from the experimenter. Evaluation apprehension and its effects can be reduced or minimized by assuring subjects that the purpose of the experiment is more of mathematical or technical nature than an investigation of the subjects' personalities, and that the experimenter is not interested in individual responses but in normative or nomothetic aspects of responses from groups of individuals.
Bacharach, S. B. (1989). Organization Theories: Some criteria for evaluation, Academy of Management Journal, 14, 496-515.
Brief, A. P., & Dukerich, J. M. (1991). Theory in organizational behavior: Can it be useful? Research in Organizational Behavior, 13, 327-352.
Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research. Chigago: Rand McNally, p. 1-13.
Dipboye, R. L., & Flanagan, M. F. (1979). Research settings in Industrial and Organizational Psychology. American Psychologist, 34, 131-156.
Fromkin, H. L., & Strenfert, S. (1976). Handbook of Social Psychology. Chapter 10. Laboratory Experimantation.
Gioia, P. A., & Pitre, E. (1990). Multiparadigm perspectives on theory building. Academy of Management Journal, 15, 584-602.
Kerlinger, F. N. (1973). Foundations of behavioral research (2nd ed.). New York, Holt, Rinehart & Winston.
Kuhn (1970). The Structure of Scientific Revolution. (2nd. ed.). Chicago: University of Chicago Press.
Lundberg, C. C. Hypothesis creation in organizational behavior research. Academy of Management Review.
McGrath, J. E. (1964). Toward a "Theory of Method" for research on organizations. In W. Cooper, H. Leavitt, & M. Shelly II (Eds.), New Perspectives in Organizational Research, New York, Willey, 533-556.
Nachmias & Nachmias (1987), Chapter 1, (3-27). The Scientific Approach.
Pfeffer, J. (1993). Barriers to the advance of organizational science: Paradigm development as a dependent variable. Academy of Management Review, 15, 599-620.
Platt, J. R. (1964). Strong inference. Science, 146, 347-353.
Sackett, P. R., & Larson, J. R., Jr., (1990). Research strategies and tactics in industrial and organizational psychology. Handbook of Industrial and Organizational Psychology, Vol. 1, Chapter 8, 419-428.
Whetten, D. A. (1989). What constitutes a theoretical contribution? Academy of Management Journal, 14, 490-495.