Philosophy of Management Science (0LM05) Exam preparation Author Student-id M.A.J. Kusters 0557541 Exam date 23 June 2009 Version – 2.1 1. Fallacies and management fashions 1.0 Practicalities Review and discussion question 1-5 + exercises sheet Question 6-7 are introduction of the assignment Baggini “basic tools for argument” Godfrey Smith Abrahamson “Management Fashions”. 1.1 The theory of argumentation Argument: premise i (i ∈ {1,..,n}) conclusion. An argument can be evaluated by (a) evaluating its premises, (b) evaluating the claim that its premises support the conclusion, and (c) evaluating its conclusion. Types of arguments: 1) deductive, conclusion follows necessarily, 2) inductive, conclusion follows with probability and 3) defeasible, conclusion follows with plausibility. Validity and soundness: argument is valid if whenever the premises are true, so is the conclusion. Argument is sound, if it has true premises and is valid (an argument with false premises can be valid). Formal fallacy: an argument that seems but isn’t valid. Subtypes: modus ponens (If A then B, A, thus B) = valid, case of affirming the consequent (if A then B, B, thus A) = invalid. Fallacy: argument that illicitly, but often convincingly, leads to a conclusion. Types: 1) fallacies that rest on dubious premises, 2) fallacies of irrelevance (premises irrelevant to the conclusion), 3) fallacies based on hasty conclusions or suppressed evidence. 1. Begging the question/circular reasoning: if one or more premises take a stand on an issue or claim that is in dispute in the debate. 2. Straw man argument: an argument that presents an attack on a weak position as an attack on another more credible one. 3. Appeal to ignorance: argument for or against a proposition on the basis of a lack of evidence against or for it. ~ More fallacies: false dilemma (there are more options to choose from than presented), appeal to authority / expert opinion, overgeneralization (to conclude something about a whole set on the basis of a subset), false causes (to conclude to a cause where there is only statistical covariance), slippery slope (to argue on the basis of unproven consequences in the long term), suppressed or selective evidence (to use evidence misleadingly, by no citing counter-evidence), non-sequitur (invalidity of argument, inference is wrong). Invalidity is about deductive arguments, f.i. a missing premise which should be added. Fallacy is different, it is about arguments that look right. DN mode of explanation: Deductive Nomologic: meaning that explanation is logically deducted from nomothetic premises; argument is based on general laws and antecedent conditions. An DN explanation consists of statements of two kinds: explanandum and explanans. The explanans is a class of sentences. This indicates the qualitative character of the DN mode of explanation. However, it can be applied to natural and social science, and therefore it should be possible for the event-to-beexplained (explanandum) to be a formula on a relationship. The explanans are then the conditions on which this relation or formula occur. Example of quantitative explanation: general law: Y= MC2 , Antecedent conditions: M=5, C= 2. Thus Y=20. 1.2 MS: paradigms or research programmes? Does Management Science (MS) consists of fashions / paradigms or of progressive research programs? Use of scientific theories, models and hypotheses: lead to more accurate descriptions, predictions and explanations of reality. Worries: 1) in management the goal is not just to understand reality, but to improve it. 2) “social” reality is much harder to describe and explain, and almost impossible to predict. Rationality: to claim that x is rational is to make an evaluative or normative claim. It is to say something about how things (our thought and actions) should be (irrespective of how they are). Traditional view: science is a rational activity that slowly progresses towards the truth (Lakatos’ programmes: “science is rational”). Scientific thought is guided only by evidence, experiment and its rational evaluation, not a 2 matter of opinion or prejudice, not affected by emotion or society. Research programme (RP): compromised of: i) a hard core: basic theoretical assumptions, ii) protective belt: auxiliary factual and theoretical assumptions, and iii) heuristics: rules that guide the modification of the protective belt in order to protect the hard core from anomalies (neg: don’t change the hard core / pos. how to change protective belt so as to explain anomalies away). How science works (Lakatos): scientist work within RP, there can be more than 1 programme within a given field at a given time, programmes develop determined by its pos. and neg. heuristics, one programme can outperform another: thus possible to make an objective, rational decision about which is preferable. Rationality of RP evaluation: RP A is rationally preferable to RP B (A is and is more progressive than B), RP is progressive when it continues to predict and explain new phenomena. Alternative view: there is no progress towards the truth in science, but just a sequence of hypes and fashions, subject to social processes (Kuhn’s paradigms: “science is irrational”). Kuhn’s view: scientific thought is not just a matter of evidence and reason: which scientific framework / paradigm one accepts is (partly) irrational. Paradigm: is a scientific achievement (way of viewing and investigating the world) that is: i) sufficiently unprecedented so as to attract adherents ii) open ended (by adoptions there remain unresolved research problems), iii) will involve some of special methods of inquiry / instruments, characteristic types of explanation and argument theories and laws, exemplars, normative or prescriptive assumptions. Scientific (paradigm) lifecycle: pre-scientific investigation (absence of paradigm, slow progress) normal science (paradigm is adopted and researchers are either converted to it or marginalized, activity of puzzle solving, failure to resolve puzzle (anomalies) is blamed on scientist not the paradigm) revolution (anomalies accumulate and change of paradigm might occur, + something special, furthermore a paradigm is not abandoned unless a competing paradigm is available) normal science. Theory-laden nature of observation: Kuhn: what we report depends on our standards and theories (paradigms), does not allow comparison of paradigms: rational paradigm evaluation is not possible. Incommensurability: “not comparable by use of a common (not only different but also objective) standard or measure”: competing paradigms have different stands of justification and evidence, paradigms lead to incompatible conclusions: no neutral or objective way of deciding which is correct, proponents of different paradigms inevitably beg the question against each other. The two aspects of the incommensurability problem are: M.A.J. Kusters (0557541) i) people from different paradigms are not able to communicate with each other. They use terms in different ways. ii) even when communication is possible, people in different paradigms will use different standards of evidence and argument. This second aspect of incommensurability refers to theory ladenness. This causes the phenomenon that when scientists work in a paradigm, they cannot objectively assess another paradigm. Because their paradigm shapes their world and includes their own standards for what counts as a good argument/research. Incommensurability refers to comparisons between paradigms. Theory-ladenness refers to something inside a paradigm. So the unit of analysis is different. So theory-ladenness is one of the two aspects of incommensurability. The communication aspect can be solved, so theory-ladenness is the most important aspect for Incommensurability. Theory-ladenness is therefore a stronger argument for Kuhn’s view. Relativism and social constructivism: what is reasonable or counts as known depends on one’s paradigm. Since the choice for a paradigm cannot be rationalized; Irrationalism: choosing basic assumptions cannot be reasoned. Kuhn is ambiguous about the one-paradigm-per-field issue. But the number of paradigms it is not essential for his view as long as they are all incommensurable. When they are, there is no rational comparison possible. Abrahamson’s “Management Fashions” (1996) addresses the question the rationality of MFs: i) rationality plays a role in his definition of MF’s: “A MF is a transitory collective belief, disseminated by management fashion setters, that a management technique leads rational management progress.”, ii) Abrahamson studies MFs as social phenomena, but examines the role of “norms of rationality” and of “techno economic forces”. One tool to sell MFs is rhetoric: advertise technique as increasing rationality by successful examples, empirically validated theories. Popularity of MF is described by a bell shaped curve, explanation: i) social factors (need to seem cutting-edge, confirmation of status, boredom), technical / economical factors (change of economic or political situation). MF: not all fashions may possess the same ‘rationality’ to adopt. To argue that a MF was adopted on an irrational way, you have to come up with standards from the MF and its predecessors. You also have to show that there is no objective standards on which managers could base their choice. To argue whether a MF is a RP (rational) you should know the standards of it and its predecessors, the anomalies that happened during the use of the predecessors, and whether the MF is capable to deal with these problems. You should be able to show objective standards to compare the MF with its predecessors to argue against incommensurability. EXAM QUESTIONS: MAR06 0LM05 – Exam preparation 3 1. Consider the following argument in a student paper. The students argue that adopting the management technique “total quality management” (TQM) cannot be rational: “When we look to the evolvement of TQM we can see a pattern similar to a paradigm shift. After World War II, US manufacturers found themselves in heavy competition with Japanese firms. At first, they thought that Japanese success was price-related, and thus responded to Japanese competition with strategies aimed at reducing domestic production costs and restricting imports. This, however, did nothing to improve American competitiveness in terms of quality. As years passed, price competition declined while quality competition continued to increase. By the end of the 1970s, the American quality crisis reached major proportions, attracting attracted attention from national legislators, administrators and the media. The chief executive officers of major U.S. corporations stepped forward to provide personal leadership in the quality movement, developing the technique that became known as TQM. What we witness here is the sudden change from the conventional paradigm to the TQM paradigm. The American economy suffered greatly from the “Japanese way of doing” without really understanding why. The reason for this is that the Americans lived in a completely different paradigm with different values, beliefs, and rules of conduct which didn’t resembled the Japanese paradigm. But when there became more and more anomalies, which they couldn’t resolve within their paradigm, the faith in the conventional paradigm had declined so much that there was no other way than embrace the new paradigm.” Additional information should be given regarding anomalies of earlier techniques and TQM’s (in-)capacity to solve them; and about central standards of TQM and predecessors, in order to argue for/against incommensurability. a) Briefly explain the Kuhnian notion of ‘paradigm shift’. Use the notions ‘anomaly’, ‘crisis’ and ‘rationality’ in your answer. This comes straight from the Godfrey-Smith text in the reader. Do not say that anomalies cause crises in some rational way. There are always anomalies, and crises are irrational phenomena. Sometimes, they lead to paradigm shifts in which, of course, there is no rational transition from one way of doing science to another. b) Explain how the change from earlier management techniques to TQM might involve a paradigm shift. The change might be a paradigm shift, because TQM might be incommensurable with earlier techniques. Only if there is no rational way of comparing the techniques, there is an irrational transition. c) The argument given by the students is not sound. Explain why, by arguing that the change from earlier management techniques to TQM need not involve a paradigm shift. Follows more or less directly from previous answers. TQM might be comparable with earlier techniques. Indeed, the student description strongly suggests that it arose out of problems experienced with the earlier techniques; if it solved these problems (and did not suffer “Kuhn loss”, i.e., problems that earlier techniques solved suddenly without solution), transition to TQM is management progress. d) What kind of information would be needed, in addition to that given by the students, to show that the transition to TQM is not rational? JAN07 1. in a paper written for this course, students argue that “total quality management” (TQM) can be characterized as a Lakatosian RP, and that its adoption can be explained as a rational decision. First they offer the following characterizations: “Hard core: the core of TQM is the customer-supplier relationship where the processes must be managed. The ‘soft’ outcomes of TQM – The culture, communications, and commitment provide the foundations of TQM. Protective belt: the Protective Belt of TQM consists of the following factors: committed leadership, adoption and communication of TQM, Closer customer Relationships, Closer supplier relationships, Benchmarking, Increased training, Open organization, Employee empowerment, Zero-defects mentality, flexible manufacturing, Process improvement, Measurement.” In another paper, the following anti-Kuhnian argument is given: “According to Kuhn, only one paradigm at a time is right for a certain topic. But in management, several mutually exclusive paradigms can co-exist in a continuing state. For instance, at a particular time, in different companies, different techniques for improving quality are used. These co-existing techniques cannot be paradigms, so Kuhn’s model makes no sense. e) Explain why the first sentence in this quote seriously misrepresents Kuhn’s views and rephrase the sentence in a Kuhnian way. The big mistake is that paradigms are not “right” for a field. There is no absolute truth, but this statement says that there is (but it changes over time?). One reformulation is: “According to Kuhn, one paradigm dominates each field at a certain time.” f) A Kuhnian can easily respond to this argument. Show how, and examine whether the argument can be strengthened in such a way that this response is blocked. Although critics like to point out that Kuhn often suggested a one-paradigm-per-field view, he is at least ambiguous in this respect. And it is not essential that there is only one paradigm: as long as they are all incommensurable, rational comparison is still impossible. a) Briefly characterize a Lakatosian RP, including the notions ‘hard core’ and ‘protective belt’. Use the phrase ‘succession of theories’ in your answer. A RP is a historical entity; it evolves over time. It will contain a succession of (related) theories. A RP has two main components, in Lakatos’s view. First it contains a hard core. This is a set of basic ideas that are essential to the RP, and these ideas are similar in every theory that is part of the program. Second, a RP contains a protective belt. This is a set of less fundamental, auxiliary claims and ideas that are used to apply the hard core to actual phenomena and to keep it from being falsified. The 4 protective belt may undergo changes in the development of the RP. b) Assume that the students list actual factors of TQM. Have they succeeded in characterizing a hard core and protective belt, according to your definition in 1.a? As indicated in 1.a) both the hard core and the protective belt consists of claims. A claim is a statement that is either true or false, for instance: “there are seven planets in the solar system”, or “managers ought to be committed”. However, the things listed by the students are not claims, only factors like “culture”, and therefore not appropriate elements of a hard core or protective belt. Thus, the student have not succeeded in characterizing a hard core and protective belt. The students also give some arguments for the rationality of adopting TQM: “The proof of rationality in adoption of TQM in Japan is straightforward. It was invented as a result of their poor economy. American manufacturing industry realized their competiveness became very weak compared to Japanese manufacturing industry. They began searching the reason for their decline. They compared their traditional approach to management with the Japanese approach. They found out that what TQM had done for the Japanese, could also be done for them. Therefore they have adopted TQM. Another argument for the rationality of the adoption of TQM is the fact that it has evolved after the implementation. First, only the manufacturers adopted TQM, but later the nonmanufacturers followed.” c) Evaluate the arguments in this short section. Do so separately for each argument. If you think the arguments are fallacious, characterize the fallacy involved. 1) “[…] adoption […] is strainghforward. It was invented as a result of their poor economy.”: The clear cause for invention (poor economy) does not make it a rational decision (taking LSD may cause you to believe that you can fly, but it does not make jumping off a building a rational decision); 2) “[…] what TQM had done for the Japanese, could also be done for them”: at best a very weak inductive claim; it is reasonable to expect all kinds of dissimilarities; 3) “Another argument […] is the fact that is has evolved after the implementation”: this is not rational; evolution is just change, and it might be change for the worse. In another paper students argue for the incommensurability of TQM and another technique, “Business Process Re-engineering” (BPR). One argument they give is announced under the heading: “theoryladenness of observation”. d) Describe the phenomenon of theory-ladenness of observation, preferably with a small example. Why does the phenomenon support the Kuhnian view of science? Theory-ladenness of observation means that your background beliefs or theory influence your perception of reality. When you believe that Jupiter has moons, you are prone to interpret a blurry image in this way. The phenomenon supports the Kuhnian view because it means that researchers in two paradigms literally see different M.A.J. Kusters (0557541) things when they look at the world, perform experiments, etc. This makes resolving their debate far more difficult. The students present the following argument: “The [two techniques] have a different focus, TQM promotes improvements in performance by statistical methods and BPR promotes improvements in performance by applying an IT structure. The common element in both approaches is their focus on performance improvements through improvements of processes across the whole company. But the term ‘process’ is used differently in each approach, because the focus and the manner of change are different. In the TQM approach, the term ‘process’ refers to a way of splitting the organization in order to improve quality, while the BPR uses the term to implement workflow technology. The concept ‘process’ is a good example why researchers cannot rationally compare paradigms. When followers of both paradigms discuss the term ‘process’, confusion will arise because they give a different meaning to the same world.” e) How effective is this argument for the theory-ladenness of observation? The argument may be effective, but not for theoryladenness. TQM and BPR define a single notion in different ways, but these different definitions may not affect perceptions of reality at all. f) The students conclude from their argument that “the concept [of process] cannot be discussed rationally”. Write a short response from a Lakatosian / Laudanian perspective, defending rationality. Since both definitions are (presumably) clear, the opponents can at least consult each other’s results, and interpret them in the light of the alternative definition. Furthermore, they both aim at improving performance, albeit by different means. Therefore, they have a shared goal, which makes comparison possible: it is at least conceivable that a neutral party investigates which of the two means actually succeeds most in improving performance. JAN08 1. consider a a hypothetical management fashion, Punish for Profit. Under a. to d., four arguments are offered in favour of, or against, the introduction of Punish for Profit in a specific organisation. Discuss the validity of each of these arguments separately. Examine whether the argument is valid or fallacious, and argue why. a) "Introducing Punish for Profit is irrational, because it would decrease job satisfaction among employees." INVALID. There is no premise that states that it is irrational to introduce any MF that decreases job satisfaction among employees. Thus there is a missing premise. The missing premise is: “it is never reasonable to decrease job satisfaction”. b) "Introducing Punish for Profit is irrational. All studies that show its advantageous effects have been done by consultants who make millions by helping organizations introduce Punish for Profit." INVALID. Ad hominem argument: appeal to the position of the speaker “You are so stupid; your argument couldn’t 0LM05 – Exam preparation possibly be true”. In case above, that the consultants are making millions by helping organizations introducing Punish for Profit does not mean that the results of their studies are incorrect (and if they would be biased and present advantageous effects for that reason, still these or other advantageous effect could be really present). Thus criticizing the authors of the article instead of their work misinterpreting the situation. c) "Introducing Punish for Profit is rational. It performs better than our current management technique, which does less to increase customer satisfaction." [NB: Analyze this as a pragmatic argument] INVALID. Form of pragmatic argument: i want A, doing M is a means to bring about A, therefore I should do M. In this case A would be the highest goal for on organization, which is profit. To increase customer satisfaction could be a (one possible) means of attaining profit. Introducing Punish for Profit seems to be one possible mean of attaining the intermediate goal of increased customer satisfaction. Note first that these arguments are never valid. The conclusion is normative, and the premises descriptive, so it does not follow that I should do something. I might also conclude that I will not satisfy my desire, or that I abandon it altogether. Critical questions: (1) What other goals do I have that might conflict with A? (2) What alternatives to doing M would also bring about A? (3) Are some of these alternatives more efficient ways of bringing about A? (be better than Punish for Profit) (4) Is it practically possible for me to do M? (possible problems of implementations are unknown) (5) Are there negative side effects of doing M? (may be there, is also unknown) d) "Introducing Punish for Profit is rational. It closely resembles the technique that we are using now, so it may be regarded as the next stage of a Lakatosian research program." INVALID. The fact that the technique closely resembles the present technique does not support the conclusion that the introduction of Punish for Profit is a rational next step. e) Give a question-begging (circular) argument in favour of Punish for Profit. Premise: all other MFs underperform Punish for Profit. Conclusion: Punish for Profit is the best MF. f) Can an argument be valid without being sound? If not, argue why. If so, give an example. YES. An argument can be valid without being sound. Validity: the argumentative structure can be correct, with the conclusion necessarily following from the premises. Not sound: one or more of the premises can be untrue (also making the conclusion untrue). When both apply, the argument is valid but not sound. Example: premise1: all charismatic people are a (former) chairman of study association Industria; premise2: Ronald is a charismatic person; conclusion: Ronald is a (former) chairman of study association Industria. In this example the argumentative structure is correct, however one of its premises (premise2) is untrue (Ronald is more like a dictator ;-), therefore making it an argument which is not sound. 2. Rigour and relevance 2.0 Practicalities 5 Review and discussion question 1-4 Daft, Lewin, Shrivastava, van Aken, Bertrand & Fransoo, Humpreys 2.1 The articles Should we aim for scientific rigour or for practical relevance in MS? Diagnosis of the controversy: multiplepersonality disorder. Case for rigour: “nothing is so practical as a good theory”. “Knowledge is power”close tie between rigour and relevance. “By definition science is rigorous, otherwise it is just common sense”. Case for relevance: social phenomena are so complicated, you cannot make predictions about them. People have a perspective = interpretation, by formulating and framing you lose that aims, ideas, etc: aim should be to improve actual organizations to solve existing practical problems. Because of the complexity issue this can only be realistically done on a case-to-case basis. This calls for open-mindedness about approaches: no straitjacket of “normal science”. Issues at stake: five ideas / distinctions in the clusters: i) law-based (general) vs. ideographic (specific), ii) formal vs. informal, iii) theory-based vs. data-based, iv) descriptive vs. prescriptive, and v) science (research) vs. engineering (design). I) Laws (nomothetic): scientific explanation uses laws: general, causal statement of a quantitative relationship between variables. Laws don’t have to be exceptionless, but can have ceteris paribus clauses (hold so long as no interfering factors are present). I) Ideographic: looking for accurate descriptions, not for general explanations. There could be a spectrum, ranging from purely nomothetic, via hybrid to purely idiographic sciences. Where to put management studies is not clear. II) Formalization: laws are quantitative, but not all quantitative methods are law-based (optimization, statistics). Non-laws based models: thought of as tools which do not “show the system as it really is”. II) Problems with the use of models: every model of a complex system involves idealizations (applicable to many different cases): miss out (very important) factors (abstraction), assumptions needed for mathematical tractability, less controversial what “force” means than what “wealth” or “performance” means. Real problem is robustness: adding influencing factors (in social sciences) will not get something that is approximately correct. II) Mathematical modeling (Humpreys): difficult in social sciences: i) most traditional models consider the individuals to be homogeneous, each individual in the population is considered to have the same basic properties as any other, and ii) from the start a model is defined followed by altered behavior of individuals. Individuals react differently to a proposed “fixed model”. But not impossible: i) to effectively model traditional models, it is required to utilize an essentially dynamic approach, and ii) the employment of simple mechanisms is acceptable if there is good reason to hold that these “mechanisms” correspond to genuine features of the 6 systems under investigations. Certainly, quite sophisticated physical models employ radically simplified physical mechanisms to construct descriptively accurate computational models. III) Theory or data-based: does theory or data comes first? Models and data need to be related ((formal) models need verification, otherwise empty explanation). Theory driven: rigorous people want to do this on behave of set of relations among well defined quantities. Using models is using formal methods (some sort of rigour involvement), there is a need, and freedom to engage in stipulation (defining the terms in the model in a rather abstract way) and abstraction. Problems: is only a computational template (won’t give understanding), testability becomes difficult. Data driven: start from data to find relations between quantities. Problems: step from correlation to cause is contentious (f.i. there could be hidden or intervening variables), garbage data = garbage model. Data trouble: scarcity, not reproducible (systems change), not gathered for the model at hand, perhaps unreliable. So theory-driven models might be too abstract and datadriven models too unreliable to be of any use. IV) Science is not enough: we need description (how the world is), but also how to handle the situation = prescription (how the world should be (like normative rules), and how you ought to behave). An accurate analysis of the current situation and, if possible, an explanation of the current crisis seems necessary to change things for the better. So description without prescription is sterile, but prescription without description is blind. V) Science and engineering: main goal in engineering design is changing reality or maintaining some favorable state of affair (prescription is thus essential). Design: aimed at the artificial (to make), problem-oriented (goal is to contribute to the solution of a practical problem), typically involved sets of prescriptions or rules. Designing often involves performing a fixed series of actions, following a determinate set of rules: form of a rule: ‘if you want to achieve Y in situation Z, then perform action X’. X is the solution concept for a type of problem, the core of the rule. The remainder of the rule specifies application conditions and relevant goals. Daft and Lewin – heretical research (departure of from established beliefs): The point of heretical research methods is to find new channels through which to obtain organizational insights and to change the mix of research methods. Although no method is truly heretical, researchers should be encouraged to do whatever it takes to learn about organizations. Measuring the gap: there may be a gap, even if it is merely one of degree. Shrivastava found and measured the gap (criteria are clustered) in strategic management: three criteria for rigour (conceptual adequacy, methodological rigour, accumulated empirical evidence) and five for relevance / usefulness (meaningfulness, goal relevance, operational validity, innovativeness, cost of implication). There are thus two sets of standards at work in strategic management research: the field has a “dual orientation”. Suggestion of Shrivastava: gap may be closed, and interaction between rigour and relevance, not a clearer distinction, would be a sign of real progress. M.A.J. Kusters (0557541) Rigorous research vs. heritical research: i) conceptual adequacy: doesn’t fit with heretical research: the researcher gives an own interpretation of a study which is not generally supported by the research field. I.e. the research program is not well grounded in the initial base discipline. The conceptual framework is not consistent any more with the existing theories in the field, ii) methodological rigor: this will remain unchanged although the results are likely to interpret differently (think of using the outliers in research instead of the main stream results), and iii) accumulated empirical evidence: nothing will change for the empirical evidence. Applicability of Shrivastava to research: Shrivastava’s criteria are applicable to both nomothetic as ideographic research, although it is far more difficult to reach rigour in ideographic science. In contrast with nomothetic were rigour is reached easily while being relevant (for example physics in the construction industry). The same explanation holds for research-like and design-like studies. Mutually independency of Shrivastava’s criteria: not for the following criteria: A) Conceptual adequacy vs. methodological rigor: conceptual adequacy also is about the methods chosen for empirical examination of research questions. B) Methodological rigor vs. Empirical Evidence: when the methods to empirically examine research questions are not right, the evidence cannot be empirical accumulated. C) Meaningfulness vs. goal relevance: if the goal is not relevant, how can the study be meaningful? D) Negative correlation between the rigour criteria and goal relevance and operational validity. (as stated in the text: “the tendency of researchers to generate non-specific, generalizable and broadly applicable knowledge reduces the operational validity and goal relevance of research results”). E) Innovativeness vs. meaningfulness: when a result is obvious, the meaningfulness will be less. Shrivastava with respect to rationality of RP in MS: Shrivastava is a Lakatosian: i) he considers more research programs in one period, ii) he thinks the subject is open for discussion: “these criteria reflect a synthesis of ongoing debates in management literature on the issue of rigor versus relevance in research “, iii) He gives objective criteria to make a rational choice about whether a research study is rigour or relevance. With these criteria research studies can be compared and they can no longer be considered incommensurable. EXAM QUESTIONS: MAR06 a) List Shrivastava’s criteria for evaluating the rigor of a research project. Briefly describe each of these criteria. Conceptual adequacy is the framework that guides the research program and shows whether the project is guided by a theoretical framework; a function of how well the research program is grounded in theoretical frameworks provided by underlying disciplines. It measures the extent to which research programs apply the knowledge of developed in their base discipline to generate theoretically interesting issues and choose appropriate research settings and methods for empirical examination of research questions. 0LM05 – Exam preparation Methodological rigor: refers to the extent to which qualitative or descriptive data is used versus analytical modeling methods and objective quantifiable data to empirically examine research questions. Accumulated empirical evidence: supporting the theoretical structures. The accumulated evidence lends credibility to research findings and legitimizes them to other researchers and managers. It provides a basis for accepting research findings as being empirically validated and thereby grounded in objective or projected reality. Meaningfulness: refers to the comprehensibility of research results to users. It is a function of the extent to which research findings capture and adequately describe organizational reality and whether the findings actually present what the problem-owners within the organization wanted to be explored (Shrivastava, 1987). Goal relevance: the extent to which primary variables are relevant to organizational and managerial goals Operational validity: or actionability, the extent to which research results are operationalizable through concrete actions (Shrivastava, 1987). Innovativeness: or non-obviousness research results. New and non-obvious research results provide the organization with innovative knowledge which can be used to differentiate themselves from competitors. Cost of implementation: of research results. b) Describe the distinction between a design-like and a research-like project. Can a rigorous project be designlike? Design-like projects aim at solving a practical problem; research-like projects exclusively aim at prediction, description and explanation. A rigorous project can be design-like if it is also relevant. This may be difficult to realize, but the second assignment was aimed at showing that it is possible. Now read the Appendix, which contains a short summary of a Master project. c) Evaluate the rigor of this project, using Shrivastava’s criteria, and describe whether it is research-like, designlike or both. If you think data are lacking to “score” the project, describe these relevant data (“If the summary would have described X and Y, I would have been able to apply criterion Z”). - The conceptual adequacy of this project is mediocre at best. No framework of generally accepted scientific concepts is used, the researcher constructs his own. It is also entirely unclear what the underlying theoretical discipline is: this is almost phenomenological research. - Methodological rigour: the researcher does use several accepted methods for his (highly) qualitative research, so I would charitably score it “high” on this criterion. - Accumulated empirical evidence: this is a tough one. The researcher did not gather many data, but perhaps this is not necessary for qualitative research. I hesitate to score it “high” here, but since this may just reflect my scepticism regarding qualitative research, let’s say that I would like to have more data. In questions such as this one, many answers are defensible. We will look whether you understand the criteria, and are able to apply them in a well-argued way. 7 In their paper “Can Organization Science begin to break out of the normal-science straitjacket?” (1990), Daft and Weick argue that organization studies have “prematurely settled into a normal-science mindset. This seems inappropriate because organizations are complex, variable-rich phenomena that can be studied from multiple perspectives” (p.2). They promote the development of “heretical” (ketterse) research programs, which search new channels for insights into organizations rather than build upon existing research (pp.6–7). d) Discuss to what extent this emphasis on heretical programs means that Daft and Weick reject rigorous research, as defined by Shrivastava’s criteria. The emphasis on heretical methods means that conceptual adequacy and accumulated evidence might well remain very low for all research in the field: there are no generally accepted frameworks, and no method can prove its merits in a sustained manner. Thus, fragmentation of the field (or “pluralism”) at first glance decreases rigour. Make sure you answer this question referring to Shrivastava’s criteria, and not just with the aid of common sense – there is nothing wrong with common sense, but the question was about the criteria. JAN07 Read summary of Master Thesis “The effect of national culture on business-to-business relationships” (2005), by Victor van den Broek. a) Describe the distinction between nomothetic and ideographic research. How can the project of van den Broek be described in terms of this distinction? Nomothetic research: aims at the formulation of laws (i.e., universal, explanatory statements); Ideographic research, aims at descriptions of specific phenomena that are as accurate as possible. Although van den Broek wants to include differences between companies in his model, he ultimately aims at a general model of the relation between B2B relations and national culture. This may fall short of formulating laws, but it is closer to the nomothetic ideal than to ideography. b) Describe the distinction between theory-based and data-based modeling as introduced by Paul Humphreys. Is the Master project of van den Broek theory-based or databased? Theory-based modeling: starts from an existing theoretical framework and derives a model from it, which is subsequently confronted with data; Data-based modeling: starts with collecting data, and constructs a model directly from the data. Van den Broek uses Hofstede’s existing model of national culture, and also constructs his own – not from data, but from literature review (which, presumably, does more than list data). Therefore, his modeling is theory-based. c) Evaluate the rigour and relevance of van den Broek’s project, using Shrivastava’s criteria. If you think data are lacking to “score” the project, describe these relevant data (“If the summary would have described X and Y, I would have been able to apply criterion Z”). 8 1) Conceptual adequacy: van den Broek uses Hofstede and existing literature to construct his own model, and therefore scores “high” on this criterion. 2) Methodological rigor: we do not get lots of information on the construction of the model, so the “analytical methods” remain unclear. The only data used come from interviews, and are probably highly qualitative. At best a “mediocre” score. 3) Accumulated empirical evidence: there is no evidence whatsoever for van den Broek’s own model, but the criterion is primarily about the program it is part of. Presumably, there is evidence for Hofstede’s model and the other literature used – van den Broek does not choose an entirely new approach. Somewhat mildly, a “mediocre / high” score. 4) Meaningfulness: the concepts used in the model, such as “culture” and “trust” are probably meaningful and understandable for decision-makers, but there is no connection to a specific strategic problem facing these people. So “mediocre” at best. 5) Goal relevance: there is probably some connection to manager’s goal, but no performance indicators. So “low / mediocre”. 6) Operational validity: the final sentence of the summary is perhaps an action implication, but it is not very clear and difficult to implement. “Low”. 7) Innovativenessh: there may be more information in the full thesis, but the claim that B2B relations are influenced to national culture and that it is important to realize this is very trivial. “Low”. 8) Cost of implementation: because there are no action implications, I would say that this criterion does not apply. In sum, this thesis scores higher on rigor than on relevance, but that says more about the lack of relevance than about the excess of rigor. d) Suppose van den Broek had been able to carry out his quantitative project. How would this have affected his “Shrivastava scores”? Carrying out a quantitative project would imply that van den Broek would have scored higher on “methodological rigor”: there would be more quantitative data. Furthermore “accumulated empirical evidence” might increase slightly, and there might have been more material for performance indicators and action implications; but the latter two are speculative. In sum, rigour would have increased. e) In the course, five “interpretive” phenomena were examined that might affect research in the social sciences. Describe for at least one of these phenomena how it might have undermined van den Broek’s research into interaction between national culture and business-tobusiness relationships. The project provides ample opportunity for hermeneutic circles. Van den Broek brings his own prior understanding of important aspects of national culture (the Hofstede dimensions) to the interviews. And although he does eliminate some elements of his model after the case study, nothing is added to it, meaning that the model is treated as the correct framework for the interpretation of all further cases……. JAN08 M.A.J. Kusters (0557541) Read summary of OPAC Master thesis “Recovery chain planning: Allocating used products to the appropriate recovery option” (2000), by Martin Muysers. a) In their 2002 paper in the reader, Bertrand and Fransoo follow Mitroff et al. in distinguishing four steps in the operational research method: (1) conceptualization; (2) modelling; (3) model solving; (4) implementation. Can each of these steps be found in Muysers' thesis? If so, point out where. If not, discuss whether the absence of this step indicates a lack in scientific rigour. 1) conceptualization is found in the sections “background” and “The Multis chain”, where Muysers attempts to make a conceptual model of the problem and system he is studying. Muysers makes decisions about the variables that need to be included in the model, and the scope of the problem and the model to be addressed. 2) modelling is found in the sections: “value cost decision tool”, “recovery chain planning”, and “information requirement”. However it is not clear if Muysers has build a quantitative model. Furthermore concerning step 3) model solving, there is no information given about the mathematical structure and the mathematics involved in Muysers’ model. Muysers’ thesis also lacks on implementation, the summary ends only with some recommendations (which seem to indicate that Muysers did derive results from his model). The absence of how Muysers derived results from his model, which is step 3) model solving, indicates a lack in scientific rigour, especially in the criterion methodological rigour. Furthermore the work done remains very conceptual and abstract without step 4) implementation. It is still unclear to what extend the model of Muysers is working for Baan, which according Shrivastava indicates a lack of relevance. Last concerning the implementation step: testing the model supports the theoretical structure, which supports its scientific rigour. Also on this point Muysers’ Master Thesis seems lacking. b) According to Van Aken (2005), also included in the reader, management research should end in one or more field-tested and grounded technological rules. Explain the meaning of "technological rule", "field-tested" and "grounded". Technological rule: “a chunk of general knowledge linking an intervention or artefact with a desired outcome or performance in a certain field of application”. It has the form: “if you want to achieve Y in situation Z, then perform action X”. A technological rule is field-tested if “it is tested in its intended field of application”; this entails multiple tests or cases in which the rule is tested. A technological rule is grounded if “it is known why the intervention or artefact gives the desired performance”; grounding can be done with insight from the social sciences. c) Can (one of) Muysers' recommendations be regarded as a field-tested, grounded technological rule in Van Aken's sense? If so, show this. If not, discuss whether the lack of technological rules shows that Muysers' study is not design-oriented. 0LM05 – Exam preparation Muysers’ recommendations vs. field-tested: no evidence of testing is shown (see answer question a), thus also no multiple tests are carried out. Muysers’ recommendations vs. grounded: the concept of a generative mechanism does not apply to Muysers’ recommendations: the question “why does this intervention (in this context) produce this outcome” remains unanswered. Despite the lack of technological rules, Muysers’ study is a design-oriented study: about the practical problems of one company and Muysers tries to contribute to the solutions of these problems. However, to what extend Muysers solves the practical problems is unclear, but it is certain that his recommendations cannot be regarded as technological rules. 3. Positivism and interpretationism 3.0 Practicalities Review and discussion question 1, 2, 6 and 7. Weick, “Organizing and the process of sensemaking” Scherer, “Modes of explanation in Organization Theory” Schutz, “Concept and theory formation in the social sciences” Nagel “Methodological problems of the social sciences”. 3.1 The articles Is MS about the explanation or about understanding of phenomena in industrial organizations. View of Social Sciences: there is a tangle of (probably) related and (certainly) conflicting views according to which social science cannot be as objective and disinterested as the natural sciences. Suggestive view: social science is subjective because: i) human beings study other human beings and their organizations, ii) these are very complicated and dynamic, iii) humans are conscious of themselves, of each other, and of their environment. Electrons and anemones are not, and iv) social sciences are always done “from the inside”: social scientists are part of the societies they study. BUT, does this justify thinking that social sciences cannot be objective. Social positivism view: i) there is a causal structure that underlies phenomena, ii) explanations point out such fixed or dynamic structures, iii) idealizations are needed and acceptable ways of constructing general explanations, and iv) the social sciences should be modeled as much as possible on the natural sciences. View: people’s thoughts do not matter for the underlying causal structure of society, social scientists should only study this underlying structure. Aim: explanation, not understanding. Positivist position: People are constrained by “situational imperatives”; they should increase an organization’s efficiency or fail. What they think or want is irrelevant to explaining what happens. So looking inside people’s heads, or talking to them, let alone constructing meanings, is a waste of research time. 9 Interpretive view: “Human beings are different from dead matter and even most animals because the world is meaningful to them, and the social world is different from the natural world because it is based on meaning.” Core: when we want to describe social phenomena, we have to grasp their meaning = “understanding” or “interpretation”. The question lies around the interpretation of social phenomena: it is not clear whether the social sciences need their own method. Method for interpreting meaning: needed because meaning is not perceived directly (= underlying). Hermeneutics is the study of (the real) meaning. Constructing meaning: i) social sciences: use hermeneutics to reveal the underlying meaning of social phenomena, versus ii) natural sciences: use some other method to reveal the underlying causes of natural phenomena. Difference: meanings cannot be revealed like causes. Via interpreting, you make something meaningful to you, thus construction (based in presuppositions, assumptions and prior knowledge) rather than revealing. Meaning and objectivity: If correct, this means that objectivity is lost: the social sciences do not study some fixed order, independent of the researcher. Instead, the researcher interprets phenomena on the basis of his/her own ideas. Even changing these ideas is a constructive activity. There is at most intersubjectivity: a community of researchers agreeing on an interpretation of phenomena. This supposedly affects the methodology of scientific research: there are no objective data, no independent validations, no general explanations. Interpretive studies: interpretativists study how people make sense of their circumstances (involves participating in the observed processes). Interpretivism and rigorousness: Shrivastava considers research to be rigour if it scores high on his criteria for rigour, and low on the relevant criteria. For interpretative research: 1) conceptual adequacy – low, 2) methodological rigor – mostly qualitative data, so low, 3) accumulated empirical evidence – single case studies, low, 4) meaningfulness – high, 5) goal relevance – high, 6) operational validity – quite high, 7) innovativeness – uncertain, 8) cost of implementation – not applicable. Concluding: interpretivist research is mostly not very rigorous. Changes to the method, for example gather data with the same approach in more situations, use common methods to describe phenomena and so on can make it more rigour. Interpretivism and qualitative data: interpretivist research doesn’t necessarily have to be qualitative, although it is hard to gather quantitative data for this approach because it is concerned with the interpretation of social phenomena and. Magnan associates interpretivism with only qualitative data. Interpretivism with respect to rationality: according to Lin, interpretivists and positivists live in a different paradigm. They are incommensurable as she states, although she doesn’t clearly speak of paradigms. To be a Lakatosian, an interpretivist accepts more than one competing research program to exist. Their hard core would consist of the interpretive techniques by which they study phenomena, and the way they gather data and draw 10 conclusions about it. It seems that something about rationality is missing. Interpretivism and positivism with respect to paradigms: The interpretivists and positivists each look in a very different way at a certain phenomenon. The positivists think that science can be objectively observed and general laws can be derived from studies. The interpretivists on the other hand consider social phenomena as meaningfull, and that naturalistic techniques pass by the meaning. This difference in believes leads to very different approaches in research regarding data collection, and how the data is interpreted. Both schools will ask very different questions to study a problem, and thus observe different things. This can be seen as theory-ladeness, causing incommensurability. Either of the schools will not understand the other one’s practices, given that they are in their school. Therefore we can in fact see interpretivism and positivism as paradigms. This also has some implications for the discussion between the approaches. There is incommensurability, so either of the two groups will not accept or understand the way the other group does their research. Interpretive arguments against positivism: 1) people respond to purported social laws, so these are violated, 2) explanation (in terms of underlying causal structure) and idealisation are just styles of interpretation, 3) Twoparadigm model: we have here two different paradigms, so it is impossible rationally to decide between them, and 4) interpretationists point out phenomena that positivists cannot accommodate. Schutz and Numagami’s arguments against positivist research: Schultz: 1) if there is a theory predicting human behavior, than it will not tell us anything about social reality as experienced by men in everyday life, it will be highly abstract and the concepts will be far from reality, and 2) the object under study in the social science, (social reality) has its own meaning: the actors in this reality have already interpreted the world they experience as the reality. -> inquiry can’t be objective because of double hermeneutics. Numagami: 1) human conduct creates and recreates lawlike regularities in social phenomena, so they are not ‘out there’, 2) human beings have some capacity of reflection on their own thinking and that of others , which enables them not to behave as a lawlike regularity would predict, and 3) methodological argument based on validity of lawlike regularities as support for invariant laws. Differences: main differences in arguments are that Numagami uses a researcher point of view, and elaborates on the mismatch of object under study and the application of lawlike regularities, while Schulz argues from a field perspective and indicates that human behavior can’t be modeled or predicted, and therefore no lawlike regularities or invariant laws can excist. He also argues that a theory on social reality will be inconsistent in itself. Phenomenon 1: enactment: the process of creating social phenomena through action and preconceptions (e.g. creating an emergency situation). As a result, researchers participate in enactment, they participate in the construction of their subject matter, whether they want to or not (searching for information, selecting relevant information, expressing uncertainty, publication of data, etc.). Eg. Hawthorne effect. Problem: A researcher doesn't M.A.J. Kusters (0557541) know how she'll affect what she is investigating, so her explanatory endeavour is frustrated. Phenomenon 2: priming: the impact that the activation of previously activated representations have on subsequent judgment and perception. Data gathering involves multiple interpretations. The researcher always creates a framework for data collection; comparable with Kuhn’s theory-ladenness of observation. Problem: the aim of investigating a researcher independent world, and thus of uncovering objective findings, is undermined. Phenomenon 3: the hermeneutic circle: the phenomenon that in interpreting something (a text, a situation, an event) we always bring 'prejudices' and preconceptions along that guide our interpretation and influence our first impressions of something. As a result, our interpretation tends to affirm and reinforce our preconceptions; a specific variant is the parts-whole variant, where our initial understanding of something depends on our first impressions (influenced by preconceptions) of some small part(s) of it, and we continue to interpret the rest and the whole in terms of our previous interpretations of its parts, and vice versa. The circularity might be vicious: our faulty assumptions about part and whole might reinforce each other. The circularity might not be vicious: our assumptions might correct each other endlessly. Problems: i) even when one’s assumptions change, one still remains with assumptions, ii) one cannot find out what a phenomenon “really means”, and iii) testing becomes problematic as results depend on interpretation. Phenomenon 4: situatedness: all research is done in a specific context or situation. This context includes f.i. a background of interests, assumptions, and previous experience. Situations always change, and actions and knowledge change along with situations. Formal representations are misleading in this respect: they look fixed, but their meaning and use is situated and changeable. Problem: this undermines the ideal of general knowledge. Phenomenon 5: sense-making and storytelling: most of the time, people act unreflectively. Still, people often need to and do make sense of their actions, e.g. when they justify them to others or themselves (rationalizations). They do so by telling stories (“narratives”): coherent sets of statements, making their motives explicit and comprehensible. These stories are constructed afterwards, but they strongly determine our understanding of what happened, the role we played, the measure of control we had, etc. Claim: every representation of social events is such a story. Problem: So much for objective explanations of social phenomena! End of social science: concluding from the five phenomena: i) there are no objective data ("data illustrate, not validate"), ii) there are no objective tests of hypotheses, iii) people act unpredictably, and iv) scientific explanations of behaviour are just stories to influence behaviour. If this is correct, research about social phenomena would not be predictive, explanatory, objective science. It would just be an ongoing, ever-changing conversation between people. However, positivists argue for the possibility of objective, explanatory research 0LM05 – Exam preparation despite the problems raised by the interpretive phenomena. Strategies for combining positivism and interpretivism: LIN’s approach: Lin makes a distinction between positivists, who’s research is focused on finding causal relations across situations, and interpretivists, who focus on describing causal mechanisms as accurate as possible. Both types of research make the researchers ask different questions, use different methods and draw different conclusions. Lin says, that each of the streams, although they may study the same phenomenon, can complement the other. Positivists can find causal relations between two variables and tell in what kind of other situations they can be observed, but cannot tell how the one variable influences the other. Interpretivists on the other hand can tell this, because their goal is accurate description and studying how and why two variables influence each other, but they cannot tell in which other situations the phenomenon can be found. Lin suggests four strategies: 1) literature reviews quickly narrow one’s scope into method specific research. Better approach is to track down literatures that have the same substantive focus as one’s own question but use a different approach, 2) outline studies from different approaches that would complement their own work, 3) if researchers have a chance to combine two types of research in their work, they should start with an exploratory interpretivist study, and 4) comparative case study makes it possible to incorporate some of the best features of both streams in the same study. MANGAN-LALWANI-GARDNER’s approach: Mangan e.a. offer a different approach, that they call triangulation. They distinguish four types of triangulation: 1) data: collect data from different times and different sources, 2) investigator: different investigators independently collect data, 3) methodological: quantitative and qualitative techniques are employed, and 4) theory: theory taken from one discipline and used to explain in another. Conclusion about differences in both strategies: it is clear that interpretivists and positivists perform their research with very different approaches, Lin even calls it incommensurable. Both approaches criticize each other’s work in obvious ways. Lin’s approaches are centered on the use of different studies from the other approach in one’s study. This requires the researchers to let go their skepticism about the other stream, which may be very difficult. Mangan offers a more hand-on approach with his triangulation. It doesn’t require researchers to take the other approach seriously, but are small useful techniques which can narrow the gap between both practices. EXAM QUESTIONS: JAN06 The debate between positivists and interpretivists is one of the most fundamental controversies in the social sciences. Many philosophers and scientists take one side or another, and some present arguments for their view. Another view is that both approaches can co-exist, more or less peacefully. One way to support this view is by arguing that positivism and interpretivism are different paradigms. A very influential presentation of this argument has been given by Gibson Burrell and Gareth Morgan, in 11 their Sociological Paradigms and Organisational Analysis (Heinemann, 1979). Take, for instance, the following quote: “All social scientists approach their subject via explicit or implicit assumptions about the nature of the social world and the way in which it may be investigated. First, there are ontological assumptions– assumptions which concern the reality and nature of the phenomena under investigation, [for instance] whether reality is a given ‘out there’ in the world, or the product of one’s mind. Associated with these ontological issues, is a second set of assumptions of an epistemological nature. These are assumptions about the grounds of knowledge – about how one might begin to understand the world and communicate this as knowledge to fellow human beings.” (p.1) a) Describe briefly these basic ontological and epistemological assumptions of positivism and interpretivism. In other words: what do these approaches say about the reality and nature of organizations and about our knowledge of them? Ontology of positivism: there are underlying causal structures. Mental attitudes do not feature in this structures as causes (otherwise they should be studied). Ontology of interpretationism: social reality is constituted by mental attitudes and mutual understanding. People jointly construct social phenomena and their meaning. Epistemology of positivism: find out causal structures by hypothesizing social laws, idealizing from particular situations, and ignoring mental attitudes. As much like natural science as possible. Epistemology of interpretationism: impossible to find general laws. Instead describe particular situation, your own role and understanding of it, and participate in social construction of phenomena. Not imitation of natural science. Burrell and Morgan go on to identify a basic distinction between “nomothetic” and “ideographic” methods; they associate the first method with positivism and the other with interpretivism. b) Explain the terms “nomothetic” and “ideographic”. Should positivist research necessarily be nomothetic, and interpretivist research ideographic? Explain why (not). Nomothetic methods look for laws; idiographic methods give detailed descriptions of particular situations. These methods fit positivism and interpretivism, respectively, quite well, but not one-on-one. Positivists might use idiographic explanations as a substitute for laws (like in parts of biology and chemistry). Interpretationists, on the other hand, seem “stuck” with idiography. In their final chapter, Burrell and Morgan discuss the relation between the two approaches. They start by noting how easy it is to launch “destructive critiques” of each: “By assuming a posture in a rival paradigm, it would have been possible to demolish the contribution of any individual paper or theoretical perspective, by providing a comprehensive critique in terms of its underlying assumptions.” (p.395) It is, indeed, not difficult to criticize positivism from a interpretivist perspective, and vice versa. 12 c) Write a short interpretivist critique of positivist research, and a positivist critique of interpretivist research. Use a concrete example (e.g., try to criticize interpretive research into re-organizations, as in the paper by Isabella in the reader, from a positivist point of view). This is fun. There are many slanderous descriptions to be given. Some things that you might mention: Int on pos: overgeneralization, neglect crucial details and “human” factors. Pos on int: mere anecdotes, no validation of hypotheses, bankruptcy of social science. Yet Burrell and Morgan note that these criticisms only show how different both approaches are. They then argue for peaceful co-existence, at least for the time being: “Each paradigm needs to be developed in its own terms. What we are advocating amounts to a form of isolationism. (...) Contrary to the widely held belief that synthesis and mediation between paradigms is what is required, we argue that the real need is for paradigmatic closure (afsluiting). The paradigms reflect alternative realities. They stand as mutually exclusive ways of seeing the world.” (pp.397–398) d) Discuss whether this is a good way to resolve the positivism-interpretivism debate. Give at least one argument why it is good, and one why it is not – and then make a final judgment. What is nice about this “solution” is that it prevents counterproductive controversy and polemics. If there is some truth in both views, this truce (wapenstilstand) at least allows both parties to find it; otherwise they waste time trying to convince or slander each other. [What if, e.g., Lex Donaldson would have done positivist research instead of defending and promoting it against rival views?] What is, of course, awful about this solution is that it is cowardly. It assumes that there is no truth of the matter, and allows possibly mistaken views to continue. If positivism would be wrong, all research in it would be a waste of time, and finding the correct anti-positivist argument is, in a way, cost-efficient. JAN07 In the best-selling Freakonomics Steven Levitt describes an explanation for the dramatic decrease in crime rates in the USA in the 1990’s. first, he discusses several candidates explanations and finds them either incomplete or irrelevant. Then, he presents his own – controversial – alternative: crime rates dropped in the 1990’s because abortion had been legalized in 1973, with the Roe vs. Wade decision, made by the US Supreme Court. According to Levitt, this decision explains the decrease in crime rates in the following way: “What kind of woman would have benefited most from the Roe vs. Wade decision? She would be unmarried, under twenty years, or poor, and sometimes all of these. What kind of future would the child have had? One investigation shows that children that were not born in the first years after legalization of abortion would have had a 50% greater chance of living in poverty, and a 60% greater chance of being raised by a single parent. These two factors – poverty and single-parenthood – are among those that most strongly contribute to a criminal future for the child. M.A.J. Kusters (0557541) Being raised by a single parent almost doubles the chance that the child will later commit crimes; the same goes for children of teenage mothers. Another investigation has shown that having a relatively uneducated mother is the factor that contributes most strongly to criminality. In other words: exactly those factors that made millions of American woman decide to have an abortion appear to determine that their children, if they would have been born, would have had an unhappy and possible criminal life.” a) Describe briefly the distinction between positivist and interpretationist research. Of which type of research is Levitt’s explanation an example? Positivist research: looks for general statements about social reality, and for explanations of social phenomena, preferably stated in lawlike form. Their explanations refer to causal mechanisms. Interpretationist research: looks for understanding of specific phenomena in social reality; they interpret the actions of (groups of) people and try to find out/construct the meaning of these actions. These descriptions are specific to the situation. Levitt’s explanation: is positivist research: Levitt makes a general statement about a causal relation between two variables – abortion rates since the 1970’s and crime rates in the 1990’s. b) Describe the DN-model of explanation, as presented by Scherer in the reader. Does Levitt’s explanation fit the DNmodel? DN-model: every explanation has the form of a deductively valid argument, which has as its premises one or more universal (lawlike) statements and one or more statements about boundary and/or initial conditions. Levitt’s model does not, strictly speaking, fit the DN-model, because there is no logically necessary relation between abortion and crime rates: there are only probable relations. However, his explanation does fit the IS-model, which is a close variant of the DN-model (and also mentioned by Scherer). c) In her paper in the reader, Ann Chih Lin describes a way to combine positivist and interpretationist research. Briefly describe this combination strategy, and then apply it to Levitt’s research: if it is positivist, describe how it can be extended with interpretationist elements; if it is interpretationist, describe how it can be extended with positivist elements. Lin argues that it is possible to combine positivism and interpretationism in the following way: positivists discover general causal relations between variables, and interpretationists add to these relations an understanding of specific mechanisms. Thus, one general explanatory schema can be extended with several more detailed interpretations of specific instances. For the Levitt case, the general explanation of rising abortion rate and (later) dropping crime rate can be supplemented with a detailed understanding of what led pregnant women to their decision, and which aspects of this situation might have brought their future children to a life of crime – the text in the exam already indicates ways of providing such a detailed understanding. d) In another paper in the reader, Tsuyoshi Numagami argues that it is difficult, or even impossible, to find 0LM05 – Exam preparation invariant laws in management studies – because of “the reflective capacity of the agents involved”. Briefly desrive Numagami’s argument, and discuss whether it also affects Levitt’s explanation. Numagami’s argument against general laws in management studies is that agents will always have the capacity to respond to their knowledge of this “general” law – and that they might respond by violating it. Therefore, no law describing the behaviour of these knowing agents can be general. It is hard to see how this might undermine Levitt’s analysis, however. It seems that pregnant women who are in the situation that he describes (low income, low education, etc) have little choice but to have an abortion, or to see their children predisposed towards a life of crime. The only way to violate Levitt’s statistical law would be to have a group of these women consciously refusing to have an abortion and raise their children, against the odds, to be responsible citizens. JAN08 In one of the student papers that defended positivism against the interpretive phenomena, an argument was given that can be read as follows: "Interpretivists criticize positivist research, because it encounters various complicating phenomena – such as the hermeneutic circle and storytelling. However, these phenomena are also found in interpretivist research. Therefore, the presence of these phenomena does not count against the possibility of doing positivist research." a) Discuss the validity of this argument. INVALID. That these phenomena are also found in interpretivist research bears no relation to the (im)possibility of positivist research; it does not make the problems positivists (may) have with these phenomena any less that others also encounter them. In many of his works, including Personal Knowledge (1964), Michael Polanyi discusses the notion of 'tacit knowledge'. One characterization of this notion is the following: "The concept of 'tacit knowledge' refers to the fact that 'we can know more than we can tell'. For instance, we know people's faces and recognize them among a million, but we cannot make explicit how we do this. Any description of (what we know about) a face will also fit other faces. The same goes for a scientist who uses all kinds of instruments. He has mastered the skill to deal with these instruments, but as with riding a bike, it is impossible for him to make explicit all the knowledge he relies on in using the instruments. In short: people rely on tacit knowledge all the time. We rely on knowledge we CANNOT make explicit to acquire knowledge that IS explicit. So behind every bit of explicit knowledge, there is an unspecifiable amount of tacit knowledge." b) Discuss to what extent tacit knowledge presents an obstacle for positivist research. Start your answer with a short characterization of positivism. Positivism: is a “paradigm” or a way of thinking in the social science, that aims to model social science as in the natural science. Thus positivists look for general (invariant) laws to explain social phenomena. How people interpret 13 the world is irrelevant to positivists, as they are looking for underlying patterns. Tacit knowledge cannot be shared, because it cannot be made explicit. Therefore in research, where also tacit knowledge plays are role, not all factors involved are replicable and controllable. This implies that tests cannot be replicated in exactly the same way, under exactly the same conditions. Tacit knowledge thus presents an obstacle for positivist research, to the extend the scientist use certain research methods (like questionnaires), because these, like the scientist, rely on tacit knowledge. Furthermore, by considering tacit knowledge, also generalization of research results to social laws is problematic (because of uncontrollable and research that cannot be replicated). c) Discuss the similarities and differences between tacit knowledge and the phenomena of priming, sensemaking and situatedness. Start your answer by giving short characterizations of the three phenomena. Priming: literally refers to laying a ground layer of paint. In the present context, the ground layer is applied to someone’s mind, influencing this person's subsequent perceptions. It makes use of the selectiveness of our attention, guiding our attention to particular things, and influencing our evaluation of what we perceive. The person who is primed is normally not aware of this; thus, his attention is guided and his perception influenced by a ‘hidden’ factor. Versus tacit knowledge: priming can be caused by tacit knowledge (but priming is influence by others, and tacit knowledge is inherent and necessary thus not a distortion of perception)??? Sensemaking: refer to the interpretive process in which people are constantly engaged. We try to make sense of what has happened, is happening, and will happen, while acting at the same time. We often have to make sense of things of which we have only limited knowledge, or events that are unfinished. So we try to frame coherent interpretations that are in constant need of revision. These interpretations tend to involve rationalization, especially (though certainly not exclusively) when it comes to our own role in events. Versus tacit knowledge: in sensemaking one relies on tacit knowledge, making it explicit.. (but sensemaking is a process, while tacit knowledge is only an element in certain processes)??? Situatedness: refers to the characteristic of, e.g., research that it is always done in a particular context and against a particular background, and is necessarily influenced (or partly determined) by this. It implies that the results of one research cannot simply be transposed into another context. Versus tacit knowledge: it is impossible to make explicit all aspects of the situation in which research is conducted, just as it is impossible to make explicit all the tacit knowledge one relies on in acquiring explicit knowledge (situatedness is a characteristic of research and can be made explicit, while tacit knowledge is something possessed by people which by definition is impossible to make explicit)??? d) Is the hermeneutic circle equivalent to the theoryladenness of observation? If not, how are both phenomena related? Discuss, starting with a characterization of both the hermeneutic circle and theoryladenness. 14 Hermeneutic circle: In short: our interpretation of the whole depends on our interpretation of the parts and the other way around. Theory-ladenness: the term Kuhn used to denote the fact that all scientific observation depends on and is shaped by the theoretical background of the scientists, more specifically, on the paradigm within which they are working. Not equivalent, hermeneutic circle is broader than theories and observations. Furthermore theory ladenness is only about the meaning about the observations taken (only about the direction). Not about what the meaning is, and this does count for the hermeneutic circle. M.A.J. Kusters (0557541)