Investigating the Psychological World: Scientific Method in the Behavioral Sciences

Notes

Notes 1 Method, Methodology, and Realism 1. In a sympathetic appraisal of Laudan’s account of the transition from inductive to hypothetico-deductive method, Ernan McMullin (1984) took issue with some of its detail and emphasis. McMullin agreed with Laudan’s central contention that since 1700 the philosophy of science had to face the fact that science increasingly appealed to theoretical entities. However, he maintained that the acceptance of the hypothetico-deductive method in the seventeenth century was prompted more by the “corpuscular philosophy” of thinkers such as Robert Boyle and John Locke than the successful use of the hypothetico-deductive method in science. 2. Bert Uchino, Dustin Thoman, and Sari Byerly (2010) sampled over 230 articles from the prominent Journal of Personality and Social Psychology from 1982 through 2005 and found that the large majority of articles favored a testing strategy of confirmation. Considerably fewer favored a strategy of falsification, and even fewer favored a strategy of employing crucial tests of multiple hypotheses or theories. These findings square with the author’s casual impressions and speak against the claim sometimes made that Popperian falsification is psychology ’s hypothetico-deductive method of choice. 3. Of course, there are other prominent accounts of scientific method. Two of the best known are T. C. Chamberlin’s (1965) method of multiple working hypotheses and John Platt’s (1964) advocacy of strong inference. Although they promote important ideas (theoretical pluralism and strong tests, respectively) and receive regular endorsement by methodologists, they seem to have had a limited influence on scientific practice. O’Donohue and Buchanan (2001) provide a thoughtful critique, written for psychologists, of Platt’s theory of strong inference. 4. Strictly speaking, the claim that there cannot be a logic for discovering hypotheses is a corollary to the hypothetico-deductive method, not a part of it. Some descriptions of the method speak about the amethodological formulation of 164 Notes hypotheses to explain the data. However, because hypothesis generation is not part of the method proper, I do not include it in my description and discussion of the method. 5. Erotetic logic, the logic of questions, is the obvious source for a theory of questions, but in my view it is too formal to be readily applicable to most of our scientific problems or to help researchers directly. However, this is not to deny that models of interrogative inquiry may give us some useful insights about inquiry processes generally. 6. More recently, Wimsatt (2007) extended his list of the important properties of heuristics. In addition to the four just mentioned, he noted that heuristics are purpose relative (they are useful for something) and are also derived with modification from other heuristics to better perform a new role. 7. Preliminary results from a 2009 PhilPapers survey of over three thousand philosophers showed that 66 percent either accepted or leaned toward scientific realism, whereas 18 percent favored scientific antirealism. This is in keeping with results from the same survey on a number of more specific philosophical categories , which favored a naturalistic metaphilosophy, a nonskeptical realism about the external world, a correspondence view of truth, and a non-Humean conception of laws. 8. The term causal mechanism is ambiguous. In ATOM, the generation of theories involves explanatory inference to claims about the existence of causal entities . It is not until the subsequent development of these theories that the mechanisms responsible for the production of their effects are identified and spelled out. Also, ATOM assumes that the productivity of causal mechanisms is distinct from the regularities that they explain (Bogen, 2005; cf. Woodward, 2003). Importantly, this allows for the methodological use of generalizations that describe natural regularities to help identify the causal mechanisms that produce them. 9. Note, however, that the strategy of analogical modeling is essential for theory development in ATOM, and the theory of explanatory coherence does heavy-duty work in the theory because it is the best-developed method of inference to the best explanation currently available. 2 Detecting Psychological Phenomena 1. When contrasting explanatory theories with claims about phenomena, Bogen and Woodward focus on what they call systematic theories. For them, systematic theories properly explain phenomena by showing in detail how the phenomena result from the causal factors appealed to in their explanation, and by unifying, and therefore systematizing, the phenomena claims. Psychology seems to have few well-developed theories of this sort.Although it constructs theories of various kinds, most of them are modest theories with low, but genuine, explanatory power. 2. Bogen...