Theory & Psychology

Miller's (1969) influential paper argued that psychology must `give away' its scientific outputs to the people, thus signaling a process of democraticization based on a conventional understanding of the science and practice of psychology; a view that privileges the scientifically generated findings over other forms. In this paper, I argue for a democraticization of inputs, based on an ethnopsychological view of the science and practice of psychology and calling for a psychology that gives voice to the people as dialogic partners and co-authors of our work. The implications of this different understanding of democraticization is examined with respect to psychological theory, research, education and practice. The theme of experiential or epistemological democraticization that this paper reflects is seen to be part of a larger movement appearing in various segments of primarily Western societies and challenging the hegemony of currently prevailing frameworks for understanding the nature and character of the person.

In this paper I use a cultural studies framework in order to draw an analogy between methods and systems of representation. I then argue from Marxist and structuralist perspectives that method, like any coding practice, is ideological in that it defines the terms through which we experience and explain the world. Further, I identify ideological analysis in psychology as dealing mostly with ideology as a set of meanings, not as a set of practices, and treating method as derived from theory. I propose the reverse argument that regards method as creating and not as reflecting the phenomena, and argue that all methods are ideological. Finally, I illustrate the argument by an example which shows how a structured scale and an unstructured questionnaire construct different models of self-presentational behaviour, and different notions of reality.

The theoretical problem of explaining how people can produce and understand `sentences' they have never heard before has, for some time, been at the center of psycholinguistic and cognitive theorizing. It has rarely been questioned as a coherently formulated `problem' in itself. In recent years, however, a range of conceptual criticisms has been advanced from within the Wittgensteinian tradition. A consideration of these criticisms in relation to Chomsky's various formulations constitutes the main purpose of this paper. Among the critical issues discussed are: (1) the compositional theory of linguistic behavior which was developed as part of the solution to the new sentence (projection) problem; (2) the conception of linguistic comprehension which underlies it; (3) the `ideas/utterance' translation assumption; (4) the characterization of language learning embraced by the treatment of the projection problem; (5) the internal rule-system model; and (6) alternative versions of the phenomena grouped under the `new sentence problem'.

The argument that empirical studies should be appraised in terms of multiple context-dependent criteria is reviewed. A distinction is made between `practical impact' appraisal and `conceptual rigor' assessment. The idea of replacing a test of statistical significance with multiple criteria is assessed and found inappropriate when we are concerned with conceptual rigor. A test of statistical significance can be profitably used to ascertain whether or not a theory is tenable vis-a-vis a set of experimental data. It is an indispensable step in the multi-step theory-corroboration procedure which is hypothetico-deductive in nature. The binary nature of a significance test is compatible with the evolving nature of scientific knowledge, which should be based on internally and externally valid research results. The use of multiple context-dependent criteria of impressiveness, on the other hand, is antithetical to the requirements of internal validity and of external validity. A case is made that criticisms of the use of significance tests and of the hypothetico-deductive approach do not detract from the contribution to conceptual rigor due to the use of a test of significance and of the hypothetico-deductive nature of theory corroboration experimentation.

In his paper entitled `Conceptual rigor versus practical impact', Chow (1991) argued against Rosnow and Rosenthal's (1989) recent call for caution and sophistication in the evaluation of the results of empirical studies. Chow also holds forth upon the purported failings of meta-analysis as a form of integrating research findings that is antithetical to the purpose of reviewing research. This commentary considers the paucity of evidence in support of Chow's assertions and illuminates the syllogistic fallacies used to disguise the deficiencies rife in Chow's paper. This commentary also provides correct information regarding fundamental issues in meta-analysis, the relationships between effect size, sample size and significance level, and the putative distinction between the concern for practical impact and Chow's notion of `conceptual rigor'.

Chow (1991) has defined the experimental logic and methodology described by Rosnow and Rosenthal (1989) as being incompatible with his notion of `conceptual rigor' and even antithetical to the requirements of internal and external validity. The purpose of this response is to suggest: (a) that the two approaches in question are not, in their entirety, mutually exclusive and differ only at one precise step within the theory-corroboration process; and (b) that to rely exclusively on probability of outcome as the sole decision criteria in theory corroboration without being cognizant of the theoretically irrelevant factors (e.g. sample size of study) which affect them seems, to this author, antithetical to science.

Chow (1991) distinguishes between `practical impact' and `conceptual rigor' research, and he concludes that effect-size estimation is useful only in practical impact research. I argue that significance tests do not answer substantive questions about the data and are useful only as a check that the results are unlikely to have occurred by chance. Chow's decision to regard the similarity between data and prediction as being a dichotomous judgment made on the basis of significance testing is therefore unwise. I conclude that effect sizes are the single best index of the relationship between theoretical predictions and the obtained data. The role of replications and meta-analysis in advancing theory is also discussed.

Conceptual rigor is indeed a desideratum worth dedicated pursuit; in fact, one might wish that Chow had pursued it somewhat more diligently in his present essay. I suggest that the approach to data interpretation he advocates here is an etch-a-sketch draft whose prospect for refinement into an operational logic of inference that professional scientists can live by appears minuscule.

This is Chow's response to four comments on his critique of the view that research conclusions be based on multiple context-dependent criteria. Five themes could be identified in the comments. In reply, it is argued that care should be taken not to use the alpha level whimsically because the continuum similarity is being used as a dichotomy in theory corroboration. The superiority of effect-size estimates to statistical significance is more apparent than real. Chow's assessment of meta analysis is illustrated with the apples and oranges issue.