Life-Span Developmental Psychology

Methodological Contributions

Stanley H. Cohen, Hayne W. Reese, Stanley H. Cohen, Hayne W. Reese

• Publisher: Psychology Press
• Year: 2019
• Print ISBN: 9780898599718
• eBook ISBN: 9781317739333
• Topic: Psychology
• Subtopic: Developmental Psychology
• Index: Psychology

Dealing with the methodological and data analytic problems in developmental research, this book presents solutions advanced from the disciplinary perspectives of psychology, behavior analysis and behavioral systems, sociology, and anthropology. Topics addressed include:
* the metatheoretical issues about the relationship between data and theory
* the identification and analysis of age, cohort, and time-of-measurement effects
* the assessment of quantitative and qualitative change
* the use of group and single-subject designs for control by systematic variation
* the use of systems methodology to investigate the developmental continuity and organization of behavior
* the analysis of data from repeated measures designs
* the use of structural equations and path analysis to test causal hypotheses
* the use of structured relational matrices to study development and change

This unique volume offers students an unusually wide range of research tools for identifying and studying specific developmental problems.

1 The Data/Theory Dialectic: The Nature of Scientific Progress

Hayne W. Reese

West Virginia University

Kant (1781/1896) said, “Thoughts without contents are empty: intuitions without concepts are blind” (p. 41). Thoughts, for Kant, can be called ideas, conceptions, concepts, or the like, and intuitions are sense impressions. Thus, Kant’s “celebrated dictum” (Cassirer, 1954, p. 56) means that concepts are empty of empirical content unless they refer ultimately to sense impressions; and sense impressions are literally meaningless unless they refer to concepts, that is, in information-processing terms, unless they are encoded conceptually. This proposition is widely generalized. For example, in a manual on Marxism-Leninism published in the former Soviet Union: “Divorced from practice, theory is barren. Unguided by theory, practice is doomed to grope in the dark” (Fundamentals, 1961, p. 114). In other words, theory without data is groundless, and data without theory are uninterpretable. This proposition is tautological because its truth is given by the definitions of the terms it contains; but it provides a model for analyzing the nature of scientific progress. The purpose of this chapter is to examine this model.

THE RELATION BETWEEN DATA AND THEORY

Discussion of the relation between data and theory might begin with the question of which came first, the data or the theory? Before this question can be answered, both “data” and “theory” need to be defined.

Definitions

Data constitute a body of information that principled criticism indicates is true (the kind of criticism needed is discussed in the section The Methodological and Theoretical Domains). “Data” in this sense includes both “data” and “danda” as defined by Pepper (1942, chap. 3). A theory is a set of definitions of concepts and hypothesized principles and laws relating these concepts to one another.

Which Is Primary?

Agassi (1964, p. 193) said that when one of Hegel’s doctrines turned out not to agree with the facts, Hegel said “so much the worse for the facts.” Agassi commented, “Rarely has anyone paid more dearly for a silly joke.” I do not know the context of the quotation from Hegel (Agassi gave no reference), but taking the remark as serious instead of a silly joke would be consistent with Hegel’s philosophy. This conception of theory as primary over data is idealistic, and it is contrary to the popular conception of data as primary. Relative to data, theory has often been denigrated not only by laypersons — “I’ll believe it when I see it” —but also by scientists (Anastasi, 1992; Overton, 1991). For example, Karl Dallenbach (1953) said, “There is such a thing as theory-blindness, i.e., being so blinded by adherence to a specific theory as to be unable truly to see the facts” (p. 34). Another example is a statement by Cofer (1968) referring to an exchange at a colloquium in which he briefly described an informal model of controlled association. A colleague remarked that the model sounded more like information processing than stimulus-response learning, and Cofer said:

I could not care less. We obtain data on problems of interest and we attempt to make sense of the data, trying to remain moderately rigorous in its formulation in the sense that its terms are open to test by further observation and experiment, (p. 534)

Skinner (1981) was more emphatic:

[Models] evoke contemplation rather than action. The theoretical physicist wants to represent reality; the laboratory physicist wants to do something about it. One changes a model to produce a different picture; the other manipulates independent variables to change a dependent variable. A model is what something is to be done about; it is not what is to be done. Model is little more than another word for idea — something known by acquaintance. I look forward to greater recognition of the importance of laboratory scientists. The theorists have been sponging on them for decades and getting most of the credit, (pp. 173–174)

One problem with such a position is that data by themselves are meaningless: Without an integrating theory, they are completely dispersed. In discussing this passage from Skinner, Parrott (1986) commented:

To characterize the activities of theoretical scientists, or system-builders, as “sponging” on their experimental colleagues certainly does little to promote theoretical work. Moreover, because this passage appears without a context, whereby the types of theory typically and appropriately criticized by Skinner in this manner may be identified, the passage is damaging to theory building efforts of any sort. (p. 39)

Scientific facts and theories are inseparable (e.g., Kessel, 1969; Kuhn, 1970, p. 7). As Dallenbach (1953) said, a study without a theory is not a scientific experiment, it is only busy work; without a theory, such work has no meaning and does not contribute to scientific knowledge. Lakatos (1978) also considered empirical fact-finding to have little importance unless it is theory-guided. In these views, the legal maxim “The fact speaks for itself” (Res ipsa loquitur) has no basis in the philosophy of science. A rationale for this position is provided by Rychlak’s (1976a) distinction between the methodological and theoretical domains of science.

THE METHODOLOGICAL AND THEORETICAL DOMAINS

Rychlak (1976a) distinguished between methodological and theoretical domains of science:

A theory is simply a stipulated (hypothesized, suggested, believed-in, etc.) relationship between two or more constructs (items, terms, observed events, etc.). A method is the means or manner of testing the theoretical proposition we have formulated. Knowledge is the understanding of why a given theoretical proposition (construct, statement, etc.) is true or false after this proposition has been put to test. (p. 221; reference citation deleted)

The methodological domain refers to methods of investigation, or ways of obtaining information. The theoretical domain refers to interpretation. For example, the methodological domain refers to information about antecedent–consequent relations, or more generally, relations between independent variables and dependent variables; and the theoretical domain refers to stimulus–response relations, or more generally, cause–effect relations. The relevant point here is that causality does not refer to observables, but rather refers to interpretations (for further discussion, see Reese, 1986, 1993). Observables are in the methodological domain; interpretations are in the theoretical domain.

The methodological domain is used to obtain data, but data are not scientific knowledge until they have been interpreted within the theoretical domain. The methodological and theoretical domains are interrelated (Overton & Reese, 1973). On the one hand, the selection of independent and dependent variables, which occurs in the methodological domain, is motivated by at least a low-level theory regarding what the independent and dependent variables “mean”—that is, what their construct validity is (Cook & Campbell, 1979, pp. 59–64) — and why they might be related. But on the other hand, ideas about what the independent and dependent variables mean and why they should be related are influenced by observed relations between these variables (e.g., Mulaik, 1985).

THE CONTEXTS OF DISCOVERY AND JUSTIFICATION

Rychlak’s distinction between the methodological and theoretical domains is related to a distinction by Reichenbach between the contexts of discovery and justification in science (1938, pp. 6–7; 1951, p. 231). Reichenbach used the words discovery and justification in their everyday senses, not in technical senses that required deep philosophical analysis. The distinction is between the psychology of scientists and the epistemology of science (Reichenbach, 1938, p. 5): (a) The context of discovery is the context in which discoveries are made. It includes the motivations, insights, values, and other personal characteristics that underlie the way a scientist selects a problem, designs a study, and analyzes, interprets, and reports the results. (b) The context of justification is the context in which discoveries are fitted into the body of scientific knowledge. To the extent that these contexts have the same rational basis, that is, the same underlying world view, they should be related; but the context of discovery seems often to be related to a different world view from that of the context of justification.

Arguments about the Distinction

The commonsense definitions — or, better, characterizations — of the contexts of discovery and of justification have been overlooked by some writers and the issues have become almost mystical. Overton (1984, 1991) assigned the distinction to a “conventionalistic” view of science, which is inferior, in Overton’s opinion, to the organic view he espoused. Conventionalism is a philosophy of mechanistic science (Overton, 1984, 1991); but of course it is not therefore inferior to an organic philosophy. And anyway, Reichenbach explicitly rejected conventionalism (1938, pp. 14–16).

Overton was incorrect if the contexts are understood in Reichenbach’s senses; but the phrase “context of discovery” has come to include some of what Reichenbach included in the context of justification, and with this changed meaning, Overton was correct. The context of discovery is unquestionably an important part of the history of science, but in Reichenbach’s sense it is also absolutely excluded from current science as such. Otherwise, psychology would be an integral part of current knowledge in any science and the new philosophies would be proved false by their own categories. The psychology of discovery includes motivations, which are culturally instilled; insights, which perhaps require certain minimal biogenetic endowments; and values, which are again culturally instilled (on the last point, see Reese & Fremouw, 1984).

Reichenbach made the distinction in order to delimit the tasks of a philosophy of science, which he asserted deals only with the context of justification. In agreement, Brodbeck (1953) asserted that the philosophy of science is a logical analysis of concept formation, theory construction, the nature of laws, and verification. It does not include (1) the “science of science” (p. 3), that is, the history of ideas or the sociology or psychology of science — the scientific study of scientists’ motives, the social impact of science, and scientists’ activities; (2) the ethics and morality of science; and (3) the philosophy of nature, metaphysics, and epistemology.

This position has been challenged on the argument that a philosophy of science must cover both contexts. However, the arguments have been specious. Feyerabend (1978) argued that the domains of discovery and justification “are equally important to science and… must be given equal weight. Hence… we are dealing with a single uniform domain of procedures all of which are equally important for the growth of science. This disposes of the distinction” (p. 167). However, Feyerabend took a great leap from “equal weight” to “uniform domain of procedures.” The procedures differ, as dialectical opposites, and the fact that they are united in a broad domain of procedures does not make this domain uniform and therefore does not dispose of the distinction.

As defined by Reichenbach, epistemology is a “rational reconstruction” of the context of justification (1938, p. 5). For Lakatos (1978) and other proponents of the view that the context of discovery must also be covered, the context of discovery also requires rational reconstruction. In other words, the wider view of the philosophy of science — redefined as the history of science — requires the philosopher to be also a psychologist, or to pretend to be, but from the armchair of rational reconstruction rather than the field hand’s work at empirical demonstration. Empirical demonstration is lacking, incidentally, not only with respect to the validity of these philosophers’ interpretation of the context of discovery but also with respect to the historicity of their reconstructions of the history of science.

Rational Reconstruction

Reichenbach attributed the phrase “rational reconstruction” to Carnap, citing the first edition of Carnap’s Der logische Aufban der Welt, published in 1928 (Reichenbach, 1938, footnote 1, p. 5). In the second edition of the work Reichenbach cited, Carnap (1967) said:

[A rational reconstruction] does not represent the actual process of cognition in its concrete manifestations, but… is intended to [represent] the formal structure of this process. This viewpoint allows an...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Preface
• Acknowledgments
• List of Contributors
• 1. The Data/Theory Dialectic: The Nature of Scientific Progress
• 2. Reflections on a “Model” Approach to Metapsychology
• 3. Developmental Designs Revisited
• 4. Survey Methodology in Life-Span Human Development Research
• 5. Single-Subject Designs and Developmental Psychology
• 6. Notes from the Field: On the Coordinated Use of Quantitative and Qualitative Data
• 7. Behavioral Systems Methodology: Investigating Continuity and Organization in Developmental Interactions
• 8. Computer-Assisted Qualitative Data Management: Some Recent Innovations in the Management and Analysis of Field Notes
• 9. Repeated Measures Analysis in Developmental Research: What Our ANOVA Text Didn’t Tell Us
• 10. Using Multivariate Data to Structure Developmental Change
• Author Index
• Subject Index