One Theoretical Issues in the Psychology of Aging

Timothy A. Salthouse

Much of the research literature in many scientific disciplines can be categorized as primarily addressing empirical, methodological, or theoretical issues. Empirical articles attempt to provide tentative answers to specific questions, methodological articles tend to focus on how to ask answerable questions, and a major goal of theoretical articles is to specify which questions are the most important to ask. The current chapter is primarily concerned with theoretical issues relevant to contemporary research in the psychology of aging, and thus it is focused more on questions than on answers. However, at the outset it is important to acknowledge that no attempt was made to be comprehensive in the coverage of theoretical issues, nor to link the material to the literature in the philosophy of science. Furthermore, the goal was not to review contemporary theories in the field of aging, but rather to stimulate consideration of a number of issues that appear to be central to theoretical speculations in the psychology of aging. Indeed, a major thesis of this chapter is that there is currently a great deal of confusion about what should be considered a psychological theory of aging, at least in part because there is little consensus with respect to the critical questions that should be addressed in order to qualify as a theory of aging. Because the author is most familiar with research on age differences in cognitive functioning, most of the examples are drawn from that research domain, but the issues are assumed to be applicable to many different domains.

At least two well-recognized advantages of theories are that they organize a great deal of information in a parsimonious manner and they serve to guide future research. One of the primary ways in which theories influence the direction of research is by identifying major issues that ultimately must be addressed by a satisfactory theory. By specifying which questions are interesting and important and, by omission, which are not, theories serve to focus research.

The focusing-of-research function is essential because a nearly infinite number of questions could be asked in most research areas, and the vast majority of them could be justified by the claim that their answers are not yet known. However, not all questions are equally important or informative, and a major role of theories is to specify which research questions are central and which are peripheral or irrelevant.

Table 1.1 Major Theoretical Questions in the Psychology of Aging

I. A Taxonomy of Major Questions

It is helpful to begin by considering a taxonomy of major questions that ultimately must be addressed by a successful theory. In addition to providing a systematic basis for evaluating theories, the taxonomy may also be useful in understanding how theories differ from one another and why it is often not feasible to make direct comparisons among theories.

The taxonomy, which is summarized in Table 1.1, conceptualizes major issues in the psychology of aging in terms of questions concerning the what, when, where, why, and how of the phenomenon of age-related differences in some aspect of behavior. Each of the questions is important because a phenomenon could be considered to be well understood, and the theory would be considered to provide a compelling explanation of the phenomenon, if all of the questions had convincing answers.

The question of what is clearly relevant to distinctions among theories because the answer will affect the intended scope of the theory. For example, if the theory is primarily concerned with age differences in a limited aspect of behavior, then it would probably be considered relatively narrow in scope. In contrast, if the phenomenon is defined in general terms that encompass many different types of variables, then it would be viewed as fairly broad.

How a theory answers the question of what will also determine which aspects of the phenomenon are considered primary, and perhaps somewhat analogous to the central “disease,” and which are considered secondary, and possibly more analogous to “symptoms” of the disease. For example, a variable that is interpreted as representing merely one of many possible manifestations of the phenomenon of primary interest might not be considered a high priority for theory-relevant research.

Because the answer to the question of what serves to specify the primary focus of a theory, this information needs to be considered before attempting to compare two or more theories. For example, if theories differ in their answers to the question of what, comparisons among them may not be meaningful because at a fundamental level they are not addressing the same phenomenon. To illustrate, two theories may both be concerned with adult age differences in measures of speed, and hence from a certain perspective they might be considered similar. However, one of the theories might be primarily concerned with describing the relations among reaction time measures of speed in adults of different ages (Cerella, 1990), whereas another might attempt to explain the relations between a theoretical construct of processing speed and adult age differences in a variety of measures of cognitive functioning (Salthouse, 1996). Because in the former case the answer to the what question refers to the relation among reaction times in different age groups whereas in the latter case it refers to the role of speed in age-related differences in cognitive performance, the theories are not addressing the same phenomena, and consequently it may not make sense to attempt to make direct comparisons among them.

The question of when is relevant to the evaluation of theories because if the theory assumes that the phenomenon begins early in adulthood, then the theorist needs to consider what can be learned by studying age differences very late in life, whereas if the theory assumes that the phenomenon begins late in life, then the relevance of observations in early adulthood needs to be considered. In other words, if a theory's answer to the question of when is very late in life, then research on young adults may not be directly relevant to the theory, but if the phenomenon is assumed to begin early in life, then research restricted to older adults may be of limited value. Whether findings from research on adults from different age ranges are relevant to the theory will therefore be determined by the theory's conceptualization of the phenomenon. Many studies compare a sample of young college students with a sample of adults between 60 and 80 years of age, others restrict their focus to adults above a certain age, such as 50 or 70, and a few studies compare adults across a wide range of 18 to 90 or older. However, depending on the theory's answer to the question of when, it may not be meaningful to treat each of these types of data as equally applicable to the theory because, for example, changes before age 60 may not be viewed as reflecting the same phenomenon or might not be postulated to involve the same mechanisms as changes that occur after age 80.

Most of the theoretical attempts to address the question of where have tried to localize age-related effects within some type of conceptual or neuroanatomical model of the behavioral variable of interest. To illustrate, at least five approaches to localization have been employed by researchers investigating influences associated with increased age on aspects of cognitive functioning. In each case a number of conceptually distinct “loci” for age-related influences have been postulated, and a primary goal of the research conducted within that perspective was to determine which of the possible loci have the greatest relations to age. For example, researchers working with correlation-based structural models have attempted to localize age-related influences within models of the organization of cognitive variables such as at the level of individual variables, the level of first-order factors, or the level of higher order factors. Researchers working with componential models have attempted to localize the influences within qualitatively different processing components that are postulated to contribute to performance on the task. Theorists who have proposed stage models attempt to localize age-related influences within components that are postulated to represent an ordered sequence between input and output, such as encoding, storage, and retrieval in the case of memory. Theorists working with computational models attempt to determine which specific parameters of one or more...