Aging and Skilled Performance

Advances in Theory and Applications

Wendy A. Rogers, Arthur D. Fisk, Neff Walker

• Publisher: Psychology Press
• Year: 2014
• Print ISBN: 9780805819106
• eBook ISBN: 9781317779445
• Topic: Psychology
• Subtopic: Developmental Psychology
• Index: Psychology

The term "skill" encompasses an array of topics and issues. For example, individuals are skilled in a variety of domains such as chess, typing, air traffic control, or knitting; researchers study skill in a variety of ways, including speed of acquisition, accuracy of performance, and retention over time; and there are a variety of approaches to the study of skill such as computer modeling or experimental analysis. Contributing to the understanding of whether, how, when, and why skills may decline as a function of age is the goal of this volume.

This book is based on the Aging and Skill Conference sponsored by the Center for Applied Cognitive Research on Aging. The broad focus of the conference was to discuss cognitive theories underlying age-related skill acquisition, transfer, and retention and to discuss applications of these theories to such issues as age-adaptive training, compensatory strategies and devices, and utilization of new and existing technology. The contributors were asked to discuss the cognitive theory relevant to their topic, explain how the theory informs the field about aging, examine where gaps exist among general cognitive theory in this area and theories of aging, and demonstrate the practical relevance of the theory to enhancing or enabling activities of daily living--for work, home, or leisure--for older adults.

This is the first book to focus exclusively on aging and skill. It covers a range of abilities, provides the theoretical basis for the current status of age-related differences in skill, and offers direct evidence of the applicability of research on proficiency to aspects of daily living. Each chapter was written either by an expert in the field of aging, or by an expert in the field of skill--many expert in both areas.

CHAPTER 1

Practical Relevance and Age-Related Research: Can Theory Advance Without Application?

Arthur D. Fisk
Alex Kirlik

Georgia institute of Technology

Progress in a scientific field can be measured in a variety of ways. Certainly, theoretical maturity and advancement is one such measure. The ability of a scientific field to motivate and inform solutions of practical importance seems also to be a key measure. As researchers, the authors are often called upon by our students or colleagues to explain the relevance of our work to one theory or another, to defend the theoretical positions we hold, or to simply explain the theoretical advances we hope to make as we embark on a new set of experiments. Generally, such requests provide us little reason for pause. However, when we leave our academic offices we sometimes encounter less than rewarding experiences. More often than not, we feel awkward when discussing our field, say, with acquaintances at parties or during visits from our relatives. We can discuss how our research has been directly applied to the design of training systems to support skill refinement of American football quarterbacks (“Inside the South,” 1994); yet, when discussing cognition and aging, the examples are not as forthcoming.

Sometimes we are asked how our work in aging can be used to solve a specific problem such as designing a plan to make use of public transportation easier (from a navigation perspective), designing an effective interface and training program to make use of an on-line card catalog system easy and effective, or documenting the best way to ensure that older adults can use various types of new technology currently existing only in the minds of some design team. Such encounters sometimes constrain our otherwise high enthusiasm about the true progress of our theoretical advances.

Psychological research, especially within the field of skill acquisition, should be able to answer the questions mentioned earlier. Our goal for this chapter is to explore the need for an expansion of both domains of study and research techniques to afford the opportunity for more ease in answering questions of a practical nature. We also have set out to discuss details of a set of criteria that would allow such expansion to be productive as well as to discuss the process to achieve such a goal. We highlight evidence in the literature from psychology and other fields that suggests that such a research approach will not only improve the applicability of research findings but also will result in theoretical advances that cannot be gained otherwise. To summarize, our plan for this chapter is to: (a) discuss our definition of research that has practical relevance, (b) outline the characteristics of practically relevant research, (c) present and attempt to dispel some common misconceptions about this class of research, and (d) discuss why such work is important.

WHAT WE DO NOT DEBATE

As a beginning point we wish to emphasize what this chapter is not concerned with: This is not a debate about the merits of so called “basic” versus “applied” research. Such an ill-founded debate has been ongoing for far too many years. In such a confrontation one often hears from one side that too few results from the laboratory can be applied because “academic” researchers study meaningless problems and on the other side one can hear that “applied” research has no generality and is conducted by individuals who are not talented enough to do “real science.” These are simply inflammatory comments that miss the important issues that should be discussed. The discussion should concern how we solve important scientific and practical problems and why we should decide upon and make use of research techniques that can be most appropriate to address a specified research issue.

DEFINITION OF RESEARCH THAT HAS PRACTICAL RELEVANCE

The concept of practically relevant research is certainly not new. Indeed, the discussion of the need for research with practical relevance has survived the lifetimes of many psychologists. To have practical relevance, research should derive from real, practical problems with the intent to solve those problems, but the work also should be designed to incorporate, build on, and advance theory. It should be clear that every single study conducted will not necessarily solve a real-world problem, just as a single study may not necessarily resolve a complex theoretical issue. Hence, practical relevance must be thought of more broadly in regard to a research program.

Many examples of practically relevant research programs come to mind, for example, Broadbents early work on attention was driven by the need to solve practical problems encountered by air traffic controllers. Communication theory originated from efforts to deal with the problem of maximizing information transmission over the telephone line. Speech research at the Haskins laboratories began with the problem of building a reading machine for the blind. The ecological approach to perception emerged, in part, from the issue of landing an aircraft.

Adams (1972) pointed out that in the modal case many years pass before the results of research on human abilities have a meaningful impact on technology or in dealing with real-world problems. Yet, attention to real, practical problems when undertaking research can provide valuable tools for delimiting a problem and selecting the most fruitful course for the investigation. In addition, such efforts can speed the recognition of the value of a body of research from a given area and more quickly allow useful application.

CHARACTERISTICS OF PRACTICALLY RELEVANT RESEARCH

This type of research transcends many domains and approaches used to address psychological research issues. Fisher (1993), when discussing the concept of “Optimal Performance Engineering,” cogently argued that the well-known trio of empirical, theoretical, and analytical studies are candidates for what we have refered to as practically relevant research. It is not the class of research technique but rather the characteristics of the approach that determine the practical relevance of the work. One useful approach, and an important characteristic, is to design the tasks to be studied in the spirit of Brunswik’s concept of representative design (see Hammond, 1966).

Representative design encourages the researcher to give the same care and attention to the issue of sampling task environments as is currently given, say, to the issue of sampling subject populations. The generalizability of research is constrained by the validity of induction from a subject sample to the relevant subject population, as well as by the validity of induction from a sampled task to the population of relevant tasks. Currently, the dimensions of subject variability are perhaps better understood than the dimensions of task variability, and, thus, subject sampling is a more mature enterprise than task sampling. To improve this situation, the theory and logic underlying the selection of an experimental task must be made explicit so that the class of tasks to which findings should generalize can be identified. Thus, the validity of the task sampling theory can be treated empirically and therefore subject to refinement or falsification.

Such an approach requires the use and thorough understanding of task analysis. Task analysis is a concept of complete familiarity to the human factors engineer (Drury, Paramore, Van Cott, Grey, & Corlett, 1987), yet it is a remarkably foreign concept in most psychology laboratories. Use of a task analytic method is important when designing research because it forces one to attend closely to the important factors of the task, environment, subject and subject population, and choice of dependent variables.

The Task

How can such a focus help the research enterprise? Consider first an analysis of the task. There are numerous definitions of what constitutes a task. However, there can at least be consensus that a task is a set of activities—for our purposes, mental activities that contribute to a specific functional objective. Drury et al. (1987) outlined the characteristics of a task: Actions are grouped together by their objective and their temporal nature. These actions include perceptions, discriminations, decisions, control actions, and communications. Every task involves some combination of cognitive and physical action. Each task has a stimulus or cue that identifies its starting point. A task also has a stopping point that is identified by some related information or feedback. Tasks usually are defined as units of action performed by a single individual (but see, e.g., issues in team training, Swezey & Salas, 1992).

To perform a proper task analysis one must be able to describe the task, understand the task requirements, and analyze and interpret the task requirements in terms of knowledge and theory about human characteristics. For our purposes, major attention should be directed at understanding cognitive processes and their interaction with the targeted task. Assume we were interested in the issue of age-related time-sharing during multitask performance. One strategy might be to design a dual-task test where two tasks must be performed simultaneously. It is fair to say that most laboratory experiments examining multi-tasking performance allow (or require) the two tasks to be integrated to the point that successful performance could depend more on the effectiveness of task integration than on the effectiveness of true task time-sharing (e.g., Korteling, 1993). Such research may well inform us about aging and task integration; however, an analysis of activities encountered in daily living suggests that to achieve the overarching goals required for successful performance of most complex tasks there is a requirement for coordination of cognitive processing associated with the task activities. Such analysis would make salient the idea that independent actions must be planned and coordinated at different times, different amounts of resources must be applied to the components at different times, and some components must be put on hold while others are emphasized at a given point in time. Such observations have been the basis of formal models that optimally describe time-sharing performance (Schneider & Detweiler, 1988) and lead to experimental design drastically different than the traditional dual-task experiment.

Other examples of the usefulness of this type of approach can be found in various areas. For example, the skills-analysis based training movement in the United Kingdom made substantive use of task analytic methodologies to translate issues related to perceptual and motor difficulties of industrial skills learning to research questions (e.g., Crossman, 1956; Seymour, 1968). Schneider, Vidulich, and Yeh (1982), Eggemeier and Fisk (1992), and more recently Kirlik, Walker, and Fisk (in press), made use of extensive task analysis to translate real-world training problems involving complex perceptual rule-based learning into complex, but well controlled laboratory research questions.

Gibson (1965) stated well the need for task analysis when she emphasized the need to understand the important elements of stimuli for research on learning to read:

Some alphabetic writing systems have nearly perfect single-letter-to-sound correspondence, but some like English, have far more complex correspondence between spelling patterns and speech patterns. Whatever the nature of the correspondence, it is vital to a proper analysis of the reading task that they be understood…. It would be useful to know just what the distinctive feature of letters are. What dimensions of difference must a child learn to detect in order to perceive each letter as unique? (pp. 1067–1068)

From the analysis she performed, Gibson was able to create relevant stimuli and experimental tasks to bring the research questions under conditions of control while maintaining close connection with the real-world problem of interest.

The Environment

Now consider the environment. Numerous authors noted the failure of psychology to meaningfully contribute to understanding how to solve practical problems to some degree because of a lack of appreciation of the context within which cognition and behavior occurs (but see Chapanis, 1988; Fisher, 1993; Kirlik, in press, for a review and a rebuttal). If these suggestions, that the products of a basic science do not provide effective resources for application, are even only partially true then such realization can provide important lessons for basic science itself. Such “lessons learned” point to the need to understand the context in which cognitive activity occurs.

The endeavor of understanding the environment in which cognitive activity will take place may be as important as understanding the overarching task in which the cognitive activity of interest is embedded. Thus, one must focus on determining under what environmental conditions various cognitive activities will be activated, and required, for effective task performance. The issue is not only to understand cognitive processes such as problem solving, working memory, or skill acquisition but also to determine in what context problems must be solved, when decisions must be made and what factors affect their outcome, what task characteristics lead to working memory demands, or what skills must be acquired given various environmental constraints.

An example provided by Kirlik (in press) may better make the point. Kirlik pointed out that one of the earliest attempts to model human-machine interaction concerned manual control behavior, such as steering a car or flying an aircraft. Engineers familiar with the design of electromechanical feedback control systems turned their attention to modeling the human as a feedback control system in order to assess human capabilities and limits so that vehicles could be designed so that control demands were within these limits. Control theory has a well-specified language for environmental modeling. The thing being controlled can be described in terms of a transfer function that relates system inputs (steering adjustments) to system outputs (headings). The goal of this endeavor was to discover the human transfer function, that is, a description of the function relating stimuli to responses during manual control behavior.

As Kirlik (in press) stated, the engineers were in for a rude awakening. The empirical results indicated that there was no single human transfer function. Rather, the human transfer function appeared to adjust to changes in the dynamics of the controlled system. Birmingham and Taylor (1954) noted that the ability of the human to adjust to the environmental transfer function was so great that the control system designer was doomed to failure. Subsequent modeling attempts (McRuer & Jex, 1967) were successful only once the search for invariance in behavior shifted to the level of the human-machine system rather than in human behavior alone.

Given this finding concerning human perceptual-motor behavior, why should we expect cognitive-level behavior to be any less adaptive to environmental structure than is perceptual-motor behavior? Much modern psychology research, especially in cognitive aging, paints a rather dismal picture of human cognitive abilities and limitations, leaving us to wonder how it can ever be possible for human cognition to allow our subjects even to find their way into the laboratory. But rarely does the cognitive system operate in isolation from external environmental aids.

How often have we discovered the cleverness of participants of our experiments who are able to confound our studies of complex cognitive activity by cueing off what we might think of as extraneous aberrations? Often our response is to bring tighter and tighter con...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table Contents
• Preface
• 1 Practical Relevance and Age-Related Research: Can Theory Advance Without Application?
• 2 State Models of Paired Associate Learning: The General Acquisition, Decrement, and Training Hypotheses
• 3 The Use of Signal Detection Theory in Research on Age-Related Differences in Movement Control
• 4 Control Theoretic Approaches to Age-Related Differences in Skilled Performance
• 5 Aging and Dual-Task Performance
• 6 Aging and Memory: Implications for Skilled Performance
• 7 Intelligence as Process and Knowledge: An Integration for Adult Development and Application
• 8 The Effects of Display Layout on Keeping Track of Visual-Spatial Information
• 9 Assessing Age-Related Differences in the Long-Term Retention of Skills
• 10 Aging and the Acquisition of Computer Skills
• 11 Cognitive Theory and Word Processing Training: When Prediction Fails
• 12 Instructional Design for Older Computer Users: The Influence of Cognitive Factors
• Author Index
• Subject Index