Commentary Human Development 2007;50:359–366 DOI: 10.1159/000109836
Understanding the Aging Decision Maker Rui Mata University of Michigan, Ann Arbor, Mich., USA
Key Words Adaptive decision making Aging Ecological rationality
From his early 60s to his late 70s, Alan Greenspan was chairman of the Federal Reserve, the central bank of the United States of America, and thus one of the most important decision makers in American economic policy. Although not without critics, many economists agree that during his almost 20 years in office Greenspan managed the Federal Reserve effectively and was able to deal successfully with various stock market crises. Moreover, Greenspan remained a sought-after expert after stepping down from his position as chairman, being appointed to influential positions in major financial institutions. Alan Greenspan is a prototypical case of successful decision making in older age and represents a particularly encouraging example of successful aging at a time where more and more people are asked to make important decisions well into old age. Living Longer and Better: The Importance of Understanding the Aging Decision Maker
More people are living longer than ever before: human longevity increased steadily at 3 months per year in the last century [Oeppen & Vaupel, 2002]. Increases in life expectancy have been accompanied by falling or stagnating birth rates in many countries, leading to larger numbers of retirees compared to workers. This imbalance has put pressure on welfare systems in many countries that have initiated reforms to allow for longer working lives and longer postretirement autonomous living [National Institute on Aging, 2007; The Swedish National Institute of Public
© 2007 S. Karger AG, Basel 0018–716X/07/0506–0359$23.50/0 Fax +41 61 306 12 34 E-Mail
[email protected] Accessible online at: www.karger.com www.karger.com/hde
Rui Mata Department of Psychology, University of Michigan East Hall, 530 Church St. Ann Arbor, MI 48109 (USA) Tel. +1 734 763 7344, E-Mail
[email protected]
60
Number of papers
50 40 30 20 10 0
1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Year
Fig. 1. Published papers on decision making and aging in the last 20 years. Results are obtained
from the Web of Science databases using the keywords ‘aging’ and ‘decision making’.
Health, 2006]. As a consequence, more and more people will be making important work-related, as well as personal, medical, and financial decisions in later stages of their life span. Increased age is associated with decline in decision-making abilities [Thornton & Dumke, 2005], thus posing a potential challenge to better and longer lives. As a consequence, the gains in understanding the relation between aging and decision making are clear. First, describing the potential losses due to age-related cognitive decline will allow us to predict the challenges ahead, such as potential decline in productivity or people’s quality of choices of medical care. Second, understanding the causes of age-related decline in decision abilities will allow us to devise compensatory strategies and intervention programs targeting the older population: just as there may be considerable gains in investing in health care programs [Manton, Lowrimore, Ullian, Gu, & Tolley, 2007], there could be considerable gains in cognitive training of decision abilities. Unfortunately, decision scientists have paid little attention to development. For example, Newell and Simon’s [1972] Human Problem Solving pioneered the application of the information-processing approach to higher-order cognition. In their book, Newell and Simon introduced a list of important dimensions of human cognition, which included a developmental facet. However, this component received virtually no attention as the work was ‘concerned primarily with performance, only a little with learning, and not at all with development or differences related to age’ (p. 4). Mirroring Newell and Simon’s seminal work, decision-making researchers have largely neglected developmental change in cognitive function. Moreover, the little existing research on aging and decision making has focused mostly on dem-
360
Human Development 2007;50:359–366
Mata
onstrating age differences in various tasks ‘rather than examining specific mechanisms underlying the differences’ [Finucane, Slovik, Hibbards, Peters, Mertz, & MacGregor, 2002, p. 159]. Consequently, our knowledge of the aging decision maker is still poor [Hanoch, Wood, & Rice, 2007; Mather, 2006; Peters, Finucane, MacGregor, & Slovic, 2000; Sanfey & Hastie, 1999; Yates & Patalano, 1999]. Fortunately, as the review of Hanoch et al. [2007] attests, interest in the relation between aging and decision making is on the rise (fig. 1) and there is potential to ‘study the ontogenetic change of the adaptive toolbox of Homo heuristicus’ [Gigerenzer, 2003, p. 434]. The Adaptive Toolbox Approach: From Bounded to Ecological Rationality
The adaptive toolbox approach [Gigerenzer, Todd, & the ABC Research Group, 1999; Gigerenzer & Selten, 2001] is a research program concerned with higher-order cognition, and in particular, decision making. A main idea that guides this approach is that of bounded rationality [Simon, 1956]. Simon described the bounded rationality framework as the attempt to investigate how cognitive systems solve the problems they are faced with given constraints in available resources (e.g., time, computational power). This idea of bounded rationality is deeply associated with the thought that cognitive systems are fundamentally adapted to their environments. Simon illustrated this idea with a metaphor: the mind and the environment as blades of a pair of scissors. The adaptive toolbox approach coined the term ecological rationality [Gigerenzer et al., 1999] to emphasize the mind-environment link as a solution to the problem of how bounded rational beings can behave adaptively. The principle of ecological rationality proposes that simple cognitive mechanisms can achieve good performance by exploiting the structure of an environment in which they operate. In other words, ecological rationality states that the fit between mind and environment allows mechanisms to be both simple and successful. Underpinning the ecological rationality idea is the metaphor that the cognitive system relies on an adaptive toolbox of simple mechanisms – a repertoire of strategies – with each strategy tuned to a particular environment. These simple strategies sidestep the need to postulate psychologically implausible conceptions of rationality requiring unlikely levels of knowledge and computational resources. A model of mind based on an adaptive toolbox is therefore boundedly rational in the sense of relying on realistic levels of mental resources, and ecologically rational in the sense of being tuned to characteristics of natural environments. The adaptive toolbox program strives to identify different types of environmental structures in terms of quantifiable statistics, and propose simple mechanisms that take advantage of such structures to make correct decisions. One type of environment that has received attention is that in which the probability of observing an object is correlated with some of its characteristics. For example, the probability of a city name appearing in a newspaper article is correlated with that city’s population size. Given the task of determining which of two objects scores higher on a criterion, such as which of two cities has more inhabitants, can a simple mechanism exploit this statistical property? Goldstein and Gigerenzer [2002] calculated the success of a very simple algorithm that uses only recognition information (‘if you recognize one object but not the other, choose the recognized object as being
Understanding the Aging Decision Maker
Human Development 2007;50:359–366
361
the larger of the two’) in 10 environments (e.g., city sizes, mountain height) and found that it did astoundingly well, providing more than 90% accuracy in some environments. Similarly, Borges, Goldstein, Ortmann, and Gigerenzer [1999] showed that recognition can be a successful (and profitable) cue for picking high-performing companies in a bull stock market. In fact, recognition and similar cues, such as fluency, can be informative whenever the occurrence of events is systematically associated with the objects’ characteristics, in other words, when ignorance is systematic [Schooler & Hertwig, 2005]. As one would expect, single individuals and individuals in group settings have been shown to act in accordance with the recognition cue when it is informative [Goldstein & Gigerenzer, 2002; Pachur & Hertwig, 2006; Reimer & Katsikopoulos, 2004], but not when recognition is not a valid cue [Oppenheimer, 2003]. Successful Aging and the Adaptive Toolbox Approach
‘Win some, lose some,’ so the saying goes. Striking a balance between losses in basic abilities and gains in knowledge is, according to the selection, optimization, and compensation (SOC) framework, the key to successful aging [Baltes, 1997; Baltes & Baltes, 1990; Baltes, Staudinger, & Lindenberger, 1999]. The SOC approach proposes that successful aging involves the selection of actions that fulfill a specific goal given a person’s resources and environmental demands. For example, in old age, to behave adaptively, one may try to compensate for losses in the speed with which cognitive operations can take place by using different strategies or selecting different environments in which to act. Research on aging has described well the gains and losses in intellectual resources and how these relate to behavioral performance in fairly low-level domains, such as memory [e.g., Baltes et al., 1999; Schaie, 1994]. In contrast, there is no comprehensive theory of environments and how changes in intellectual functioning are related to cognitive performance in different settings. Consequently, the need for a theory of environments and their relation to successful behavior in later life is clear. The proposal of Hanoch et al. [2007] of connecting the adaptive toolbox approach to theories of aging may help fill this gap. The adaptive toolbox approach [Gigerenzer et al., 1999] shares with the SOC framework the assumption that, to perform adaptively, individuals have to strike a balance between their resources and the task demands by selecting the appropriate tool from their repertoire. The study of adaptive strategy selection may prove a particularly interesting test bed for theories of successful aging. First, it is a domain with real-world significance and applied value, and thus amenable to direct measurement of success. Second, the decisionmaking domain allows examining adaptive strategy selection as a function of individual resources and environment structure. A few efforts have now been made to understand the relation between aging and successful, adaptive decision making. For example, Pachur, Mata, and Schooler [in preparation] investigated the adaptive use of recognition in a probabilistic inference task in which young and old adults had to either infer which of two German cities had more inhabitants, or which of two diseases was more prevalent in Germany. In the cities environment, recognition is a very predictive cue: choosing the recognized object in a pair in which only one object is recognized will lead to almost 90% cor-
362
Human Development 2007;50:359–366
Mata
rect responses. In contrast, in the diseases environment, using recognition as a cue will provide a correct answer around 60% of the time, and thus should be used less often as a cue compared to the cities’ environment. Both young and old adults performed well and adjusted the use of the recognition cue as a function of environment structure. However, old adults showed overall increased reliance on the recognition cue, and the use of the recognition cue was negatively related to measures of processing speed and inhibitory function, suggesting older adults’ increased reliance on recognition was due to deficits in fluid intelligence. Similarly, Mata, Schooler, and Rieskamp [in press] have recently investigated the impact of cognitive aging on the ability to select more complex inference strategies as a function of environment structure. In their experiment, participants made decisions in either an environment that favored the use of information-intensive strategies or one favoring the use of simple, information-frugal strategies. The results indicated that both younger and older adults seem to be adaptive decision makers in that they adjust their information search and strategy selection as a function of environment structure. Nevertheless, old adults tended to look up less information and relied more on simpler, less cognitively demanding strategies compared to young adults. In accordance with the idea that age-related cognitive decline leads to reliance on simpler strategies, measures of fluid intelligence explained age-related differences in information search and strategy selection. Overall, the results of the two studies support the idea that cognitive aging is associated with some loss in adaptivity through increased constraints on the strategies available to older adults. Nevertheless, the results also suggest that older adults are adaptive and may still be able to perform successfully by relying on simpler strategies in the appropriate environment. In other words, simplicity can pay. Future research should investigate the statistical structure of environments in which younger and older adults routinely perform decisions and evaluate whether simple strategies provide satisfying outcomes. Even in those cases in which simple strategies are not the most accurate, they may perform nearly as well as more cognitively demanding ones. For example, Fasolo, McClelland, and Todd [2007] conducted simulations in the consumer domain (e.g., digital cameras) to show that consumers can neglect most product information and still make good enough decisions in some circumstances, namely, when there is little conflict between attributes or when there are clear differences in importance between attributes. The Ecological Rationality of Emotional Decision Making
Hanoch et al. [2007] raised the possibility that older adults’ deficits in strategy selection and application could possibly be offset by increased reliance on emotional decision making. Nevertheless, Hanoch et al. remained ‘largely agnostic as to whether the shifts towards greater reliance on emotions lead to better (or worse) decisions’ (p. 5). An important goal of future research is to assess in which circumstances reliance on emotions is adaptive, that is, we must study the ecological rationality of emotional decision making. The relation between emotion and decision making is likely to be complex: a possible divide-and-conquer strategy to study these issues will allow us to understand the different ways in which emotion can influence decision outcomes.
Understanding the Aging Decision Maker
Human Development 2007;50:359–366
363
Emotion as Information: Cue Selection Emotion can contribute to a decision outcome as a piece of information or cue to be fed into the decision process. This is the concept underlying the affect heuristic [Finucane, Alhakami, Slovic, & Johnson, 2000] and the somatic marker hypothesis [Bechara & Damásio, 2005]. In these proposals, emotion serves as an indicator of value which can be utilized to decide between options. In this case, the success of the emotional cue in inference lies in its validity, that is, its ability to predict the criterion of interest [Gigerenzer et al., 1999]. For example, saying that the recognition cue has a 0.9 validity in the German cities environment means that in 90 out of 100 pairs of cities in which one city is recognized and the other is not, the recognized city has more inhabitants. Unfortunately, we know little about the validity of emotional cues, how they fare compared to nonemotional cues, such as recognition, and whether older adults are able to rely more on these successfully. However, this is an issue amenable to scrutiny: the validity of an emotional cue such as liking or dreading can be examined empirically and compared to other cues. For example, the paired-comparison paradigm employed to study the use of recognition in inference [e.g., Pachur & Hertwig, 2006; Pachur et al., in preparation] can be extended to assess the predictive value of dread towards a particular disease. In a typical paired-comparison task assessing the use of recognition, each participant is asked to recognize as well as make inferences about pairs of objects. One can additionally calculate the validity of an emotional cue by asking participants to report the dread felt towards each object, say diseases, and then calculate in how many cases the most dreaded disease is the most prevalent – thus answering the question of how valid dread is as a cue in the diseases environment. In addition, the predictions of a strategy using dread as a cue can directly be compared to those of recognition, fluency, or other strategies, thus giving us insight into the use of different cues by young and older adults. Emotion as State: Strategy and Environment Selection Emotional and motivational states influence information processing and attentional processes which are likely to influence strategy selection and application and, consequently, decision outcomes. For example, Löckenhoff and Carstensen [2007] investigated the information search patterns of young and older adults in a task asking participants to decide between different health care plans, and showed that motivational states of the two groups influenced their attention to different types of information. Löckenhoff and Carstensen [2007] did not assess the use of particular decision strategies by individual participants so we do not know whether the agerelated differences in information search were due to differences in strategy application, such as using different weightings of an information-intensive strategy, or strategy selection, say reliance on alternative information-frugal strategies. The specific answer to this question matters to the extent that the different emotionally induced weightings or information-frugal strategies differ in their ecological rationality, that is, are adapted to different environments. In other words, according to the ecological rationality perspective, understanding the potential success of the aging decision maker will involve, first, studying the environment, namely, which statistical struc-
364
Human Development 2007;50:359–366
Mata
tures favor which strategies, and second, the aging mind, in particular how people’s strategy selection may be influenced by particular emotional or motivational states and manipulations. Another way in which emotional and motivational states may influence decision outcomes is through environment selection. The study of ecological rationality provides clues to the types of naturally occurring environments in which people make decisions. However, we still know little about how younger and older adults select environments, in particular, to what extent individuals choose environments as a function of their personal characteristics and goals. For example, it would be sensible for someone who must rely on simpler strategies to seek environments where these provide good payoffs. Likewise, some environments may be more amenable to the use of emotional cues and future research should investigate whether young and old adults are able to adaptively select environments accordingly. The Challenges and Opportunities ahead
Alan Greenspan gives proof of adaptive decision making well into old age. Unfortunately, there are many more cases of impaired decision making in later stages of the life span. Although decision and developmental scientists are increasingly concerned with decision making and aging, progress will require both methodological and theoretical advances. First, more studies are needed which include longitudinal and training data concerning different areas of decision-making competence. These will allow describing individual patterns of decline and maintenance in different decision-making components including emotional decision making, as well as assess the role of experience in offsetting some deficits due to age-related cognitive decline. Second, more detailed models of decision processes are needed. In particular, researchers need to relate theories of aging and computational models of decision making to make more precise predictions about performance as well as design interventions, decision aids, and environments that foster successful decision making by the elderly. Only then will we better understand as well as help the aging decision maker. Acknowledgment This work was supported by a postdoctoral fellowship from the German Academic Exchange Service (DAAD).
References Baltes, P.B. (1997). On the incomplete architecture of human ontogeny: Selection, optimization, and compensation as foundation of developmental theory. American Psychologist, 52, 366–380. Baltes, P.B., & Baltes, M.M. (1990). Psychological perspectives on successful aging: The model of selective optimization with compensation. In P.B. Baltes & M.M. Baltes (Eds.), Successful aging: Perspectives from the behavioral sciences (pp. 1–34). New York: Cambridge University Press. Baltes, P.B., Staudinger, U.M., & Lindenberger, U. (1999). Lifespan psychology: Theory and application to intellectual functioning. Annual Review of Psychology, 50, 471–507. Bechara, A., & Damásio, A.R. (2005). The somatic marker hypothesis: A neural theory of economic decision. Games and Economic Behavior, 52, 336–372.
Understanding the Aging Decision Maker
Human Development 2007;50:359–366
365
Borges, B., Goldstein, D.G., Ortmann, A., Gigerenzer, G. (1999). Can ignorance beat the stock market? In G. Gigerenzer, P.M. Todd, & the ABC Research Group (Eds.), Simple heuristics that make us smart (pp. 59–72). New York: Oxford University Press. Fasolo, B., McClelland, G.H., & Todd, P.M. (2007). Escaping the tyranny of choice: When fewer attributes make choice easier. Marketing Theory, 7, 13–26. Finucane, M.L., Alhakami, A., Slovic, P., & Johnson, S.M. (2000). The affect heuristic in judgments of risks and benefits. Journal of Behavioral Decision Making, 13, 1–17. Finucane, M.L., Slovic, P., Hibbard, J., Peters, E., Mertz, C.K., & MacGregor, D.G. (2002). Aging and decision-making competence: An analysis of comprehension and consistency skills in older versus younger adults considering health-plan options. Journal of Behavioral Decision Making, 15, 141–164. Gigerenzer, G. (2003). The adaptive toolbox and life span development: Common questions? In U.M. Staudinger & U.E.R. Lindenberger (Eds.), Understanding human development: Life span psychology in exchange with other disciplines (pp. 423–435). Dordrecht: Kluwer. Gigerenzer, G., & Selten, R. (2001). Bounded rationality. Cambridge: MIT Press. Gigerenzer, G., Todd, P.M., & the ABC Research Group. (1999). Simple heuristics that make us smart. New York: Oxford University Press. Goldstein, D.G., & Gigerenzer, G. (2002). Models of ecological rationality: The recognition heuristic. Psychological Review, 109, 75–90. Hanoch, Y., Wood, S., & Rice, T. (2007). Bounded rationality, emotions and older adult decision making: Not so fast and yet so frugal. Human Development, 50, 334–359. Löckenhoff, C.E., & Carstensen, L.L. (2007). Aging, emotion, and health-related decision strategies: Motivational manipulations can reduce age differences. Psychology and Aging, 22, 134–146. Manton, K.G., Lowrimore, G.R., Ullian, A.D., Gu, X., & Tolley, H.D. (2007). Labor force participation and human capital increases in an aging population and implications for U.S. research investment. Proceedings of the Natural Academy of Sciences, 104, 10802–10807. Mata, R., Schooler, L.J., & Rieskamp, J. (in press). The aging decision maker: Cognitive aging and the adaptive selection of decision strategies. Psychology and Aging. Mather, M. (2006). A review of decision making processes: Weighing the risks and benefits of aging. In L.L. Carstensen & C.R. Hartel (Eds.), When I’m 64 (pp. 145–173). Washington: The National Academies Press. National Institute on Aging. (2007). Why population aging matters: A global perspective. http://www. state.gov/g/oes/rls/or/81537.htm (accessed September 1, 2007). Newell, A., & Simon, A.H. (1972). Human problem solving. New Jersey: Prentice-Hall. Oeppen, J., & Vaupel, J.W. (2002). Broken limits to life expectancy. Science, 296, 1029–1031. Oppenheimer, D.M. (2003). Not so fast! (and not so frugal!): Rethinking the recognition heuristic. Cognition, 90, B1–B9. Pachur, T., & Hertwig, R. (2006). On the psychology of the recognition heuristic: Retrieval primacy as a key determinant of its use. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32, 983–1002. Pachur, T., Mata, R., & Schooler, L.J. (in preparation). Aging and the use of recognition in inference. Peters, E., Finucane, M.L., MacGregor, D.G., & Slovic, P. (2000). The bearable lightness of aging: Judgment and decision processes in older adults. In National Research Council, P.C. Stern, & L.L. Carstensen (Eds.), The aging mind: Opportunities in cognitive research (Appendix C, pp. 144–165). Washington: National Academies Press. Reimer, T., & Katsikopoulos, K. (2004). The use of recognition in group decision-making. Cognitive Science, 28, 1009–1029. Sanfey, A.G., & Hastie, R. (1999). Judgment and decision making across the adult life span. In D. Park & N. Schwarz (Eds.), Aging and cognition: A primer. Philadelphia: Psychology Press. Schaie, K.W. (1994). The course of adult intellectual development. American Psychologist, 49, 304–313. Schooler, L., & Hertwig, R. (2005). How forgetting aids heuristic inference. Psychological Review, 112, 610–628. Simon, H.A. (1956). Rational choice and the structure of the environment. Psychological Review, 63, 129–138. The Swedish National Institute of Public Health. (2006). Healthy aging: A challenge for Europe. http:// www.healthyageing.nu/templates/Page.aspx?id = 1258 (accessed September 1, 2007). Thornton, W.J.L., & Dumke, H. (2005). Age differences in everyday problem solving and decision making effectiveness: A meta-analytic review. Psychology and Aging, 20, 85–99. Yates, J.F., & Patalano, A.L. (1999). Decision-making and aging. In D. Park, R. Morrell, & K. Shiffren (Eds.), Processing of medical information in aging patients: Cognitive and human factors perspectives (pp. 31–54). Hillsdale: Erlbaum.
366
Human Development 2007;50:359–366
Mata
