⬍ .10. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱⴱ p ⬍ .001. 6 REED, MIKELS, AND LÖCKENHOFF
to many choice domains. Findings from the behavioral component confirm that age effects extend beyond self-report measures, supporting the construct validity of choice preferences. The present research also reports the first evidence for an empirical link between choice preferences and information search. Consistent with our hypotheses, older adults desired fewer options for the behavioral decision task and viewed less information than younger adults. Whereas the average younger adult desired roughly half as many options as were available and viewed nearly two thirds of the information grid, the average older adult desired only one third of the total number of options and failed to view half of all available information. Combined with results linking choice preferences and information search, these findings raise the possibility that older adults seek less information because the amount of choice and/or available information exceeds their preferences. This notion is buttressed by the observation that age differences in information search are exacerbated in studies where relatively more information is available to view (Mata & Nunes, At the same time, the lack of support for the hypothesized mechanisms underlying age differences in choice preferences is puzzling. Although we tested fora wide range of possible explanations, none of the covariates could account forage differences.
The finding that maximizing did not play a role is especially surprising given prior research that maximizers prefer more choice relative to satisficers (Dar-Nimrod et al., 2009) and that older adults are less likely to maximize than younger adults (Tanius et al., 2009). It is possible that eliciting choice set size preferences through a hypothetical and abstract self-report measure one that provided no incentive to select the optimal choice—may have diluted the effect of maximizing. However, maximizing also failed to account forage differences in behavioral choice preferences. Furthermore, in contrast to previous findings in younger adults (Reed et al., 2012), DMSE was not associated with choice set size preferences. It is possible that this effect is specific to student samples or does not generalize across contexts. The mechanisms underlying the age trajectory of choice preferences therefore remain unclear. There are several important limitations to the present study that should be noted. First, although our findings cast doubt on a range of plausible theoretical explanations forage differences in choice preferences, ruling out these factors will require future replication using alternative measures and more diverse samples. Additional factors that were not measured could potentially explain age differences in choice preferences. For instance, although we incorporated measures of both crystallized (i.e., vocabulary) and fluid cognitive abilities (i.e., STM), we did not measure working memory. Because working memory is critical to effective decision making (fora discussion, see Mather, 2006) and declines with age across adulthood (fora review, see Salthouse, 2004), it may influence age differences in choice set size preferences. Alternatively, it is possible that older adults prefer fewer options than younger adults because of their accumulated life experience with respect to decision making. For instance, they might be more cognizant than younger adults of the cognitive burdens and potential regret associated with decisions among larger versus smaller choice sets. Thus, their preference for reduced choice may reflect age-related development of everyday decisional wisdom and strategies for dealing with choice. It is also possible that the driving factors behind choice set size preferences may not be accessible to conscious thought or insight and therefore ill-suited to measurement via self-report. This interpretation is supported by mounting evidence that unconscious, automatic, and/or intuitive processes play a significant role indecision making (Simonson, Future research on choice preferences would benefit from considering abroad array of factors such as these.
Given the lack of reliable correlates of choice preferences observed in the present study, it is also possible that age differences in choice set size preferences are the product of cohort effects, as opposed to developmental changes. Older adults formative years are likely to have occurred before the recent proliferation of choice, whereas contemporary younger adults live in an era of unprecedented choice in almost every domain imaginable (Iyen- gar, 2010; Schwartz, 2004). Thus, older adults may desire fewer options than younger adults because they are relatively more accustomed to limited choice indecision environments, an interpretation that could be tested by adopting longitudinal, as opposed to cross-sectional, designs. 6 Although the age trajectory of choice preferences may result from cohort effects instead of age effects per se, the practical implications of the present findings are nonetheless significant. For instance, given that older adults desire fewer choices across a wide range of domains, and because having too much choice is often counterproductive to decision quality and satisfaction (fora review, see Scheibehenne, Greifeneder, & Todd, 2010), public policymakers would benefit from considering the present findings when designing decision contexts for individuals of varying ages. For instance, the Medicare Part D prescription drug plan offers a degree of choice—typically dozens of options—that individuals of all ages, but especially older adults, would likely find excessive. Restricting choice sets for such decisions maybe more beneficial than detrimental. Conversely, it could be argued that public services that already offer comparatively restricted choice sets (e.g.,
Medicare Parts A and B, Social Security) maybe serendipitously tailored to the choice preferences of their older beneficiaries, in which case, increasing choice for those domains would be mal- adaptive. Reducing discrepancies (and promoting concordance)
between available choice and individual preferences for choice would therefore benefit decision makers across the life span. 6 It should be noted that the age differences in choice preferences observed by Rozin and colleagues (2006) appeared relatively stable across cultures. At first blush, this could be interpreted as evidence against the cohort effect explanation. However, their study incorporated only Western cultures (i.e., the United States and five European countries, and the authors themselves argued that the historical increase in choice among foods pervaded the developed world. Thus, study samples across all included countries were likely to contain latent age differences in the experience of food choice parallel to the observed differences in preferred choice.