The attraction effect in decision making: superior performance by older adults

Abstract

Previous work showed that older adults' choice performance can be wiser than that of younger adults (Tentori, Osherson, Hasher, & May, 2001). We contrasted two possible interpretations: a general expertise/wisdom view that suggests that older adults are generally more skilled at making decisions than younger adults and a domain-specific expertise view that suggests that older adults are more skilled decision makers only in domains in which they have greater knowledge. These hypotheses were contrasted using attraction effect tasks in two different domains: earning extra credit in a course and grocery shopping, domains presumed to be of different levels of knowledge to younger and older adults. Older adults showed consistent choice for both domains; younger adults showed consistent choice only for the extra credit problem. Several explanations of these findings are considered, including Damasio's somatic marker theory and age differences in reliance on heuristic versus analytic styles.

Acknowledgements

This research was supported by a grant from the National Institute on Aging (R37 AGO4306). We thank all the people who helped us in the data collection process of this project, including Cynthia May, Cindy Lustig, Carrick Williams, Jason Blevins, Stephanie Davis, Rachelle Ta-Min, and Sudipa Bhattacharyya. We also extend thanks to John Payne, James Bettman, David Rubin, Amy Needham, Matt Serra, Andrew Mitchell, and David Goldstein for advice at various stages of this project.

Notes

¹Given that invariance or consistency is a basic assumption of rational choice theories (e.g., Tversky & Kahneman, 1986), the term “rational” is also widely used to refer to these decision patterns.

²In this research tradition, it is quite common to give multiple problems to the same participants (e.g., Huber et al., 1982), and we also wished to maximize the number of participants. Analyses comparing the participants who received one vs. two problems altered none of the conclusions based on collapsing these participants.

³Studies with college students have used both monetary compensation and course credit (e.g., Sedikides et al., 1999; Sen, 1998); participants showed the attraction effect in both instances. Thus, the fact that the two age groups received different compensation for participation should not be a factor for the attraction effect. We note that most studies in cognitive gerontology done in North America use a similar compensation scheme.

⁴A logistic regression analysis on the level of domain interest and the size of the attraction effect collapsed across both ages was not significant, β = −.13, χ²(1, N = 740) = 2.35, p = .13.

⁵When all data are reanalysed using a three-question knowledge scale (averaged over the first three questions, taking out the last question on interest), no conclusions change except that younger adults' knowledge score for grocery shopping is higher than that for extra credit, which does not match their choice patterns anyway.