An investigation of the effects of normal aging on reasoning ability : A dual-process approach
Prowse Turner, Jamie
MetadataShow full item record
Research has provided a great deal of evidence that reasoning performance declines with age (De Neys & Van Gelder, 2009; Fisk & Sharp, 2002; Gilinsky & Judd, 1994; Salthouse, 2005). Understanding these age-related differences is important because reasoning is an integral part of everyday cognitive functioning, the decline of which may result in older adults relying on heuristic strategies that can result in bias (Bacon, Handley, & Newstead, 2003). The objective of the current research was to use a dual-process theory framework to explain why there are age-related differences in reasoning. In addition to a variety of reasoning tests (i.e., a syllogistic task, base rate task, and the Cognitive Reflection Test), the present dissertation included independent tests of capacity (working memory, processing speed, and inhibition) and thinking styles to account for these age-related differences. Chapter 2 focused on two recently proposed levels of Type 2 analytic thinking, algorithmic (individual differences in capacity) and reflective (individual differences in rational thinking dispositions) (Stanovich, 2009). It was hypothesized that a) both reasoning performance and capacity performance would differ with age, b) these components would contribute independently to performance on three reasoning tasks, and c) that they would explain at least some of the age-related differences in reasoning performance. Older adults demonstrated lower algorithmic capacity relative to younger adults and measures of capacity were related to performance on all three tasks. Furthermore, capacity attenuated the age-related differences in reasoning. Older adults also demonstrated a lower score than young adults on the measure of thinking dispositions; however, this predicted age-related differences only on the base rate task (and marginally on the syllogistic task). Furthermore, on the syllogistic reasoning task, a belief-bias component of reasoning was related to the age-related differences in reasoning. Chapter 3 focused on whether the performance differences between young and older adults demonstrated in Chapter 2 could be attributed to differences in metacognitive skills. Four aspects of metacognition were examined: 1) differences in conflict detection, 2) confidence in individual answers, 3) confidence in overall performance, and 4) scores on the self-report measure Metamemory in Adulthood. There was little evidence to suggest that there were difference in metacognitive ability between young and older adults, thus the results were not consistent with the hypothesis that metacognition plays a role in the age-related reasoning differences. In Chapter 3, I also investigated the hypothesis that conclusion believability and latency may be cues to confidence, and that perceived difficulty of the task may be an alternate measure of confidence, that is related to performance and response time. The data reveal that the conclusion believability and latency cues were important predictors of confidence, especially for the older adults, and that perceived task difficulty was related to performance and latency on both the syllogistic and base rate tasks. Chapter 4 focused on the effects of perspective instructions on age-related differences in reasoning. In particular, I investigated whether reasoners would engage in a more logical, analytic style of thinking when prompted to reason from an alternative perspective. Results indicated that a shift in perspective may be advantageous for the older adults. This is promising because although older adults’ limited capacity increases reliance on heuristic output and decreases processing power, there may be a straightforward way to mediate these age-related differences in reasoning ability, simply by asking them to reason from another’s perspective. Overall, this research significantly expanded the current knowledge regarding age-related differences in reasoning. Moreover, the findings were consistent with a dual-process theory of reasoning, which provided an integrative framework that accounts for the patterns of findings presented in this dissertation.
DegreeDoctor of Philosophy (Ph.D.)
SupervisorThompson, Valerie A.
CommitteeCampbell, Jamie; Crossley, Margaret; Kirk, Andrew; Toplak, Maggie
Copyright DateDecember 2012