Strategic Difference Modulate Prediction Error Sensitivity across Age Groups
Date Issued
2015
Date
2015
Author(s)
Tang, Yong-Jheng
Abstract
In this present study, we examined how young and older adult differences in the complexity of strategy during value-based decisions involving integrating probability and magnitude contribute to age differences in neural prediction error responses for outcome learning. We hypothesized that more older adults adopt a less complex strategy than younger adults, with less effective weighting of probability or magnitude information for decisions, resulting in neural prediction error differences upon receiving feedback. Moreover, prediction error responses should not dissociate older from younger adults adopting similar strategies. We assessed complexity of decision strategy and prediction error responses in young and older adults using a lottery choice task in a functional magnetic resonance imaging (fMRI) experiment. Our results show that older adults had reduced gain prediction error (positive neural response) but greater loss prediction error (negative neural response) in fronto-striatal regions. In addition, whereas younger adult bilateral insula responded to losses, this was not observed in older adults. Further, a greater proportion of younger adults used more complex strategy than older adults. Older adults in the same strategy group as younger adults demonstrated comparable fronto-striatal prediction error responses. However, strategy did not influence age effects on insula responses. These findings suggest that decision strategy preference contributed more to fronto-striatal neural prediction error rather than biological age effects on feedback learning. Once strategy during choice is accounted for, age differences in fronto-striatal prediction error during feedback were eliminated. Notably, age differences in insula response to losses were insensitive to strategy suggesting a more prominent role of biological aging in this brain area for decision-making and feedback-learning.
Subjects
aging
decision-making strategy
feedback-learning
prediction error
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-104-R01454008-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):aaf800eec4f63026dc8834083adeae50