Associative Memory and Trustworthiness of Artificial Faces in Young and Older Adults
Abstract
Background: Older adults generally show deficits in associative memory and increased trust in faces compared to young adults. However, little research has been conducted on older adults' associative memory and trust in artificial faces. The present study investigated young and older adults' perceived trustworthiness for real and artificial faces that were associated with either a scam or neutral condition. Methods: Participants viewed the faces before and after they were associated with either a scam or a neutral condition and subsequently rated each face on perceived trustworthiness. Participants were also tested on their memory for these associations. Results: Both young and older adults rated faces associated with a scam as being less trustworthy. However, overall, older adults rated faces as more trustworthy than young adults. In addition, young adults were the only group to rate artificial faces as being less trustworthy than real faces, and older adults did not show this difference. Young and older adults also had similar accuracy for remembering the associations of real and artificial faces. However, only young adults had higher accuracy for real faces than artificial ones, while older adults showed no difference. Conclusion: These findings illustrate that older adults may perceive and remember artificial faces differently from young adults.
Study specs
Participants viewed real and artificial faces associated with scam or neutral conditions, then rated trustworthiness and were tested on associative memory.
- Authors
- KO Alberts,AD Castel
- Institution
- University of California Los Angeles
- Discipline
- Psychology,Psychobiology,Aging Research
- Study Type
- Experimental Study
- Year
- 2025
- Human Data Platform
- Prolific
- Source
- View Source Google Scholar
Measured Outcomes
Associative memory and perceived trustworthiness of real and artificial faces across young and older adults.
Peer Review & Critical Discussion
Potential Selection Bias in 2023 Cohort
The participant pool shows a concerning overrepresentation of users from high-income demographics. Looking at Table 3, we can see that 78% of respondents had annual incomes above $75k, which significantly limits the generalizability of these findings to broader populations.
Non-naive Participants Issue
I've noticed a methodological concern regarding participant naivety. Given that Prolific users often complete multiple studies, there's a real risk that participants had prior exposure to similar experimental paradigms, which could confound the results.
RLHF Applicability to This Study Design
The implications for RLHF training pipelines are understated. If we accept the authors' conclusions about preference stability, this has direct consequences for how we should structure reward model training. The temporal decay effect described in Section 4.2 is particularly relevant.
Verify your expertise to join discussion
Create an account and verify your credentials to participate in peer discussions.