Rule-based learning among older adults: Overcoming prior beliefs for better trust-related decisions

Older adults are known to prioritize unreliable facial cues (facial trustworthiness) over relevant behavioral information (behavioral evaluation) when judging trustworthiness, increasing their risk of falling victim to fraud. This study examines age-related difference in the initial preference for facial over behavioral cues, and whether older adults could overcome such biased preference through learning. An associative memory-free, rule-based one-shot trust game was used, where 104 younger adults (aged 18-28, M_age = 21.38, SD_age = 2.41; 31 males) and 105 older adults (aged 60-77, M_age = 65.55, SD_age = 4.15; 32 males) each completed 96 trials, deciding whether to invest in trustees based on their face and behavioral evaluation of trustworthiness. Computational modeling revealed that compared to younger participants, older adults relied more on perceived facial trustworthiness, rather than using the diagnostic cue of behavioral evaluation at the beginning. Nevertheless, older adults were able to learn to increase their decision weight for the diagnostic cue progressively, as younger adults did, with the feedback provided during the task. Such results showcased older adults' capability to learn the implicit rules and adjust their decisions from feedback. These findings contribute to the development of interventions aimed at preventing fraud victimization and promoting decision-making quality among older adults. (PsycInfo Database Record (c) 2025 APA, all rights reserved).