Login / Signup

Sow in tears and reap in joy: Eye tracking reveals age-related differences in the cognitive cost of spoken context processing.

Tami Harel-ArbeliYuval PalgiBoaz M Ben-David
Published in: Psychology and aging (2023)
Older adults have been found to use context to facilitate word recognition at least as efficiently as young adults. This may pose a conundrum, as context use is based on cognitive resources that are considered to decrease with aging. The goal of this study was to shed light on this question by testing age-related differences in context use and the cognitive demands associated with it. The eye movements of 30 young (21-27 years old) and 30 older adults (61-79 years old) were examined as they listened to spoken instructions to touch an image on a monitor. The predictability of the target word was manipulated between trials: nonpredictive (baseline), predictive (context), or predictive of two images (competition). In tandem, listeners were asked to retain one or four spoken digits (low or high cognitive load) for later recall. Separate analyses were conducted for the preceding sentence and the (final) target word. Sentence processing: Older adults were slower than young adults to accumulate evidence for target-word prediction (context condition), and they were more negatively affected by the increase in cognitive load (context and competition). Target-word recognition: No age-related differences were found in word recognition rate or the effect of cognitive load following predictive context (context and competition). Although older adults have greater difficulty processing context, they can use context to facilitate word recognition as efficiently as young adults. These results provide a better understanding of how cognitive processing changes with aging. They may help develop interventions aimed at improving communication in older adults. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
Keyphrases
  • young adults
  • physical activity
  • emergency department
  • deep learning
  • machine learning