چکیده (انگلیسی):

Computer-based assessments are popular means to measure individual differences, including age differences, in cognitive ability, but are rarely tested for the extent to which they correspond to more realistic behavior. In the present study, we explored the extent to which performance on an existing computer-based task of multitasking (‘cooking breakfast’) may be generalizable by comparing it with a newly developed version of the same task that required interaction with physical objects. Twenty younger and 29 older adults performed both the computer-based multitasking task and its laboratory-based equivalent. In each task, two measures determined prospective memory and one measure ascertained speed of completion. The Prospective-Retrospective Memory Questionnaire determined self-reported memory failures. In both age groups, correlations between the computer-based and the laboratory version of the task were largely restricted to a single measure of prospective memory. Whereas performance on the laboratory task correlated with self-reported memory failures across the entire sample, there was no such relationship for the computer-based task. Finally, age group by task interactions suggested that younger and older adults were differentially affected by laboratory versus computer-based assessment. Our study illustrates the need for future evaluations of computer-based psychometric instruments on younger and older samples prior to their application.