The Age Comparative Construct Validity of Speeded Cognitive Factors

The age comparative construct validity of speeded cognitive processing tasks was assessed using confirmatory factor analysis. Reaction time and card sorting tasks which varied in complexity, and which were assumed to assess aspects of memory, speed of information processing, and selective attention, were administered to 141 young (X = 24.9) and 142 elderly (X = 65.3) adults. All simultaneous models were rejected. Since the confirmatory analyses were unsuccessful, exploratory analyses were undertaken. Independent analyses yielded three factors as hypothesized for the young adults but unexpectedly five factors for the elderly adults. One extra factor for the older subjects appeared to represent strategy differences while the other may be an order artifact. Although the age comparative validity for the three primary factors seems to be supported, the results raised the possibility that speeded processing tasks with spatial content do not necessarily assess the same underlying constructs for adults at different points in the life span. These findings underscore the importance of establishing age comparative construct validity for cognitive processing tasks when the research goal is to compare the performance of diverse age groups.