A Structural Modeling Examination of the Executive Decline Hypothesis of Cognitive Aging Through Reanalysis of Crawford et al.'s (2000) Data

This study presents a new analysis of data previously published by Crawford, Bryan, Luszcz, Obansawin, and Stewart (2000) examining the role of executive function in age-related declines in general cognitive ability and memory. Although the original authors' question is relevant to understanding the role of executive function in aging, this paper argues that the methods used can be substantially improved to more accurately answer the question. Thus, Crawford et al.'s data are here analyzed using a more parsimonious methodological approach (structural equation modeling), and the results obtained depart from the authors' results but are in line with their original hypotheses. First, for younger individuals, age was differently associated with verbal measures of cognitive ability than with performance measures. Although the relation of age to executive function was stronger than the relation of age to verbal abilities, this difference was not apparent with regard to performance abilities. Second, across samples (i.e., younger and older individuals), memory was not accounted for by cognitive ability but was related to age and executive function. Finally, also across samples, executive function was strongly linked to recall and recognition measures, accounting for the largest variance in memory. These results are discussed in relation to previous findings as well as their theoretical significance for aging research.