Models of intelligence in late life: Comment on Hultsch et al. (1999)

The authors comment on D. F. Hultsch, C. Hertzog, B. J. Small, and R. A. Di xon's (1999) report of the latest results of the Victoria Longitudinal Stud y, which are presented as a test of the engagement hypothesis, and on their reanalysis of the data from the Veterans Study (Pushkar Gold et al., 1995) , which proposes an alternative model of intellectual function. Differences between the methodologies, variables, and concepts of engaged lifestyle in the 2 studies lead to different tests of the engagement hypothesis and dif ferent conclusions. Failure to report specified paths and unjustified assum ptions lessen the credibility of the Hultsch et al. model for the Veterans Study data. The authors conclude that their original model, which indicates a small but significant effect of engaged adult lifestyle on the maintenan ce of intellectual verbal abilities in later life, provides a better explan ation of the Veterans Study data.