In addition, even when there are age-based group differences in cognitive performance, there is extensive interindividual variability for specific cognitive skills within age groups. For instance, Schaie (1996), although consistently reporting mean cross-sectional differences in overall cognitive performance, pointed out impressive variability within age groups. To quantify this variability, Schaie (1988) investigated the overlap in distributions of cognitive performance among young adults and the elderly. Even in the group of eighty and over the overlap was about 53 percent, scoring well above the mean of their age group. In other words, half or more than half of individuals in the late age groups perform comparably to a group of young adults on measures of both crystallized and fluid cognition.

Moreover, there is also a considerable amount of interindividual variability in the longitudinal patterns of decline, maintenance, and improvement. Specifically, Schaie and Willis (1986) categorized older individuals (the group mean age was 72) into those who decline and those who remained stable in their performances on the Primary Mental Abilities Test (using the space and reasoning subtests) over a period of fourteen years. Forty-seven percent of the sample remained stable on both measures, whereas only 21 percent declined on both measures. Some of these individuals were followed into their eighties, and virtually none of them showed universal descent across all five subtests of the Primary Mental Abilities Test (Schaie, 1989). It is thought that those who show age-related maintenance and improvement in cognitive development differ from those showing decline on a constellation of factors, including educational background, occupational pursuits, health history, life habits, and such personality styles as rigidity and flexibility (Schaie, 1996).

The trend of cognitive development across the lifespan, however, appears to be yet somewhat different for practical skills. Williams et al. (1983) interviewed men and women over the age of 65. The questions posed to these adults had to do with their perception of age-related changes in their skill to think, reason, and solve problems. Surprisingly enough, the responses obtained from these adults were largely contradictory to the view that late development of cognition consists of decline (see Berg, in press, for review). In the Williams et al. study (1983), 76% of the elderly adults believed that their skill to think, reason, and solve problems had actually increased over the years, with 20% reporting no change and only 4% reporting that their skills had declined with age. The researchers confronted the participants with the overwhelming evidence of decline in conventional test performance upon completion of formal schooling, but the explanation of the elderly people was that they were talking about solving kinds of problems different from those found on psychometric tests. The problems they had in mind when answering the interviewer's questions were those of an everyday or financial nature. Of course, these responses might be simply discounted as self-deceiving and self-reassuring, but a number of formal psychological studies within the last decade have provided significant support for the claim made by the elderly in the Williams et al. (1983) study.

In particular, the idea that practical and academic skills might have different developmental trajectories was supported in a number of studies (see Berg and Klaczynski, 1996, for a review). Denney and Palmer (1981) were one of the first research teams to demonstrate this discrepancy. They compared the performance of adults (aged 20 through 79) on traditional analytical reasoning problems (e.g., a "twenty questions" task) and a problem-solving task involving real-life situations (e.g., "If you were traveling by car and got stranded out on an interstate highway during a blizzard, what would you do?"). One of the many interesting results obtained in this study was a difference in the shape of the developmental function for performance on the two types of problems. Performance on the traditional problem-solving task or cognitive measure declined almost linearly from age 20, onward. Performance on the practical problem-solving task increased to a peak in the 40- and 50-year-old groups, declining thereafter. Expanding on this line of research, Smith and colleagues (Smith, Staudinger, and Baltes, 1994) compared responses to life-planning dilemmas in a group of younger (mean age 32) and older (mean age 70) adults. Unlike the results of studies of aging and academic skills, which demonstrated the superior performance of younger adults over the elderly, in this study, young and older adults did not differ. In addition, each age-cohort group received the highest ratings when responding to a dilemma matched to their own life phase.