Attempted decomposition of age-related influences on two tests of reasoning.

Two studies were designed to determine which of several hypothesized components were primarily responsible for the age-related effects on matrix reasoning (Study 1) and analytical reasoning (Study 2). In neither case were the analyses successful because the variables selected to assess the hypothesized components failed to exhibit discriminant validity and also had little unique age-related influences. These results. along with the results of other analyses, suggest that large proportions of the age-related effects on different cognitive variables are shared and are not independent of one another. Implications of these findings for the interpretation of age-related influences on measures of cognitive functioning are discussed.