A framework for analyzing and interpreting differential aging patterns: Application to three measures of implicit learning

At least four distinct explanations can be proposed to account for patterns of spared and impaired performance in which the age-related effects on som e variables are weaker than those on other variables. These are: (a) some v ariables lack sufficient reliability to exhibit relations with other variab les; (b) the variables differ in their dependence on what many of the varia bles have in common; (c) variables with little or no age differences reflec t a qualitatively different form of cognition than variables with moderate to large age differences; and (d) variables with little or no age differenc es have independent positive age-related influences that offset the negativ e age-related effects shared with other cognitive variables. These interpre tations were examined with three different variables hypothesized to reflec t implicit learning obtained from a sample of 183 adults ranging from 18 to 87 years of age. Only an implicit learning measure derived from a sequenti al reaction time task had acceptable reliability at the level of individual participants, and it was negatively related to age and positively related to variables reflecting fluid cognition. These results therefore suggest th at typical measures of implicit learning, when they can be reliably assesse d, do not reflect a qualitatively distinct type of cognitive processing nor do they seem to exhibit additional compensatory age-related influences.