Two different approaches have been employed in contemporary research to investigate the effects of aging on cognitive and intellectual abilities. One approach can be termed process analysis because it relies on task analyses or formal models to attempt to identify the specific processes in a cognitive task that are responsible for the observed age differences. This approach has used a variety of analytical methods such as subtraction, additive factors, and process dissociation to attempt to partition the variance in the target variable into theoretically distinct processes. Because the primary interest is in decomposing the variance in a single variable into different processes, process analysis research has typically involved comparing performance in one or more conditions in a single task, frequently in relatively small samples of young and old adults.

The second approach taken to investigate aging and cognition can be termed covariance analysis because it attempts to specify which combinations of variables covary together with respect to their age-related influences. A primary goal of this type of research is to partition the variance in the target variable into a portion not related to age, a portion related to age and shared with other variables, and a portion uniquely related to age. Covariance analysis research necessarily requires data from multiple variables and tasks and usually involves moderately large samples of adults across a wide age range.

One way to conceptualize the difference between the two perspectives is portrayed in Figure 1.