Date of this Version
From The Influence of Cognitive Psychology on Testing, edited by Royce R. Ronning, John A. Glover, Jane C. Conoley, and Joseph C. Witt (Hillsdale, NJ: Lawrence Erlbaum Associates, 1987)
Nearly 30 years ago, Lee Cronbach (1957) distinguished between the two disciplines of correlational psychology, which investigated naturally occurring individual variance in behavior, and experimental psychology, which examined the effectiveness of certain treatments on behavior. Essentially, correlational psychology examined individual differences using factor analytic techniques; whereas experimental psychology attempted to eliminate individual differences using appropriate interventions. Cronbach believed that these two disciplines should join together to promote aptitude-treatment interaction (A Tl) research that would identify effective treatments for certain types of individuals. With this combined approach, different treatments could be prescribed for skilled and less skilled individuals.
The A Tl research methodology had limited success, however, because of inconsistency in findings and because of difficulty in replicating some of the treatments (Tobias, 1985). In addition, results rarely revealed disordinal interactions (which indicate that treatments differentially affect those on the lower and the higher ends of the performance continuum) . One explanation for the lack of disordinal interactions was that methods for identifying skilled and less skilled students on a given academic behavior were not far advanced (Tobias, 1985). What was needed were precise methods for measuring specific skills required for successful academic achievement.
Recent developments in cognitive psychology have provided more precise methods that may help to advance both A Tl research and the field of measurement. Sternberg (1977), for example, has investigated the underlying cognitive processes in intellectual behavior using componential analysis. Essentially, componential analysis investigates the underlying componenets involved in task performance. By specifying these components and the various combination rules one might employ, differences can be observed among individuals in the number of components utilized, the combination rules employed, the order of the component operations, the mode of processing (e .g., serial vs . parallel), and the time required to execute a component. This approach is more precise than the previous factor analytic approaches of correlational psychologists, because the latter measured only the end products of behaviors and not the components of mental organization (Vernon, 1970, p. 100).