Abstract:

Presents a detailed proposal for linking experimental methods and the individual-differences approach within the framework of multivariate psychology research. Essentially, a case is made for the testing of process theories through correlations as dependent variables. Topics include: (1) the experimental manipulation of correlations for the purposes of theory testing and construct validation; (2) examples of the application of nested-factor models (i.e., task-set switching and dual-task costs); (3) the testing of factor structure theories by means of specific experimental designs and exploratory factor analysis (EFA); (4) the application of facet models for the purpose of separating content and cognitive functions, enhancing symmetry, distinguishing constructs from methods, and designing test items. Two sections are devoted to the differentiation of construct validation according to the experimental tradition versus the individual-differences approach. The final section outlines a structural equation modeling framework that allows for the integration of both perspectives.