AAMC data describing medical education institutions were analyzed in order to raise researchable questions for subsequent examination. A suitable multivariate method was selected and its applicability tested. The method chosen was multiple factor analysis, which analyzes the pattern of correlations among large sets of variables. Variables were organized into categorical subsets to describe four conceptual components of medical education: faculty, curriculum, student, institution. Factor analysis was performed on each of the four sets and used to select descriptive variables that are representative of the dimensions of institutional variation in each educational component. The relationships of variables within components were discussed. The representative variables from the four educational components were merged and analyzed as one set. Apparent relationships among variables from separate components were described. Questions of potential interest were presented which address substantive issues, the quality of the data and possible effects of the method on the results. These questions or hypotheses may serve to guide the selection and design of focused research studies. (Author/LBH)