THE USEFULNESS OF A MULTIVARIATE PARADIGM FOR RESEARCH IN HIGHER EDUCATION WAS EXPLORED IN THIS INVESTIGATION. THE INSTRUCTIONAL PROCESS WAS CONCEPTUALIZED AS FOUR CLUSTERS OF INTERACTING VARIABLES--LEARNER, ENVIRONMENTAL, INSTRUCTOR, AND COURSE VARIABLES. THESE VARIABLES ARE DESIGNATED AS THE "INSTRUCTIONAL GESTALT." FOR METHODOLOGICAL REASONS, THE RESEARCH WAS CONDUCTED IN FIVE TELEVISED COURSES. "CRITICAL SUBJECTS" POSSESSING THE DESIRED COMBINATIONS OF LEARNER CHARACTERISTICS WERE IDENTIFIED IN EACH COURSE AND ASSIGNED TO EXPERIMENTAL GROUPS COMPRISING A FULL SET OF CRITICAL SUBJECTS (ALL COMBINATIONS OF LEARNER VARIABLES). THESE SETS WERE RANDOMLY ASSIGNED TO RECEIVING ROOMS PRESENTING VARIOUS CONFIGURATIONS OF INSTRUCTOR AND ENVIRONMENTAL VARIABLES. THREE CRITERIA OF PERFORMANCE WERE USED--TWO WERE END-OF-COURSE ACQUISITION MEASURES AND A THIRD WAS A MEASURE OF "THOUGHT RELEVANCE." THE DESIGN IN EACH COURSE GENERATED AN ANALYSIS OF VARIANCE MATRIX FOR EACH OF THE CRITERIA. THE PARADIGM WAS DISCUSSED AS A USEFUL STRATEGY FOR RESEARCH IN THAT IT HELPS TO CLARIFY CERTAIN DYNAMICS OF THE EDUCATIONAL PROCESS WHICH ARE ORDINARILY OBSCURED. ORGANIZING CONCEPTS WERE PRESENTED TO INTEGRATE THE RESULTS, AND STUDY IMPLICATIONS WERE MADE. (JC)