Examined the relationship between conceptual systems and preferences for behavioral or client-centered approaches to therapy. One hundred and forty college students were administered measures of their conceptual systems and of their therapy preferences; the latter included descriptions of the two therapies. Two weeks later, they were exposed to audiotaped excerpts of one of the two therapies and afterward indicated again their preferences between the two therapies. It was hypothesized that therapy preferences would be distributed nonrandomly across conceptual systems, with a greater proportion of persons in the more abstract systems who preferred client-centered approaches than in the more concrete conceptual systems. This hypothesis was supported at the .001 level. It also was found that therapy preferences were relatively stable, with very few Ss who actually changed their perferences after they had listened to the demonstration tapes.

137 college students were given the Myers-Briggs Type Indicator and O. J. Harvey's This I Believe test for conceptual systems. For each of the four Myers-Briggs scales, the proportion of each type classified in each of the four main conceptual systems was analyzed. It was hypothesized that sensing and intuitive types would be nonrandomly distributed across conceptual systems, such that the lower conceptual systems would contain higher proportions of sensing types while the higher conceptual systems would contain higher proportions of intuitive types. This was confirmed. An additional unhypothesized trend emerged in which feeling types were overrepresented among System I individuals. The main results were interpreted as supporting the construct validity of the Sensing-Intuition scale of the Myers-Briggs Type Indicator.