Educational policies proposed with and without the aid of Decision Support Systems (DSS) were compared. 120 teachers responded to a questionnaire describing two hypothetical incidents and presenting a series of options with respect to homogeneous and heterogeneous grouping of pupils according to achievement level. After selection of preferred alternatives, the participants were asked to estimate the relative importance of a number of relevant considerations. A DSS program then formulated a second decision based on calculation of the information supplied by each teacher. The findings showed that use of DSS resulted in significantly stronger support for heterogeneity. The paper discusses the differences between DSS-aided and unaided decisions and implications for future division of labor between humans and computers in decision-making processes.