Research on the use of discussion in adult education has been largely concerned with comparing it with the use of other instructional techniques and with measuring opinion change. Many studies, such as Kurt Lewin's study of food habits, have compared the effectiveness of group discussion as contrasted with lecture in changing opinions and behavior. The contradictory findings of the studies suggest that the explanations may be not in the technique used but in the educational status and cultural background of participants, the subject matter, and the facility of the leader of discussions. Studies of leadership have shown the importance of the concept of the leader as a neutral agent to facilitate consensus and also have shown that directive leadership is favored by those with less than average education and is resented by the better educated. The technique may be improperly understood and employed; a definitive study of large groups of varying purposes is called for. Such factors as optimum size, educational level of participants, objective of the discussion, subject matter, leadership qualifications, and use of a second technique, should be studied. (sg)