Abstract
This paper gives a summary of the procedures and results of 11 experiments on the effects of group discussion. A wide range of personality and situational features were manipulated as independent variables. The studies were conducted for the most part in West Germany and have been published previously in various German and English/American journals. The findings are interpreted through two principal current theories of group polarization (i.e., of the finding that group discussion enhances the initially dominant tendency of group members), the social-comparison and persuasive-arguments theories. The studies here presented (almost all providing evidence for group polarization) seem to provide somewhat greater support for the latter theory. The article also analyzes subversions of each theory and discusses the relative support these receive from the data of our studies.