| Abstract: | Studied the effects of group discussion on the extremity of person perception judgements with a view towards (a) evaluating the role of alternative ‘modes of interaction’ in selectively inducing polarization or moderation of judgements, and (b) further extending the generality of the group polarization phenomenon. Two traditional theories of group shift (leadership theory, value theory) were also evaluated. Both the video-taped stimuli and the subject, object and situation-relevant measuring instrument were specially developed in a pilot study (N = 90) in order to approximate real-life conditions. The person stimuli were rated by subjects (N = 72) in a traditional repeated measures design, with a control group for familiarization effects. Free exchange of opinions in group discussion was either encouraged or discouraged. A significant polarization effect with marked postdiscussion recidivism was found when spontaneous interaction was encouraged, while group consensus judgements were more moderate in the more formalized group interaction condition. Neither of the traditional theories examined were supported by the data. The results are discussed in term of their potential relevance to real-life groups specializing in person perception judgements (juries, interviewing panels), and the connection between the mode of interaction formulation and other recent theories of the group shift are examined, with some suggestions for the further elaboration of the present model. |
