Methods for the Measurement of Consensual Beliefs Within Groups

Lucian Gideon Conway Ill and Mark Schaller University of British Columbia The study of consensus in groups is fundamental to the understanding of group processes and the psychological experiences of individuals within groups. Measuring consensus in groups is tricky. This article reviews strengths and weaknesses of various methods for measuring the magnitude of consensus between persons on a single target belief. Considered are methods based on mean extremity, percentage agreement, dispersion, correlation, and spatial clustering. Specific advantages, limitations, and interpretational pitfalls are considered for each measure. Despi~ a diversity of personalities, roles, and histories, people often share very similar beliefs. This fact of consensus has fundamental conse- quences on group processes and the psychologi- cal experience of individuals within groups. Indeed, the study of consensual common ground has played an important role in many areas of small group research (Hoyle & Crawford, 1994; see Baron, Kerr, & Miller, 1992, for a review). For instance, examination of consensus is central to inquiry into how groups come to have the particular members that they do (e.g., Gailbreath, Wagner, Moffett, & Hein, 1997). In addition, because small group norms are, by definition, consensually shared beliefs, tests of group norm conformity models (e.g., Prapaves- sis & Carron, 1997) often involve the measure- ment of consensus. The study of consensus in small groups is not merely an academic issue; it has implications for the health and happiness of individuals in real-life groups. It appears that, within peer group networks, there emerge consensual norms governing health-related be- haviors such as binge eating (see Crandall, 1988); consequently, the impact of therapeutic interventions relevant to these behaviors may be enhanced by an understanding of the social