| Abstract: | This study examines whether the first group member to advocate a position in the group publically has disproportionate influence on the group decision, and whether group discussion by computer mail or face-to-face determines the extent of the first advocate's influence. Possible explanations of first advocacy influence were tested in an experiment that allowed simultaneous observation of three independent variables: (1) assignment of first advocate (self-selected/randomly assigned), (2) early discussion before advocacy (discussion/no discussion), and (3) mode of communication (face-to-face/electronic mail). It was assumed that a first advocate effect exists if the group choice is closer to the first advocates' stated positions than to the group's average pregroup preferences. An “influence” hypothesis predicts that first advocates who self-select will be closer to the group decision than average pregroup preferences because these are first advocates most likely to persuade other members of the group. A “listening” hypothesis predicts that first advocates who hear an early group discussion will be closer to the group decision than average pregroup preferences because these first advocates are able to anticipate the group decision as a result of listening to what is communicated in group discussion. The results indicated strong support for the “listening” hypothesis. When groups held early discussions before someone advocated a position, the first advocate effect was observed; when groups did not hold early discussions but began their task by having someone advocate a position, there was no first advocate effect. The data suggest that the content and tone of electronic group discussions was qualitatively different from face-to-face group discussions, but the process of group decision making in both conditions was about the same. The implications of these results for group discussion and computer-mediated group decision making in organizations are discussed. |
