The effect of a dissenting opinion and private behavior on conformity to a self-aggressive group norm
Results from non-experimental studies support the notion that group pressure is associated with a variety of self-destructive behaviors, including intentional self-injurious behavior. For example, heightened rates of self-injurious behavior within a social group have been observed when one or more group members engage in self-injurious behaviors. Although intriguing, results from these non-experimental studies do not necessarily support a causal relationship between group influences and self-harm behavior. One recent experimental study demonstrated that exposure to high- or low-self-injurious group norms affects self-aggressive behavior under controlled laboratory conditions. One possible explanation for this effect is that individuals were responding to the unanimous group pressures, and may be less likely to do so either if the group is non-unanimous or if the group is unaware of individual behavior. The purpose of the present study was to determine if either of these two factors is necessary to influence self-aggression in group members and to determine if gender differences exist for this group conformity effect. Participants (71 men and 93 women) were randomly assigned to one of four social influence conditions: (1) a unanimous group in which research confederates all expressed self-aggressive behavior; (2) a majority group in which one dissenter expressed non-self-aggressive behavior which preempted the behavioral expression of other group members; (3) a majority group in which one dissenter expressed non-self-aggressive behavior after the behavior of the other group members; and (4) a baseline group in which research confederates expressed no self-aggressive behavior. Participants were also randomly assigned to one of two conditions in which either they were told that the other group members could see their task behavior (public) or they were told that no one else could see their task behavior (private). After exposure to group information, participants engaged in a task that was designed to assess self-aggressive behavior that was presented as a stimulus detection exercise. It was predicted that the presence of a single dissenter would attenuate self-aggressive behavior in participants, and that this effect would be strongest when the dissenter expressed that behavior after the group presentation but immediately before the participant engaged in the self-aggression task. It was also predicted that concealing participant behavior from the group would diminish self-aggressive behavior. Results indicated that a single dissenter in a group reduced the likelihood of self-aggressive behavior regardless of the position of the dissenter, but that concealing behavior had no effect on self-aggression. A Gender by Group effect was also found, which indicated that women were more likely to be influenced by a single dissenter than were men.