Abstract
Two models of group leadership are articulated: (1) a group leadership influence model, and (2) a leaderless group discussion model. A group leadership influence model assumes that leaders are more effective when they are active within the group and that this activity level is positively associated with group productivity. A leaderless group discussion model maintains that effective leadership is more docile and that the group's overall leadership activity versus the activity of an individual leader is a better predictor of group productivity. The assumptions of these two competing models of group leadership were tested in a laboratory descriptive study. The results suggest that a leaderless group discussion model provides a better understanding of the data. Implications for group leadership from a leaderless group discussion perspective are discussed.