Information sharing in work groups: A transactive memory approach

This dissertation project consists of three studies investigating issues related to information sharing in work groups from the perspective of transactive memory (TM). Study I examined how the antecedent, transactive processes, and the group performance outcome were related to one another. The results suggested that, inconsistent with TM theory, information allocation did not affect group performance. The implication of this result is that TM theory needs to be revised and to include contextual variables to capture information sharing processes in groups. Studies II and III attempted to achieve this goal. These two studies looked at the effects of two exogenous variables---task complexity and group knowledge composition---on information sharing processes. The results in Study II suggested that an increase in task complexity did not necessarily impair group performance. Rather, group members performed better at complex tasks than simple tasks if they had a well-constructed TM system; however, they performed worse if they experienced high relationship conflict. Relationship conflict, not task conflict, directly impaired group performance. Study III examined how group knowledge diversity affected information sharing. Diversity research so far hasn't been able to draw a uniformly positive or negative link between diversity and performance. It is likely that some intervening variables determine whether diversity would facilitate or impede performance. Study III argued that, although knowledge diversity would potentially increase group performance by providing more resources to bear on the task, it could also decrease performance by increasing disagreement on knowledge criticality judgment among group members. The results supported the hypotheses. More frequent communication among members with different expertise was proposed as one solution to this problem. The results generated in the three studies provide important managerial implications for knowledge management practices in work groups. On the theoretical side, this dissertation encourages more detailed specification of information sharing processes than what is currently prescribed by TM. Additional contextual variables and information sharing processes should be considered in order to produce the theoretical guidance that is actually actionable.