Team conflict research, taken as a whole, has produced some conflicting results, especially regarding task conflict, which has demonstrated substantial heterogeneity across situations and an overall near-zero effect. Accordingly, several groups of scholars have called for new ways to study conflict. In this dissertation, I extend past conflict asymmetry research, which has considered only agreement between two parties, by investigating whether the two parties’ conflict inferences of each other are accurate. To do so, I draw from the Truth and Bias model of judgment to understand the biases and inaccuracies associated with conflict inferences. Further, I incorporate the partner’s conflict communication and the actor’s perspective taking as moderators to shed light on the contingencies of accurate conflict perceptions. To highlight the bottom-line implications of achieving accuracy, I use polynomial regression and link various aspects of accuracy to important dyadic outcomes.
To test my dissertation model, I collected data from ongoing student project teams using a time-lagged round-robin design. Results from the social relations modeling indicate that dyadic conflict inferences are inaccurate, being characterized by a negative directional bias, a significantly stronger bias force than the truth force, and low levels of actual similarity. Further, moderation analysis identifies the partner’s suppressive conflict communication as a contingency factor that can weaken the truth force, suggesting that more open communication can help dyad members achieve accuracy. Exploratory analysis also shows that perspective taking can strengthen the bias force. Further emphasizing the importance of accuracy, the polynomial regression results indicate that conflict inference accuracy (versus inaccuracy) is associated with higher levels of attributional confidence and problem-solving behaviors. Moreover, the level of accuracy in task conflict perceptions has an inverse U-shape relationship with problem-solving behaviors, whereas higher levels of accuracy in both task and relationship conflict perceptions are associated with lower levels of relationship satisfaction. Supporting the benefit of positive illusions, under-perception (versus over-perception) is related to higher levels of attributional confidence and relational satisfaction. In contrast to the importance of achieving accuracy (versus inaccuracy), agreement (versus disagreement) is not associated with positive outcomes. Supplementary analysis indicates that these dyadic outcomes, when aggregated to the team level, are strongly associated with team satisfaction and effectiveness.
Overall, this research suggests that focusing solely on the team level risks overlooking the existence and the various sources of inaccuracy in dyadic conflict perceptions. Further, the accuracy of the dyad partners’ conflict inferences of each other plays an important role in shaping their subsequent interactions. To the extent that conflict is a relational, multilevel phenomenon, dyadic conflict inferences should become an integral part of scholarly understanding of conflict; this perspective holds promise for not only accounting for the conflicting results in the conflict literature, but also informing managerial practices that are conducive to effective conflict management in the workplace.