Teams have the potential to display high performance or low performance, depending on how well team members interact with one another. Training is commonly used to maintain or enhance the performance of various team types (e.g., sport or work teams). Intelligent Tutoring Systems (ITSs) have been used for years in multiple domains to tutor individuals. However, challenges arise when attempting to develop an Intelligent Team Tutoring System (ITTS). This current work focuses on the challenge of delivering effective feedback to teams via an ITTS designed to improve team performance. This research specifically focuses on how the assessment basis of the feedback (based on individual vs. team performance metrics) affects the team.

Specifically, this research examines how feedback displaying individual vs. team performance metrics influences team performance across multiple factors. The participants in this study performed a modified version of a classic shopping task designed to test cognition known as the Multiple Errands Test (MET). The researcher created a three-person team version of the MET called the Team Multiple Errands Test (TMET) within a virtual world on desktop computers. In three different feedback conditions, teams received performance feedback with information about individual performance, team performance, or both. Dependent measures included: performance (individual and team scores), items collected (correct and incorrect), errors, time remaining, collection time per item, and task strategy. Results were analyzed at the team and individual level. The analysis was conducted in three phases. First, the researcher analyzed the influence the feedback intervention had on main performance metrics: performance (individual and team scores), items collected (correct and incorrect), errors, and time remaining. Second, the researcher analyzed how the influence of the feedback intervention on the dependent measures depended on the strategy that teams implemented. The third analysis explored the team’s perception of performance by examining the correlation between performance and how participants viewed their teammates’ performance and their own performance

The results from the first analysis suggested that the time remaining for participants in the Team feedback condition was significantly higher than in the Individual and Team (I&T) feedback condition, suggesting that feedback containing only Team-based feedback reduced the time teams spent in a session. The results from the second analysis suggested that the time remaining for participants on teams that used a specific strategy in which team members mostly stayed close to each other (designated “Go Together”), after gaining experience with the task, was significantly greater in the Team condition than in the Individual and I&T conditions. The analysis also revealed that the average collection time per item for participants in teams that used Go Together was significantly lower in the Team condition than in the Individual condition. The analysis also showed that frustration for participants on teams that used Go Together was significantly greater in the Individual condition than in the Team and I&T conditions. The third analysis demonstrated that participants in the Team condition consistently have a correct perception of their own performance and their team’s performance. The results also showed that the influence of the feedback condition on participant’s perception of individual or team performance changed depending on task experience.

This research offers a unique contribution to the growing body of research on Intelligent Team Tutoring Systems by exploring the benefits of offering real-time feedback based on individual assessment vs. team assessment. This research also demonstrates that under some team circumstances, providing more information to teams is less effective than providing less.