Examining the mental model convergence process using mathematical modeling, simulation, and genetic algorithm optimization.

摘要

The increasing implementation of teams in organizations has led to much research attention around team processes and performance. Uncertainty exists, however, in how team processes impact collaborative activities and, ultimately, team performance. Recent research has focused on team cognition as a potential means of explaining this uncertainty. Extending this line of inquiry, my dissertation research focuses on the interplay between teams' cognitive and communicative processes that have been implicitly linked in past team research. Specifically, I examine mental model convergence among team members as a specific type of team cognition. By integrating cognition and communication explicitly, the process of mental model convergence as it unfolds during collaborative activities may be analyzed via the verbal exchange of mental model content.Herein, I compare baseline, intervention, and optimal team communication processes to understand how the communication patterns evoking the underlying mental model convergence process of baseline teams may be changed by team interventions and how the process differs among them. Baseline team data comes from 60 student teams working in a laboratory setting. These data are also used to create a model of team communication processes, which is then implemented to simulate the communication processes of teams receiving interventions. The two types of team intervention conditions investigated include initiating collaborative activities with a specific topic discussion and delaying the start of task activities. The teams with optimal communication processes are obtained using genetic algorithm optimization procedures for combinatorial problems with multiple objectives. Specifically, the genetic algorithm evolves generations of team communication processes, beginning with the baseline data, toward optimal cost and time performance. In addition to examining the mental model convergence process, the performance of intervention teams, analyzed on a neural network generated performance assessment model, is compared to baseline teams receiving no interventions and optimal teams.Results indicate that team interventions do not improve team performance equally. Furthermore, event history analysis indicates a temporal shift in the timing of communication patterns between baseline teams and top intervention teams (i.e., the best performing teams receiving interventions). Moreover, top intervention teams have mental model convergence processes that emulate those of optimal teams.