Abstract
We developed and tested a partner agent for training distributed teams on a DDD/AWACS simulation. We compared 3 conditions that were identical on test trials but differed on practice trials, which had either human partners, agent partners, or no partners (control). Each condition had 20 four-member teams who trained for about 8 hrs each. Relative to the control condition, practicing with human partners and agent partners afforded about the same substantial advantage. Comparisons with formative testing suggested that the advantage of the agent partner depended on the agent responding quickly and adaptively to trainee demands. The results extended the foundation for developing and testing agent-based training in operational settings.
Get full access to this article
View all access options for this article.
