Feedforward Effects on Predictions in a Dynamic Battle Scenario

Cleotilde Gonzalez, Carnegie Mellon University
Michael Martin, Carnegie Mellon University
Jeffrey T. Hansberger, Army Research Laboratory

Abstract or Description

Commanders face many challenges in their efforts to control the battlefield. Friction (i.e., sources of delay) in the commander’s control system, coupled with the dynamics of the battlefield, requires commanders to act before threatening battlefield events occur. Effective control of the battlefield thus requires accurate predictions. This paper describes the results of a preliminary study concerned with the effect of FeedForward (FF) on the accuracy of predictions in dynamic battle situations. FF, given in the form of expert advice prior to simulated battle, did not reliably improve predictions. Exploratory analyses, however, indicate that FF guided attention to a subset of the task variables important for accurate prediction. Furthermore, FF produced quicker and more decisive victories than practice alone. In conjunction with the positive performance trend for the FF group, these findings indicate that FF facilitates strategy development and may lead to higher levels of Dynamic Decision Making (DDM) performance over time.