Performance Appraisal in a Battle Simulation

Abstract

The purpose of this research was to formulate a composite measure of mission accomplishment that could be used to assess simulated combat performance of battalion command groups training command and control procedures in a computer-driven battle simulation. Mission accomplishment was defined in terms of six measurable components or objectives of a covering force/delay mission. Four retired Army officers served as judges for all experiments. Experiment I demonstrated that two objectives could be meaningfully combined into a single measure used in previous research as an index of relative losses by opposing forces. Multiple linear regression was used in Experiment II to describe how the judges assigned mission accomplishment scores to 216 hypothetical battle outcomes including measures of each mission objective. Experiment III compared judges' assessments of 10 actual battle outcomes to the mission accomplishment scores predicted by the regression models derived in Experiment II. Inter-rater agreement was high in the assessments of actual battle outcomes, and individual predictor models accounted for more than 94% of the variance in ratings of these data for three of four judges. A composite model of mission accomplishment accounted for 96% of the variance in the combined assessments of all four judges.