Assessing Trainee Performance in Field and Simulation-based Training: Development and Pilot Study Results

Motorized Patrol Operations (MPO) provide Marine Corps commanders with effective means to watch and protect large areas of operations, as well as the ability to quickly adapt to different mission requirements. Marine leaders conducting a patrol must be vigilant and, above all, well trained, as constant changes in enemy tactics and an increasing need for motorized patrols require MPO leaders to make quick and effective decisions without command guidance. The actions taken by an MPO leader are dictated by split-second analysis and rapid decision making, which are difficult skills to teach and assess. Pre-deployment training takes a crawl-walk-run approach to teaching these skills, frequently involving a combination of classroom instruction, live field exercises and increasingly, simulated exercises. Effective performance assessment and feedback are important to both live and simulated training. Without appropriate performance measures, it is difficult to systematically assess the readiness of trainees and the effectiveness of the training curriculum. Furthermore, without any guiding instructional framework, feedback given to trainees is dependent on the style of the individual instructors. Current methods of performance assessment for MPO still are largely informal in nature, relying on the abilities of instructor/observers to accurately remember or take notes on key aspects of performance. This can result in excluding critical aspects of performance and instructor bias in interpreting the results of the training exercise. In this paper, we describe a recent effort to develop observable performance measures that assess MPO leader performance during both live and simulated training, using the COMPASS Methodology. We also describe the results from a pilot study in which we assess the utility and usability of these performance measures in both simulated and live exercises, as implemented in an observer-based assessment tool. We conclude with a discussion of potential implications of our results and future directions this research may take.