With the recovery of the global economy, commercial aviation travel has dramatically increased over the last several years and continued growth is projected at a rate of 6.5% per year. Aircraft manufacturers have an unprecedented backlog of commercial aircraft orders to meet the demand. This, in turn, is causing a surge in the demand for civil pilots to fly these new aircraft with a total of 637,000 new pilot jobs projected over the next 20 years. The increase in pilot demand will stress current commercial training system facilities. With technology advances in head-worn augmented reality and virtual reality systems (AR/VR), the potential exists to leverage this technology to increase training throughput. However, the question arises as to what will be acceptable to federal aviation regulators and aircrew in-training. There has been recent emphasis on flight crew coordination competencies being factored into refresher and type rating training. Three specific competencies: Communication, Leadership and Teamwork, and Workload Management, are prominent areas where AR/VR technologies can be utilized. In an AR/VR environment, the question arises as to how best to portray a human avatar to provide the second crew member. Previous research in the use of avatars for human stimuli theorizes that avatars with high human-likeness may evoke feelings of eeriness in the human observer (Borst & de Gelder, 2015). This paper presents an experiment that was performed to study the best representation of a pilot avatar for the purposes of competency-based training. The study utilized 4 different avatar representations portrayed in a virtual 787 cockpit displayed in an HTC Vive headset. Sixteen personnel with various piloting experience went through a within subjects experimental design with two factors. The two factors were realistic vs. abstract representation and full body vs. partial body. The results of the study are presented herein.