Exploring Gender Biases with Virtual Patients for High Stakes Interpersonal Skills Training

The use of virtual characters in a variety of research areas is widespread. One such area is healthcare. The study presented in this paper leveraged virtual patients to examine whether virtual patients are more likely to be correctly diagnosed due to gender and skin tone. Medical students at the University of Florida College of Medicine interacted with six virtual patients across two sessions. The six virtual patients comprised various combinations of gender and skin tone. Each virtual patient presented with a different cranial nerve injury. The results indicate a significant difference in correct diagnosis according to patient gender for one of the cases. In that case, female patients were correctly diagnosed more frequently than their male counterpart. The description of that case required that the virtual patient present with a visible bruise on the forehead. We hypothesize the results obtained could be due to a transfer of a real world gender bias.