Designing and Testing an Experimental Framework of Affective Intelligent Agents in Healthcare Training Simulations

Abstract

The purpose of this study is to investigate how emotionally enabled virtual agents (VAs) in healthcare provision training simulations allow for a more effective level of understanding on how an emotionally enhanced scenario can affect different aspects of learning. This is achieved by developing virtual agents that respond to the user’s emotions and personality. The developed system also provides visual and auditory representations of the virtual agents’ state of mind. To enable the fulfilment of this purpose an experimental framework for incorporating emotional enhancements (concentrating on negative emotions such as stress, fear, and anxiety) into virtual agents in virtual training applications for healthcare provision is designed and implemented. The framework for incorporating emotional enhancements is designed based on previous research, on psychological theories (with input by experienced psychologists) and from input of experts in the area of healthcare provision. For testing the framework and answering the research question of this thesis the researcher conducted nine case studies. The participants were nursing students in the area of healthcare provision, and more specifically in the area of mental health, specialising in caring for patients with dementia. The results of the study showed that the framework and its implementation succeeded in providing a realistic learning experience, stimulated a better set of responses from the user, improved their level of understanding on how an emotionally enhanced scenario can affect the learning experience and helped them become more empathetic towards the person they cared for.