Validating virtual reality for public speaking research and intervention: comparing anxiety, voice, and fluency responses to real and virtual audiences

Introduction Public speaking (PS) is a widespread activity required in many personal and professional settings. This activity is known to elicit anxiety, subsequently affecting oral communication, especially voice and speech parameters. As mastering PS skills requires practice in situations that are as similar as possible to reality, virtual reality (VR) may represent a promising method for research, training and intervention in this domain. However, it is of paramount importance to first validate VR environments in their ability to reproduce authentic anxiety responses and communicative behaviors, which are often overlooked. Methods Therefore, this study examined university students (N = 60) anxiety responses (self-reported and heart rate) as well as voice and fluency adjustments to a PS task performed either in (1) a real meeting room in front of an audience, (2) a virtual meeting room in front of an audience, and (3) the same virtual meeting room without any audience. As this last condition contained no anxious stimulus, it was included to act as a control for the anxiety induced by VR immersion. The main objective of this study was to examine the influence of the real vs. virtual nature of the audience on anxiety, voice and fluency parameters. Results Our results showed that the virtual audience elicited changes in anticipatory anxiety (increased heart rate and self-reported anxiety) compared to the control condition. The participant’s strong feeling of presence and lack of side effects such as cybersickness support the acceptability and usability of the virtual environment. Discussion Our results extend previous data and support the feasibility and relevance of using VR for PS. Additionally, we describe different VR immersion profiles among participants.