Background: Public speaking (PS) is a complex and multidimensional activity that enables the transmission of a message from a sender to several receivers. Understanding PS mechanisms is of paramount importance as our PS proficiency can have repercussions in our daily lives. Various situational variables, such as audience attitude, appear to influence PS performance by modulating anxiety. An audience displaying negative and disengaged behavior (i.e.: averted gaze, shaking head, backward posture) has been found to increase anxiety compared to a positive and engaged audience (i.e.: mutual gaze, nodding, smiling, forward-leaning posture). However, little is known about which aspects of oral communication are impacted by audience attitude.
Objective: To determine the influence of virtual audience’s attitude on voice, speech, and anxiety parameters.
Methods: We conducted acoustic analyses on the speech productions of 64 university students in front of two types of audience : (1) a positive and engaged audience (2) a negative and disengaged audience. We analyzed multiple parameters related to anxiety (self-reported, heart rate), voice (median F0, F0 IQR, Harmonic to Noise Ratio, spectral slope, center of gravity) and speech (percentage of filled pauses, length and frequency of silent pauses).
Results: Paired t-tests indicated that self-reported anxiety (SUDS) during and after PS was significantly higher in the negative condition. Regarding oral communication, no significant differences were found between conditions when measured on the entire 3-minute PS task. However, when we analyzed the PS task in one-minute intervals, we observed a significantly higher HNR in the first and second minute when immersed in the negative condition.
Conclusions: These results can be explained by the theory of vocal response under emotional arousal which posits voice adaptations rely on different independent level (Juslin & Scherer, 2005). While the negative virtual audience was found to efficiently induce psychological stress (as evidenced by the significant increase in self-reported anxiety during and after PS), this may not be sufficient to elicit changes in all voice and speech parameters. Further research on the impact of PS situational variables on voice and speech is needed to better understand and train this crucial skill.