Navigation Voice Familiarity Can Affect Driving Stress and Emotions Under Complex Road Conditions: Evidence From An Experimental Study

Abstract

Navigational voice interaction is an important factor affecting driving behavior and safety under complex road conditions. However, much of the research on navigational speech has focused on aspects such as commands and tone, and little on voice familiarity. To investigate the effects of navigation voice familiarity on driving stress and emotions, this study was conducted through simulated driving and controlled experiments. Driving stress was assessed by analyzing participants’ electrocardiogram (ECG) signals and subjective data obtained from a Driving Activity Load Index (DALI) questionnaire. Driving emotions were evaluated by participants’ facial expression data and the Self-Assessment Manikin (SAM), whereas the Post-study System Usability Questionnaire (PSSUQ) scores were collected as a complementary measure of both driving stress and emotions. We found that voice familiarity had a substantial impact on driving stress and emotions. Compared with using a familiar voice, an unfamiliar voice resulted in a 26.51% increase in participants’ ECG data and a 34.12% increase in DALI total scores. In addition, the use of an unfamiliar voice was associated with a 112.50% and 200% increase in the mean intensity value of anger and disgust among participants, respectively, as well as a decrease in their PSSUQ ratings. These findings indicate that the use of unfamiliar navigation voices in complex road conditions can exacerbate driving stress and negative emotions. This study provides valuable insights for navigational Voice User Interaction (VUI) design and ultimately contributing to driving safety.

Keywords

• Human-computer interaction (HCI)
• interaction design
• usability
• voice familiarity
• voice interaction
• voice user interface (VUI)