Don't panic, it's a robot: how the gender of the voice of an autonomous ambulance can ease patient fears

Abstract

Human interaction with autonomous vehicles predominantly occurs through voice interfaces, providing a unique context to examine gender stereotype activation in human-technology interaction. This study investigated how synthetic voice gender and personality traits affect user perceptions of autonomous ambulances during simulated emergency scenarios. Using a 2 × 2 experimental design, we tested male and female voices paired with either dominant or friendly personality traits across three dependent variables: confidence, competence, and emotional response. Results revealed significant gender × trait interactions, demonstrating context-dependent stereotype activation rather than uniform gender preferences. Male voices generated higher confidence only when paired with dominant traits, while no gender differences emerged in friendly contexts. Unexpectedly, female voices with dominant traits produced heightened emotional arousal, contradicting assumptions about feminine voices providing universal calming effects in healthcare settings. These findings reveal situational gender stereotyping, where stereotypical responses depend on the congruence between voice gender and contextual behavioral cues. This challenges existing theoretical models that assume consistent gender effects and has important implications for autonomous emergency system design. Rather than simple gender-task matching, effective voice interface design requires careful consideration of how gender and behavioral cues interact to shape user experience, particularly in high-stakes contexts where minimizing patient distress is critical.