This study experimentally examined the effects of voice type and driving context on user experience with in-vehicle voice agents in autonomous vehicles. A 2×2 within-subjects experimental design was employed with 28 participants, measuring key user e...
This study experimentally examined the effects of voice type and driving context on user experience with in-vehicle voice agents in autonomous vehicles. A 2×2 within-subjects experimental design was employed with 28 participants, measuring key user experience dimensions including intimacy, trust, intention to use, perceived safety, and perceived anthropomorphism. The results indicated that intimacy and perceived anthropomorphism were relatively higher for human voice conditions, while perceived safety was evaluated more positively in task-oriented driving contexts. Notably, an interaction effect between voice type and driving context was observed for intention to use. Specifically, synthetic voice was perceived as more suitable in task-oriented contexts, whereas human voice was perceived as more appropriate in non-task-oriented contexts. These findings suggest that the effects of voice type are not fixed but vary depending on the usage context. From a task–technology fit perspective, this study provides empirical insights into the design of in-vehicle voice agents and highlights the importance of context-sensitive voice interface design for autonomous vehicles.