Voice Similarity and its Impact on Cognitive and Affective Trust in Automated Vehicles

Zhang, Qiaoning · 2025 · Human Factors

DOI: 10.1177/10711813251364804

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates how the voice characteristics of automated vehicles (AVs) influence user trust, specifically addressing the gap in research regarding how explanation delivery methods affect cognitive and affective trust. While prior work has focused on the content of AV explanations, this research applies similarity-attraction theory and the Computers Are Social Actors (CASA) paradigm to examine whether users trust AVs more when the vehicle’s voice matches their own age and gender. The authors hypothesize that voice similarity positively influences both cognitive trust (belief in competence and reliability) and affective trust (emotional connection and comfort). The researchers conducted an online between-subjects experiment with 326 licensed U.S. drivers, recruited via CloudResearch. Participants were categorized by gender (male or female) and age group (younger adults, 18–25; older adults, 55+). They were randomly assigned to either a similarity condition, where the AV’s voice matched their demographic characteristics, or a dissimilarity condition, where it did not. Participants viewed six video clips simulating autonomous driving scenarios in urban, highway, and rural environments. The AV provided verbal explanations of its decisions using AI-generated voices (younger: “Natalie”/“Nate”; older: “Charlotte”/“Jim”). After each scenario, participants rated their cognitive and affective trust using a seven-item scale adapted for the AV context. The results confirmed that voice similarity significantly increased both cognitive and affective trust. Participants in the similarity condition reported higher cognitive trust (M = 5.277) than those in the dissimilarity condition (M = 5.112). Similarly, affective trust was higher in the similarity condition (M = 3.741) compared to the dissimilarity condition (M = 3.445). Further analysis revealed distinct impacts for specific demographic matches: gender similarity significantly influenced both cognitive and affective trust, whereas age similarity significantly affected only affective trust. Manipulation checks confirmed that participants accurately perceived the gender and age of the AV voices. These findings demonstrate that personalized voice design is a critical factor in human-AV interaction, enhancing both rational assessments of competence and emotional engagement. The study highlights the importance of distinguishing between cognitive and affective trust, showing that they are shaped by different psychological pathways. Practically, the results suggest that AV developers should incorporate customizable voice options to foster user trust and improve acceptance. Theoretically, the work extends similarity-attraction theory to the AV domain, reinforcing that social cues embedded in system design significantly influence user perceptions. Future research is recommended to explore other voice attributes, such as accent and intonation, and to examine how trust evolves with repeated use.

Key finding

Users reported significantly higher cognitive and affective trust in automated vehicles when the explanation voice matched their own gender and age, with gender similarity influencing both trust types and age similarity primarily affecting affective trust.

Methodology

survey

Sample size: 326

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via unpaywall on 2026-05-07 (4 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• acceptance adoption