What and When to Explain? A Survey of the Impact of Explanation on Attitudes Toward Adopting Automated Vehicles
DOI: 10.1109/access.2021.3130489
archive: archived pipeline: cataloged
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This review paper addresses the challenge of public trust and acceptance of automated vehicles (AVs), which hinders their widespread adoption despite potential benefits in safety, congestion, and emissions. The authors investigate how AV explanations—reasons provided by the vehicle to clarify its actions—impact driver attitudes and behaviors. The study synthesizes existing literature to determine the effectiveness of different explanation contents and timings, aiming to identify design principles for effective human-AV interaction. The paper organizes the literature into two primary dimensions: explanation content and explanation timing. Content is categorized into "what" (describing the action), "why" (explaining the reasoning), and "what + why" (combining both). Timing is classified as either before the AV acts or after the action has occurred. The review analyzes studies utilizing driving simulators, motion-based platforms, and virtual reality devices, examining outcomes such as trust, anxiety, mental workload, situational awareness, and driving performance. Specific experimental designs referenced include fixed-base simulators, desktop simulators, and Wizard of Oz methods, with explanations delivered via auditory or visual modalities. Key findings indicate that "why-only" explanations generally yield the best driver outcomes, associated with increased trust, preference, alertness, and reduced anxiety, while also improving safe driving performance. Conversely, "what-only" explanations are linked to the worst outcomes, including the most dangerous driving performance and lowest AV acceptance. "What + why" explanations produce mixed results; while they enhance trust, anthropomorphism, and driving safety, they also increase driver anxiety and annoyance compared to "why-only" explanations. The effectiveness of "what + why" explanations is further moderated by driving events, environmental risk, and the point of view used. Regarding timing, providing explanations before the AV acts (e.g., 1–7 seconds prior) is superior to post-action explanations, significantly boosting trust and reducing anxiety and workload. Post-action explanations show limited benefits for trust but can improve driver understanding of specific events, particularly for cautious drivers or in near-crash scenarios. The paper concludes that while simulator-based studies provide valuable insights, the field suffers from an over-reliance on simulation, which may limit external validity due to differences in driver emotion and behavior compared to real-world driving. Significant research gaps remain regarding the optimal modality (auditory vs. visual) for explanations, the precise timing interactions with content, and the role of individual driver differences. Future research should prioritize real-world studies using actual or "fake" AVs to validate simulator findings and explore mediating factors like perceived reliability and technical competence to better understand the mechanisms behind trust formation.
Key finding
AV explanations can promote trust and acceptance, but answers to basic questions about whether or when explanations are effective remain unclear; the review synthesizes existing literature and identifies future research directions.
Methodology
review
Sample size: None
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 tag_papers on 2026-05-30 (3 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | — | — | — | 1 | 2026-05-05 |
| archive | success | ddg | — | — | 6 | 2026-06-02 |
| extract | success | cached | — | — | 8 | 2026-08-22 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| enrich | success | semantic_scholar | — | — | 2 | 2026-06-16 |
| promote | success | — | — | — | 2 | 2026-06-16 |
| summarize | success | llm | qwen3.8-27b-gittensor | summ-v5 | 2 | 2026-08-22 |
| tag | success | vector_similarity | — | — | 9 | 2026-06-20 |
| verify | partial | — | — | — | 1 | 2026-05-07 |
Summary generated by qwen3.8-27b-gittensor on 2026-08-22; verification: pending re-verification.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Empirical Findings: self report data
- Theoretical Contribution: theory or model