Can unreliable auditory hazard warnings help the driver? The effect of timing errors and false alarms on road hazard detection in dynamic road scenes
DOI: 10.1186/s41235-026-00718-w
URL: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12996480/pdf/
archive: archived pipeline: cataloged verified
Summary
Song & Wolfe (Cognitive Research: Principles and Implications, 2026) tested whether non-spatial auditory warning cues help drivers localize road hazards in real dashcam video, and whether benefits survive imperfect timing or false alarms. Experiment 1 (N=46, Univ. of Toronto Mississauga participant pool) varied cue-to-hazard onset interval; Experiment 2 added hazard-absent trials and tested 80% vs 50% cue reliability with a fresh sample. In Experiment 1, earlier cues sped responses more than later cues, and warning cues reduced response time at all tested timings. In Experiment 2, with false alarms present and a realistic low base rate of hazards, the cueing benefit was eliminated regardless of reliability level.
Key finding
Pre-hazard auditory warnings speed road-hazard localization in idealized always-hazard conditions, but the benefit vanishes once false alarms are introduced at realistic rates, suggesting classic Posner-cue benefits do not translate cleanly to dynamic real-world driving warning systems.
Methodology
lab_experiment
Sample size: 46
Quality score: 5 / 5