Should I stay or should I go? Evidence accumulation drives decision making in human drivers

Abbink, David A. · 2020 · unknown

archive: archived pipeline: cataloged verified

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

Summary

This study investigates the cognitive mechanisms underlying left-turn gap acceptance decisions in human drivers, aiming to bridge the gap between naturalistic driving studies and laboratory-based cognitive science. While traditional gap acceptance models predict decision outcomes based on environmental factors, they offer little insight into the underlying cognitive processes. The authors hypothesize that noisy evidence accumulation, a mechanism well-established in abstract laboratory tasks, also governs complex driving decisions where perceptual information varies continuously over time. To test this hypothesis, sixteen participants performed a virtual driving task in a fixed-base simulator using the CARLA software. Participants navigated a virtual urban grid, encountering 120 left-turn trials where they had to decide whether to proceed ("go") or wait ("stay") as an oncoming vehicle approached. The experiment utilized a 3x3 factorial design, varying the initial distance (90, 120, or 150 m) and time-to-arrival (4, 5, or 6 s) of the oncoming vehicle. The researchers recorded decision outcomes and response times for "go" decisions, defined as the time between the appearance of the oncoming vehicle and the driver pressing the gas pedal. The results showed that the probability of accepting a gap increased with both distance and time-to-arrival. Crucially, response times increased with time-to-arrival but were unaffected by distance. To explain these findings, the authors developed a generalized drift-diffusion model. Unlike standard models with fixed parameters, this model featured a time-varying drift rate determined by the dynamic combination of distance and time-to-arrival, and collapsing decision boundaries that decreased as time-to-arrival shortened, reflecting increasing urgency. The model successfully captured individual and group-averaged decision probabilities and response time distributions. Cross-validation demonstrated that the model generalized to out-of-sample conditions, whereas simpler models with constant drift rates or boundaries failed to explain the response time data. The study concludes that dynamic evidence accumulation is an essential mechanism for left-turn gap acceptance. By modeling the decision process rather than just the outcome, this approach provides a more generalizable framework for understanding driver behavior. These findings have significant implications for the development of automated vehicles, offering a method to predict human driver behavior in real-time and simulate realistic human-like interactions in virtual environments. This work represents a methodologically stringent application of evidence accumulation models to driving tasks with continuously varying sensory evidence, advancing the theoretical understanding of how drivers process complex, time-sensitive information.

Key finding

A generalized drift-diffusion model with time-varying drift rates and collapsing boundaries accurately explains and predicts human drivers' left-turn gap acceptance decisions and response times.

Methodology

simulator

Sample size: 16

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 author_sweep_intake on 2026-05-27.

StageOutcomeToolModelPromptAttemptsCompleted
discover success author_sweep 2 2026-05-27
archive success canonical_url 5 2026-06-06
extract success cached 3 2026-06-10
clean success clean 1 2026-06-07
chunk success chunk 1 2026-06-07
embed success embed Qwen/Qwen3-Embedding-8B 1 2026-06-07
enrich skipped 4 2026-07-02
promote success 1 2026-06-04
summarize success llm qwen3.6-27b-prismaquant summ-v5 2 2026-06-10
tag success vector_similarity 15 2026-06-11
verify success 2 2026-06-10

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.

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).