Exploring Factors Related to Drivers' Mental Model of and Trust in Advanced Driver Assistance System

Huang, Chunxi; Wang, Jiyao; Yan, Song; He, Dengbo · 2024 · IEEE Transactions on Human-Machine Systems

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 factors influencing drivers’ mental models of and trust in Advanced Driver Assistance Systems (ADAS), specifically focusing on emerging Level 2 functions beyond traditional adaptive cruise control (ACC) and lane centering control (LCC). The research addresses critical gaps in existing literature, which has largely overlooked newer tactical driving features (e.g., automated lane changing) and relied on proportional correctness scores that fail to distinguish between objective knowledge and subjective bias. Furthermore, previous studies often used regression models that ignored underlying associations among factors. Motivated by the rapid penetration of these systems in China and associated safety concerns, the authors aimed to provide insights for tailored driver education and interface design. The researchers employed a mixed-method approach using data from a survey of 287 Chinese drivers owning vehicles with emerging ADAS capabilities. To assess mental models, the study utilized Signal Detection Theory (SDT), specifically fuzzy SDT, to calculate sensitivity ($d'$) and response bias ($c$) across four categories: ACC functions, LCC functions, functions beyond ACC/LCC, and ADAS limitations. This method separated drivers’ objective understanding from their tendency to believe in the existence of functions. Trust was measured using a five-item Likert scale. To analyze the relationships among demographic, driving-related, and psychological variables, the authors applied an Additive Bayesian Network (ABN) to model multivariate dependencies without multicollinearity issues, followed by regression analysis to interpret the graph structure. The results indicated that different factors were associated with drivers’ objective knowledge versus their subjective bias. Crucially, the study found that drivers’ subjective bias was more strongly associated with their trust in ADAS than their objective knowledge ($d'$). The ABN model successfully identified structured relationships among the influential factors, revealing how dispositional factors (e.g., technology familiarity) and situational factors interact with mental models to shape trust. The analysis highlighted that while objective sensitivity varied across different ADAS function categories, the response bias played a dominant role in determining trust levels. The significance of this work lies in its novel application of SDT and ABN to human-machine interaction research. By distinguishing between objective mental models and subjective bias, the study provides a more nuanced understanding of how drivers perceive automation. The findings suggest that interventions aimed at calibrating trust should address subjective biases rather than solely focusing on improving factual knowledge. Additionally, the study offers a methodological case study for using ABN in observational data analysis and provides culturally specific insights for the Asian market, guiding the development of customized training programs and in-vehicle interfaces to mitigate over-reliance and improve safety.

Key finding

Drivers' subjective bias regarding ADAS functions and limitations was more strongly associated with their trust in the system than their objective knowledge of those functions.

Methodology

survey

Sample size: 287

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.

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.

• trust calibration
• situational awareness
• mental model of traffic
• acceptance adoption
• trust in automation foundations
• automation

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: computational model, theory or model