Self-Perception Versus Objective Driving Behavior: Subject Study of Lateral Vehicle Guidance

Haselberger, Johann; Schick, Bernhard; Mueller, Steffen · 2024 · arXiv

DOI: 10.48550/arXiv.2402.13104

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 the discrepancy between drivers’ self-perceived driving styles and their objective lateral driving behavior, aiming to inform the development of personalized autonomous vehicle systems. The authors argue that current automated driving functions use fixed parameters that may not align with individual user preferences, potentially hindering acceptance. Specifically, the research focuses on lateral vehicle guidance on rural roads, a domain often overlooked in favor of longitudinal highway driving. The study seeks to identify objective indicators of lateral driving style and evaluate whether subjective self-reports can reliably predict these behaviors. The researchers conducted a controlled subject study with 62 German drivers using an instrumented research vehicle (JUPITER platform) equipped with multimodal sensors, including LIDAR and cameras. Participants completed an online questionnaire prior to the experiment, which included the newly translated German version of the Multidimensional Driving Style Inventory (MDSI-DE). The driving experiment involved a 71.9 km route comprising city, highway, rural, and federal road sections. The authors introduced novel objective indicators to assess lateral behavior, including a "G-G Envelope" for acceleration profiles, a "Curve Cutting Gradient" for stationary cornering, and a trajectory classification system for transient cornering behavior. These metrics were derived from raw sensor data, filtering out lane changes and oncoming traffic to isolate natural driving habits. The results revealed that lateral driving behavior is highly heterogeneous among drivers. Correlation analysis between MDSI factor scores and the proposed objective indicators showed only modest but significant associations, primarily linked to acceleration and jerk statistics. The study found that self-reports were not strong predictors of specific lateral maneuvers, such as curve-cutting intensity or trajectory class. Gender and age differences were observed in self-assessments, with women scoring higher on careful and anxious styles and men on risky and angry styles, but these subjective trends did not strongly map to the complex lateral indicators. The authors also noted that previous studies often relied on simple statistics or simulators, whereas this work provided a more nuanced, real-world analysis of lateral dynamics. The significance of this work lies in its contribution to the field of human factors in autonomous driving. By demonstrating the limitations of self-reports in predicting objective lateral behavior, the study highlights the need for direct measurement or adaptive learning algorithms to personalize automated driving styles. The authors provide a publicly available dataset containing anonymized socio-demographics, questionnaire responses, raw vehicle measurements, and derived indicators, facilitating further research. The findings suggest that integrating individual driving preferences into automated systems requires more sophisticated methods than simple self-reporting, particularly for lateral guidance, to enhance passenger comfort and trust.

Key finding

MDSI driving-style factor scores show modest but statistically significant partial correlations with objective lateral driving indicators, predominantly acceleration- and jerk-based measures (e.g., the careful and anxious factors correlate negatively with lateral acceleration and longitudinal jerk; distress-reduction correlates positively with late-counter trajectory class). Drivers tend to position roughly 0.15 m right of lane center when cornering rather than driving lane-center, and male drivers show significantly higher early-cutting and early-counter trajectory percentages and intensities than female drivers.

Methodology

lab_experiment

Sample size: N=62 (German drivers, instrumented-vehicle rural-road study)

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 arxiv on 2026-05-07 (3 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.

• lane positioning
• steering pattern
• sex gender
• cultural cross national
• decision making risk perception
• dbq psychometrics

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: behavioral performance data
• Methodological Resource: validation psychometrics
• Theoretical Contribution: computational model