Abstract

Aggressive driving is a common phenomenon with potential safety hazards. The primary aim of this study is to investigate the relationship among driving skill, driving behavior and driving aggressiveness. A questionnaire survey consisting of three subscales was conducted. Subscale 1 is about driving skill; Subscale 2 is associated with daily driving behavior, while Subscale 3 focuses on aggressive driving behaviors. A total of 239 licensed drivers completed the survey. The five-factor structure was extracted by factor analysis. Then, the correlation analysis method was used to explore the relationship between these factors and drivers’ demographic characteristics. K-means cluster was adopted to classify drivers’ driving aggressiveness into different levels using behavioral aggressiveness and emotional aggressiveness as the input features. Finally, an ordinal regression model for predicting drivers’ driving aggressiveness levels was established by driving skill factor, driving behavior factors and demographic information. The results indicated that more experienced drivers (i.e., drivers with older age, longer driving years or professional drivers) reported better driving behavior and less aggression. Male drivers were more aggressive than female drivers, especially on behavioral aggressiveness. In addition, drivers’ profession, driving years, penalty points, errors/lapses and violations can be used to predict their driving aggressiveness levels.

摘要

攻击性驾驶是一种普遍现象，存在安全隐患。本研究的主要目的是调查驾驶技能、驾驶行为和驾驶攻击性之间的关系。进行了由三个分量表组成的问卷调查。子量表 1 是关于驾驶技能；Subscale 2 与日常驾驶行为相关，而 Subscale 3 侧重于激进的驾驶行为。共有 239 名持照司机完成了调查。通过因子分析提取五因子结构。然后，采用相关分析方法探讨这些因素与驾驶员人口统计学特征之间的关系。采用K-means聚类，以行为攻击性和情绪攻击性为输入特征，将驾驶员的驾驶攻击性分为不同级别。最后，通过驾驶技能因素、驾驶行为因素和人口统计信息，建立了预测驾驶员驾驶攻击性水平的序数回归模型。结果表明，更有经验的司机（即年龄较大、驾驶年限较长或专业司机）报告了更好的驾驶行为和更少的攻击性。男性司机比女性司机更具攻击性，尤其是在行为攻击性方面。此外，驾驶员的职业、驾驶年限、罚分、错误/失误和违规行为可用于预测他们的驾驶攻击性水平。年龄较大、驾驶年限较长或专业司机）报告了更好的驾驶行为和更少的攻击性。男性司机比女性司机更具攻击性，尤其是在行为攻击性方面。此外，驾驶员的职业、驾驶年限、罚分、错误/失误和违规行为可用于预测他们的驾驶攻击性水平。年龄较大、驾驶年限较长或专业司机）报告了更好的驾驶行为和更少的攻击性。男性司机比女性司机更具攻击性，尤其是在行为攻击性方面。此外，驾驶员的职业、驾驶年限、罚分、错误/失误和违规行为可用于预测他们的驾驶攻击性水平。