Previous research has shown that the determinants of perceived and observed aggressive driving behavior may differ. However, the consideration of major sources of aggressive patterns may introduce additional variations in the effect of such determinants. This study aims to provide further insights in the variations of these two behavioral components arising from driver’s fatigue, gender as well as internal and external distractions (such as, rushing to destination, listening to music and solving logical problems) during the driving task. To identify how the factors determining perceived and observed aggressive behavior may vary across groups of drivers associated with such sources of aggressive driving, survey and simulation data are statistically analyzed. Separate models of perceived and observed aggressive driving behavior are estimated for fatigued and non-fatigued, distracted and non-distracted, male and female drivers. To address various aspects of unobserved heterogeneity, associated with the unobserved variations that are commonly shared among the behavioral components and participants, as well as their unobserved interactions, the correlated grouped random parameters bivariate probit modeling framework is employed. The results of the empirical analysis showed that the effect of the socio-demographic and behavioral factors on perceived and aggressive driving behavior may vary across the aforementioned groups of drivers, in terms of magnitude and directional effect. In addition, the identification of correlation among the unobserved characteristics further illustrates the complexities of the driving decision mechanism, especially when fundamental sources of aggressive driving are evident.

先前的研究表明，感知和观察到的侵略性驾驶行为的决定因素可能有所不同。但是，考虑到攻击性模式的主要来源可能会导致此类决定因素的效果产生其他变化。这项研究旨在为驾驶员在驾驶过程中疲劳，性别以及内外分心（例如急于到达目的地，听音乐和解决逻辑问题）引起的这两种行为成分的变化提供进一步的见解。为了确定确定感知和观察到的攻击行为的因素如何在与此类攻击性驾驶源相关联的驾驶员组之间变化，对统计和模拟数据进行统计分析。针对疲劳和非疲劳，分心和不分心的男性和女性驾驶员，估计感知和观察到的攻击性驾驶行为的单独模型。为了解决与行为组件和参与者之间通常共享的未观察到的变异以及未观察到的交互相关的未观察到的异质性的各个方面，采用了相关的分组随机参数双变量概率模型框架。实证分析的结果表明，社会人口统计学和行为因素对感知驾驶行为和激进驾驶行为的影响，在上述驾驶者群体中，在幅度和方向效应方面可能会有所不同。此外，