Cluster Modeling of Driving Behavior under Fatigue

Abstract

A road safety campaign, as one of the measures used to influence road user behavior to follow the driving regulations and to safeguard safety on the road network, may contribute to the reduction of the number of people killed on the roads. This study aimed to develop prediction models for the assessment of the impact on driving behavior of a fatigue road safety communication campaign based on the health belief model. This behavioral model took into account the behavior and the objectives of the campaign, as defined in the measurement variables of the evaluation, which were composed of self-reported data collected through a face-to-face questionnaire survey conducted before, during, and after the campaign implementation. Linear regression analyses were used to define the causal relationships between the dependent variables, intention and behavior, and other constructs. Results showed that the models developed for predicting drivers’ intention to stop and rest when they became tired and drivers’ behavior toward stopping and resting when they became tired varied significantly, depending on the characteristics of the sample tested (i.e., nonprofessional drivers versus professional drivers, younger versus older drivers, etc.). Also, it was observed that behavioral beliefs and risk comprehension were weak for the predictability of intentions and behavior; however, the inclusion of past behavior increased the predictability of the models predicting intentions, and the inclusion of past behavior and intentions (as observed also in similar studies) increased the predictability of the models predicting behavior.