DRIVERS’ PERCEPTION OF ACCIDENT RISK USING LATENT CLASS MODELS

Abstract

The human factor is considered one of the main causes of accidents. It is the driver of the vehicle who makes the decisions when driving. The task of driving a vehicle involves a set of sensory perceptions that must be in adequate conditions in order to minimise the risk of this task and avoid accidents. This work seeks to identify, through Latent Class models, how different types of drivers behave when faced with four situations/variables that constitute a risk of accidents while driving: driving speed, driving against the flow of traffic, overtaking a vehicle on a curve, and driving under the influence of alcohol or drugs. The results show that there are two classes of populations associated with risk perception in drivers, which can be categorised as Cautious and Risky. Individuals grouped in Class A would be women, with age groups between 18-35 years and over 46 years, and with a different occupation than the one asked in the survey or if unemployed. Class B would be men, aged 36 to 45, with occupations such as students and employees.