USE OF MIXED LOGIT MODELS IN THE PERCEPTION OF ACCIDENT RISK

Abstract

Traffic fatalities are high in ages between 15 and 29 years, where the human factor causes 70% and 80% of accidents. According to the above, the institutions in charge of preventing traffic fatalities must have the tools to make decisions that promote safer driving habits and thus reduce the number of accidents. The research goal is to identify how drivers behave at risk of traffic fatalities using the mixed logit models. The variables analyzed were: a) speeding, b) driving in the wrong direction, c) overtaking on curves, and d) being under the influence of alcohol. The results show that speed can vary according to driving maneuvers.