Adolescents on the Road : A Case Study of Determinants of Risky Behaviors

The 2016 report of the European Transport Safety Council claims that EU safety progress has come to a standstill. This study aims at deepening the knowledge of factors that influence adolescents' risky behavior on the road. Bayesian Networks offer a promising new way to looking at the issue. In the analysis of a dataset collected in Tuscany, Italy, called EDIT, we found evidence that the use of alcohol and illegal substances explain only part of the probability of having an accident, and that other observable variables, like the level of distress or the type of school attended are significantly related to the probability of incurring in a road crash. New and close attention should be given to a systemic approach and to a plethora of environmental and individual variables that may rise the probability of road accidents for very young drivers