Identifying driving behaviour profiles by using multiple correspondence analysis and cluster analysis

Many studies investigated the relationship between accident risk factors related to driving behaviour (e.g. distraction, aggressive driving) and the involvement in a road accident. However less attention has been given to aspects related to driver insecurity and the effects of the overall driving patterns. The main target of this research were twofold: 1) The identification of driving behaviour profiles taking into account also insecurity. 2) The analysis of the association between the identified profiles and their accident involvement. A survey was undertaken among a sample of Italian drivers to assess driving distraction, aggressiveness, indiscipline and insecurity. The items of the used questionnaire were mostly derived from the literature (e.g. Driver Behaviour Questionnaire). Behavioural tendencies of drivers to distraction, aggressiveness, indiscipline and insecurity were studied through Multiple Correspondence Analysis. By using Cluster Analysis seven groups of drivers with similar behaviours were identified. A significant association between the seven groups and road accident involvement was found. This statistical approach allows the identification of driving behaviour profiles that could be used for driving training purposes. The answers to the questionnaire can for instance highlight an aggressive and/or insecure driving behaviour thus tailoring the theoretical and practical driving exercises to the specific driver needs, especially for novice drivers.