Attitudes to road safety and its implications for public policy

This paper provides an analysis of past survey data on community attitudes to safety, and investigates the following question: Can public policy impact community attitudes to safety? This is done by analysing the attitudes to mobile phone usage among a range of demographics (for example age, sex, and education) over a number of years and identifying high-risk groups. The conclusions should enable policy makers to understand how policy can better target high-risk groups. These high-risk groups are then recommended as a target for policy makers with a range of possible policy solutions. A unique approach of this paper is the use of two different modelling methods for data analysis. A multinomial regression model and an ensemble classifier based on regression trees are used to identify high-risk groups. Both approaches provide unique insights into the data. The comparison of the two models may be of particular interest to researchers, since it demonstrates their relative advantages in predictive capacity and in data handling.