Investigation of Interaction between Age and Gender Effects of Car Users by using Log-Linear Model: A Bayesian Inference Approach

Abstract

The growth in car ownership, demand for personal mobility and traffic congestion all generate negative impacts on the environment. In addition to adverse impacts, car use contributes to urban sprawl, unhealthy micro environments and increased accidents. In western countries, chronic traffic congestion due to increasing travel demand resulting from economic growth is placing governments under pressure to promote and incentivise mode shift to non-motorised modes. Therefore, this paper aims to investigate the relationships of car users' perceptions and attitudes with age-gender factors to investigate respondents' willingness to reduce car use for the sake of the environment. A log linear model for categorical data was developed using Bayesian inference approach. Models were constructed separately with and without interaction between age and gender effect. Attitudinal data from British Social Attitudes (BSA) surveys from 2011 to 2014 were used in this study. The application of the Bayesian inference approach to the BSA data demonstrated useful properties shared by population groups. The results indicate that significant differences exist among different age and gender groups. The knowledge is useful to policy makers and transport managers in designing targeted solutions to specific population groups. The outcome of this research will be vital from a policy-making perspective, as appropriate clusters of the population can be targeted when implementing sustainability policies.