Modelling the influence of peers' attitudes on choice behaviour : Theory and empirical application on electric vehicle preferences

While the importance of social influence on transport related choices is commonly acknowledged within the travel behaviour research community, there remain several challenges in modelling the social influence in practice. Indeed, transport demand models have traditionally not considered the effects of social influence on individual choice behaviour. The present study contributes to filling this gap by providing a novel approach to model indirect social influences by combining information on attitudes and social proximity which characterises the individual’s social network. In particular, we develop an analytical framework to account for indirect social influence related to the processes of social conformity generated by peers’ attitudes. Within such a framework, a newly developed social influence variable is used to explain how peers indirectly influence the individual’s choice. We apply this approach to the empirical context of electric vehicle adoption intentions, using data from a stated preference survey. Besides stated choice experiment data on vehicle preferences, the dataset provides extensive information on social relationships amongst the respondents and on sociological constructs such as lifestyle practices, lifestyle liminality and the New Ecological Paradigm. We use a factor analysis to identify these latent constructs from the psychometric indicators available from the survey data. Having identified the subset of indicators specific to each latent construct, we use a cluster analysis on such indicators to group respondents with similar indicator levels for specific factors. Subsequently, we specify the social influence variable by interacting clusters with social relationship measures. In order to understand whether individuals with a specific attitude can influence peers in their social network, the individual’s peer attitude is included in different components of a hybrid choice model. Our results show that the inclusion of this variable indirectly affects the decision making process of the individual as the peers’ attitudes are significantly related to the latent attitude of the individual. On the other hand, it does not seem to directly affect the utility of an alternative as a source of systematic heterogeneity nor does it work as a manifestation of the latent variable, i.e. as an indicator