Research On Car-pooling Selection Based On Extended Planned Behavior Theory

The rapid increase in the number of private cars in China has led to increasingly serious problems such as urban traffic congestion and environmental pollution.In recent years,with the popularization of mobile Internet technology and the advent of the era of sharing economy,car-pooling,as a quasi-public traffic service model,has again started a wave.In order to study peoples’ choice behavior of car-sharing service in depth,a hybrid choice model which combines a multi-index multi-causal model(MIMIC)model and a discrete choice model constructed by introducing travelers latent psychological variables with the use of extended model of theory of planned behavior(TPB)as framework.Then,based on the data from the questionnaire on car-pooling,influences of object attributes and subjective attitudes on peoples’ choice of car-pooling were analyzed systematically and deeply.Firstly,literatures were reviewed for analyzing relevant characteristics of peoples’ choice of car-pooling,based on which,factors that affect this behavior,including variables of TBP and three other extended ones(perceived risks,perceived benefits and policy support),were sorted out and summarized.Therefore,a theoretical framework of peoples’ choice of car-pooling based on the TBP and its extensions was constructed,and research hypothesis was thus proposed.Secondly,questionnaires were designed,and Likert 5 scale was used to grade the observed variables that correspond to the latent variables.The car-pooling travel scenario was designed with the use of SP & RP approach.Then,final questionnaires were designed after including the personal attributes and was handed out and collected through Internet.In the end,466 questionnaires were confirmed to be useful after screening,with 2796 travel scenarios obtained.Thirdly,in order to further explore the internal relationship among latent psychological variables and their relationship with the objective attributes of travelers,a MIMIC model of car-pooling travel intention was constructed,and parameter estimation of the model was performed.The results showed that travelers with different demographic characteristics had significant influences on the latent psychological variables of carpooling behavior,which impacts peoples’ intention of travel traffic choices directly or indirectly.Finally,survey data were divided into two categories according to peoples’ experience of once having or not having car-pooling before,and a hybrid choice model of the latent variables as well as a binomial logistic model without considering the latent variables were constructed for comparative analysis.The results showed that the significant variables that affects travelers’ choice of car-pooling were different in accordance with that whether they have the experience of car-pooling or not.And the hybrid choice model,which could better explain the impacts and role of personal attributes and the latent psychological variables on travelers’ choice of car-pooling,was more accurate than the traditional binomial logistic model.According to the research results,it is possible to optimize the structure of travelling approaches by inducing peoples’ behavior of choosing car-pooling.