Effect of attribute perception factors on mode choice preferences among public transport users

Abstract

Users may form certain perceptions on the attributes of characteristics. This specific type of perception is referred to as attribute perception. Studies have shown that attribute perception influences travel decisions. Recently, in the field of transportation, there is an increased interest in exploring attribute perception and its impacts on users’ decisions on the choice of transport mode. Previous studies have identified a range of travel factors influencing passenger’s mode of choice, while there has been little effort on the exploration of the significance of these attribute perceptions. Hence, this study explored the impacts of attribute perception on the preferred choice of transport mode among public transport users in Dhaka, Bangladesh. To achieve this aim, a conceptual modelling framework was developed based on the Capability, Opportunity, and Motivation-Behavior (COM-B) model, which allows an extensive range of factors influencing mode choice to be considered. A self-administered survey was conducted to obtain users’ perceptions on public transport attributes. Cronbach’s alpha was used to identify reliability and internal consistency of the questionnaire and results showed that the questionnaire was found to have an adequate internal consistency. The Cronbach’s alpha ranged from .81 to .91 for the six attributes instrument. The data were extracted from 773 respondents from the two busiest routes in Dhaka city, namely Uttara to Azimpur and Mirpur to Motijheel. This conceptual relationship model was investigated using modelling approaches. Initially, the explanatory factor analysis (EFA) was used to determine the attribute perception factors’ underlying dimensions. Next, multiple linear regression analysis was used to investigate the relationship between attribute perception factors and travel behaviour. Finally, a multinomial logit model (MNL) was employed to explore the effect of attribute perception on the preferred choice of public transport users. The findings showed that the confidence level for regression model was 95% and its coefficient was defined as Sig < 0.05. Besides, the results showed that the attribute perception factors such as travel time, travel fare, waiting time, travel distance, comfort, and reliability had significant effects on passenger preference of travel modes among the different choices available. The results of this study have contributed to a better understanding of transport mode choice in Dhaka city and provided a foundation for future attribute perception-based mode choice research.