Increasing the number of bicycle commuters

The present study confirmed that commuter cyclists take on different mode-choice behaviours depending on whether they have experienced leisure cycling prior to becoming a commuter cyclist. In a survey conducted for the present study, about 57% of the sampled commuter cyclists had experience as a leisure cyclist. The remaining individuals were inexperienced at leisure cycling. A discrete mode-choice model, which has been widely used in mode-choice studies, was applied separately for each commuter-cyclist group, based on the stated preference data with the conventional explanatory variables (i.e. individual- and alternative-specific variables). The results showed that the model for inexperienced commuter cyclists exhibited a higher goodness-of-fit compared with the experienced cyclists. In order to address the attitudinal differences between the two groups, a binary logit model was applied again to examine the factors that may cause the difference. The extra variables associated with the previous commuting mode and the current conditions of cycling to work were put into the binary logit model. In particular, seven latent factors were derived from 28 questions by using principal component analysis. Several variables, such as distance to work, previous commuting mode, sensitivity to the environment, and symbolic motives for driving a car, were statistically significant and decisive in accounting for the differences between experienced and inexperienced commuter cyclists. Therefore, it is recommended that targeting experienced cyclists would be the optimal approach to promote the use of bicycles and reduce the use of motorised vehicles. The results of the present study could also provide policy makers with valuable background information about commuter cyclists.