Unveiling barriers and motivations to cycling across different stages of change

Despite its many benefits, cycling remains a relatively minor mode of transport in many North American cities. Hence, it is vital to investigate which perceived barriers prevent people from cycling and which aspects motivate them to cycle for transport. This study used a “stages of change” framework to identify individuals' cycling stages of change regarding their cycling behavior and future intention to cycle for transport. Then, the barriers and motivators for people at different stages were investigated. A survey of 1200 respondents was designed and administered in two Canadian cities. An interpretable ensemble learning approach, which combines multiple models to improve predictive performance, was used for modeling. The results suggest that traffic danger and the ease of getting around by other modes are the two main perceived barriers to cycling. Improving health and fitness, avoiding the hassle of parking cars, and helping to reduce carbon emissions were the top motivators for the population interested in cycling. Different barriers and motivators have different influences on people based on their cycling stage of change. For example, those not considering cycling (pre-decisional denial) do not mention traffic danger as a problem, but for those who want to cycle but feel that it is impossible (pre-decisional inhibition), traffic danger is the key problem preventing them. For the majority of those keen to start cycling (pre-actional), no problem is seen as preventing them. The results give nuance to problems and what prevents people, along with motivations, at the different stages of behavior change towards cycling. • This study investigates cycling barriers and motivators by cycling stages of change. • Distinct motivators and barriers are evident across the stages of change. • Improving health, fitness, and avoiding parking hassle are top cycling motivators. • Traffic danger and ease of other travel modes are the top barriers to cycling. • An interpretable ensemble learning approach is used for modeling.