Examining the relationship between perceived bikeability, cycling preferences and cycling distance

Understanding cycling behaviour is crucial for transport sustainability and individual health. Existing research has primarily focused on objectively measured bikeability using cross-sectional data, which are unable to assess within-person effects over time. Using longitudinal data from the Netherlands Mobility Panel, this paper applies a Cross-Lagged Panel Model and a RandomIntercept Cross-Lagged Panel Model to analyse the causal relationship between perceived bikeability, cycling preferences and cycling distance. The results show that individuals' cycling preferences and distance are relatively stable over time with positive autoregressive effects. Higher perceived bikeability and cycling distance directly encourage individual cycling distances and preferences separately in the subsequent year, with cross-lagged effects. For between-person effects, people with higher perceived bikeability generally have stronger cycling preferences and cycle longer distances. The Random-Intercept Cross-Lagged Panel Model provides a superior model fit to separate within-person from between-person effects. These findings are valuable for promoting cycling.