Associations between the objective and perceived built environment and bicycling for transportation

This paper investigates the relative associations between the objectively-measured built environment versus stated perceptions of the built environment and bicycling. Data are from a random phone survey conducted in the Portland, Oregon, region. Binary logit and linear regression models, using objective measures, perceived measures, and both sets of measures, were estimated to predict propensity of bicycling and frequency of bicycling separately. Results showed that the perceived environment and objective environment had different associations with bicycling. This suggests that future research should include both measures when possible and that a supportive objective environment is necessary for bicycling. Intervention programs to improve people's perceptions of the environment may be also necessary to reap the full potential of planning and design policies. The results also suggest that it is useful to predict bicycling propensity and bicycling frequency separately, as the predictors of each behavior do vary. For the objective measures, bicycle paths and lanes and minor streets were associated with higher propensity of bicycling for transportation, while paths and nearby retail destinations were associated with higher bicycling frequency. Perceptions of the environment did not predict the propensity to bicycle, after controlling for attitudes towards daily travel, but did predict bicycling frequency. Finally, the analysis confirms the importance of attitudes in predicting behavior.