The major goal of this research is to help create more walking- and biking-friendly policies by finding factors that influence walking and biking to transit stations. In 2005, a station user survey was conducted in Mountain View, California. Based on the survey result, this paper developed mode choice models of commuter rail user's access trips to the station. To find statistically significant variables that influence the probability of choosing walking and biking over driving, a pair of binominal logit analyses were performed with 40 travel, socio-economic, and built environment variables. The research tested 277 walkers and auto travelers living within 1.5 miles of the station and 280 bikers and auto travelers living within 2 miles of the station. This research found six predictors for the walking vs. driving model: trip distance, trip purpose, car availability, race, intersection density, and proximity to driving-friendly streets, and found five predictors for the biking vs. driving model: trip distance, trip purpose, car availability, race, gender, and proximity to driving-friendly streets. A similar set of variables entered both models, but the explanatory power of the bike model was much lower, suggesting that the factors influencing biking may be different from the traditional variables used by existing research. The influence of the built environment, which is probably more policy-relevant than travel and socio-economic variables, seems relatively low for both models, suggesting that future research needs to look at finer-grained micro-level pedestrian and biker environment.
Abstract