A key challenge facing policymakers seeking to promote bicycling is the lack of data regarding the use of current bicycle facilities. As policy makers and public employees seek to provide attractive facilities for bicycling, they are highly limited in their ability to evaluate usage levels of existing facilities or predict the usage of proposed ones. Portland, Oregon is at the forefront of bicycle facilities innovation in the United States and the Portland Office of Transportation (PDOT) has been monitoring the use of its facilities through a bicycle count program that dates back to1991. PDOT makes use of their bicycle count data to measure changes in bicycling citywide and in different parts of town. It also uses the data collected by the manual counters to understand difficult bicycle movements that may require further bicycle treatments. This research explores additional uses of PDOT's bicycle count data, using statistical analysis to examine how the attributes of different facilities impact the variation in levels of cycling across the different count locations. This paper demonstrates the power of using a linear regression model as a tool for municipalities to evaluate levels of usage on their facilities. It also makes recommendations on how PDOT could modify the way it collects and stores its dataso as to be more useful and ultimately more explanatory in similar types of statistical analysis. A general model of how a municipality might set up its own bicycle count program is also suggested.
Samenvatting