TITLE:
Causal Exploration of Bike Accidents in the Bay Area
AUTHORS:
Giovanni Bryden, Eric Catig, Wen Cheng
KEYWORDS:
Bicycles; Safety; Multinomial Logit Regression
JOURNAL NAME:
Open Journal of Safety Science and Technology,
Vol.2 No.3,
September
24,
2012
ABSTRACT: Although the proportion of bicycles to cars on the road remains low in the United States, it is on an upward trend. The high cost of gasoline and the re-introduction of bicycles as an attractive way to commute have contributed to an increase in ridership nationwide. However, as the number of bicycles on the road grows, so do the associated challenges. Without a clear understanding of how these two parties interact in a road network it is possible that we may stunt the growth of this sustainable and beneficial form of transportation and endanger riders through car-centered design practices and an unwillingness to accommodate for cyclists and their needs on the road. Five years of bicycle crash data from the city of San Francisco were analyzed. Using multinomial logistic regression, it is possible to relate different factors to the likelihood of an accident occurring, the corresponding severity, and the party at fault. Through this statistical analysis the study hopes to determine some of the statistically significant contributing factors to accidents involving cyclists in the city of San Francisco and make recommendations on how planners and design professionals can keep bicyclist safety considerations in mind. A few significant trends in the data were found, and recommendations are made to try and mitigate the risk of these factors on bicycle related accidents and their severity.