Pattern Identification from Older Bicyclist Fatal Crashes

Abstract

Estimates from the U.S. Census Bureau indicate that the elderly (age 65 and older) people represented 12% of the total population in 2005. Bicycling is becoming popular among people of all groups. In 2016, 130 elderly bicyclists were killed (20% higher than 2014) on the U.S. roadways. The sharp rise of elderly bicyclist fatal crashes calls for a rigorous study to determine the key associated factors in elderly bicyclist crashes. Graphical methods, such as joint correspondence analysis (JCA), are useful in identifying the association patterns from a complex data set with multiple variables by producing a proximity map of the variable categories in a low dimensional plane. This study used 3 years (2014 to 2016) of data on elderly bicyclist fatal crashes from the Fatality Analysis Reporting System (FARS) in the U.S. to determine the key associations between the contributing factors by using JCA. Some of the key findings include bicyclist fatal crashes on roadways with high posted speed being very random; higher crash occurrences on roadways with bicycle lane/shoulder/parking lane under dark conditions with no lighting, on two-way undivided roadways with bicyclists on the travel lane, and at signalized intersections (pedestrian/bicycle signal presence is unknown) with “motorists fail to yield” related crashes. The findings from the current study can help in refining the policies and safe design practices that explicitly recognize this issue and will better serve a growing segment of the nation’s population.