Using Fatal Crash Data to Analyze Nationwide Trends in Pedestrian Safety at Bus Stops

The multimodal nature of public transit requires bus stops to be safely accessible to passengers who begin and end their trips as pedestrians. Prior studies have analyzed bus stop safety using crash data, but many of these studies assume that all crashes within a given distance of a bus stop are directly related to the stop itself. This is likely due to the limited availability of data that specifically identifies transit bus stop-related crashes, which this study confirmed through an examination of transit-related information in safety various safety datasets. This thesis begins to address this gap by analyzing fatal "transit bus stop-related" pedestrian crashes reported in the Fatality Analysis Reporting System (FARS) and comparing them with other nearby fatal pedestrian crashes. The data was preprocessed to include only fatal transit bus stop-related crashes and other fatal pedestrian crashes within a quarter- and half-mile radii. Then, a binary logit model and hierarchical clustering compared fatal transit bus stop-related crashes with other nearby fatal pedestrian crashes. Last, an inventory of bus stop infrastructure was manually compiled to assess fatal transit bus stop-related crashes. The binary logit model indicated that midblock stops, pedestrians on the roadside, and bus-involved crashes are strong indicators of fatal transit bus stop-related crashes. The cluster analysis revealed three fatal transit bus stop-related crash scenarios: (1) crossing to/from an intersection stop, (2) waiting on the roadside at a midblock stop, and (3) crossing to/from a midblock stop. The clustering of both transit bus stop-related and other nearby crashes showed that fatal crashes involving a pedestrian waiting at a midblock stop clustered separately from other nearby fatal crashes. Finally, the inventory revealed that bus stop infrastructure generally improved after the crash, but disparities exist between crossing- and waiting-related crash locations. This thesis reveals how fatal transit bus stop-related crashes compare to other nearby fatal crashes, helping identify countermeasures specific to transit stops. Crossing improvements may benefit both transit passengers and other pedestrians in the immediate areas. Relocating transit stops to signalized intersections and converting in-lane stops to curbside pull-out stops may enhance bus stop safety.