Bicycle accommodations are an increasingly important component of the transportation system, and research has provided growing evidence that cities with higher bicycling (and walking) rates have better road safety records. In an effort to facilitate bicycle travel through intersections, newer traffic control devices have been applied, including the bicycle box, a space for bicycles to stop on a red signal ahead of the motor vehicle stop bar, and the two-stage turn box, a space where turning bicyclists can wait before making the second stage of a two-stage turn.
This study evaluated the effects of two bike boxes and two turn boxes installed in 2014 at an intersection in Charlottesville, Virginia. Videos collected during 3 days before the changes (non-consecutive over a 1-month period) and 5 days after the changes (non-consecutive in the fall and spring seasons) provided volume counts and tallies of traffic infractions and conflict events such as near misses. Data were prepared in order to pair the “before” and “after” periods, resulting in eight 12-hour sets of observations starting at 7:30 a.m., each with 48 time intervals of 15 minutes. Because the data were not normally distributed, the Wilcoxon signed-rank test was employed to compare the before and after periods. To take advantage of the paired structure of the data (i.e., before and after), a matched-pair or related-sample version ofthe test was performed.
After the main analysis, a subset of data (1 hour in the morning and 1 hour in the afternoon for three before and three after count dates) was re-reviewed by one researcher in order to address concerns about inter-rater reliability from the initial data reduction. Several methods were used to compare this re-reviewed dataset to the original review results.
Results were mixed. Among other findings, the following results were statistically and practically significant:
The two bike boxes were used properly/improperly by 46%/40% and 24%/10% of approaching bicyclists on the respective leg of the intersection.
The two turn boxes had high levels of improper (but not necessarily unsafe) use, at 57% to 100% of approaching bicyclists.
Uncategorized bicyclist traffic infractions on one approach decreased by 43% after the changes but increased by 80% on another approach.
Prohibited direct left turns increased 200% for motorists (from 0.1% to 0.4% of approaching motorists)and 290% for bicyclists (from 13.3% to 51.3% of approaching bicyclists).
The study recommends that the Virginia Department of Transportation (1) create or improve education materials related to bike boxes and turn boxes and (2) evaluate the feasibility of submitting requests for interim approval for bicycle boxes and two-stage bicycle turn boxes.