Using Data to Make Roads Safer

My post for this week can be found in the May 8, 2024 issue of the Ahwatukee Foothills News (p. 33). Norma Hubele and I wrote “Using Data to Make Our Roads Safer” to highlight the importance of using traffic safety measures that are backed by rigorous statistical studies.

Two recent front-page stories in the Tempe Tribune highlight the importance of having good data. “Demand grows for traffic signal for Corona, Kyrene kids” headlined the April 14 issue and “Photo-enforcement likely coming to Tempe this year” was on the front page of the May 5 issue. Both articles dealt with community concerns about intersections where crashes had occurred, and discussed possible city responses to these concerns.

One aspect not addressed in the Tempe Tribune articles is how the city plans to evaluate the effectiveness of the measures. The other safety measures presented in the article were deemed to have been successful because there were fewer collisions in the time period after the measure was implemented than in the time period preceding the safety measure. These statistics are promising, but could be affected by regression to the mean (see my previous post on asphalt art for a discussion of regression to the mean in traffic safety studies). The photo enforcement effort presents an opportunity to do a rigorous evaluation. Since there are many more problematic intersections than opportunities to use cameras, the city could randomly select some of the problematic intersections for camera installation (one way to do this would be create pairs of intersections that had similar traffic and types of crashes, then randomly choose one from each pair to have a camera installed). This would give an objective unbiased estimate of the cameras’ effects for crash reduction.

Sharon LohrMay 13, 2024