Safety Evaluation of Profiled Thermoplastic Pavement Markings

Chapter 6. Before–After Evaluation Results

Aggregate Analysis

Table 10 details the Florida results, and table 11 details the South Carolina results. These results include the estimates of predicted crashes in the after period without treatment, the observed crashes in the after period, and the estimated CMF and its SE for all crash types considered. The results were consistent between the two States in that no CMF results were statistically significantly different from 1.0. Both States also indicated a modest reduction in total crashes and a reduction in nighttime wet-road crashes of approximately 10 percent, although these were not statistically significant at the 95-percent confidence level.

Table 12 provides the results for the combined Florida and South Carolina data for the crash types analyzed in both States. Even with the combined data, none of the estimated CMFs were statistically significant at the 95-percent confidence level.

Disaggregate Analysis

An attempt was made to further analyze the combined dataset for nighttime wet-road crashes to identify site characteristics for which the safety benefits might be greatest. Only nighttime wet-road crashes were considered because this was a key target crash type and the only one that showed some consistency and sizable effect for both States; however, the CMF estimates were still not statistically significant at the 95-percent confidence level.

The following variables were investigated:

• Number of lanes.
• Surface width.
• Average shoulder width.
• Median width.
• AADT.
• Expected nighttime wet-road crash frequency per mi prior to treatment.

The project team saw no differences or clear trends in the estimated CMF for any of the geometric variables or AADT. Therefore, for this dataset, the expected effect of this strategy on nighttime wet-road crashes was the same, regardless of differences in these aspects of the roadway environment.

There were some indications that the CMF for nighttime wet-road crashes might be smaller (a larger benefit) for sites with higher expected nighttime wet-road crash frequency per mi prior to treatment. However, the sample was too small for a robust conclusion in this regard.