Performance of safety evaluation methods

The main objective of this study was to investigate the performance of safety evaluation methods and estimate the safety effectiveness of street bumps, closing of median openings, and divisional islands in an urban environment. Traffic accident data were obtained from the municipality of Irbid, Jordan, and its police records. In the analysis the performances of simple before-and-after and empirical Bayesian methods were compared with results obtained by comparison group methods. The results indicate that the simple before-and-after method overestimated the effectiveness of safety improvements and led to erroneous conclusions at specific locations, as well as at the aggregate level. The comparison group method and the empirical Bayesian method provided results comparable to each other. Based on the results of the comparison group and the Bayesian methods, the study indicates that street bumps and closing of median openings at intersections are effective in reducing total accidents. In contrast, implementation of divisional islands did not provide a significant reduction.