| Comments: | The authors make no mention of the traffic volumes on their sample of roads. They also state that highway segments were divided into 3 categories (divided, non-intersection, or intersection) while crashes were divided by type (pedestrian or vehicle) (p. 544). However, they don't specify how they dealt with these different categories. They also state that the "current" data are from 2002 and the "historical" data are from 2001 (p. 544). Table 3 includes 2003 data, and Table 4 excludes historical data. Since the zero-inflated regression models with the power prior use both current and historical data, the true sample sizes used for the models do not seem to be given in the paper. The method is a cross-sectional analysis, but employs Baysian techniques.Also, it should be noted that the degree of curvature in this paper assumes a 100 meter arc not a 100 foot arc as is customary in the US. |
[View score details]
