U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-10-024
Date: April 2010
Development of a Speeding-Related Crash Typology
PDF Version (1.53 MB)
PDF files can be viewed with the Acrobat® Reader®
INTERPRETATION OF RESULTS
The overall objective of this research was to obtain information on SR crashes that could be used for basic knowledge, treatment development (i.e., identifying the need for new treatments), and treatment targeting (i.e., targeting existing treatments to populations, times, or locations to maximize the effect). The methodologies used include examining single–variable tables and regression trees. As is the case with all such problem identification efforts, some caution must be observed in drawing conclusions for a number of reasons. As discussed previously, the definition of an SR crash is somewhat problematic in that at least two valid definitions exist, and both are based on the investigating officer's judgment of precrash speeds, which is a difficult judgment to make after the fact.
In addition, given either definition, the method for determining the highest priority categories varies. For treatment–related information, the variable descriptors in the single–variable tables that should be given the highest priority would be those that exhibit not only a high percentage of SR crashes, but also a high number of SR crashes. For example, the single–variable table below illustrates the number and percentage of SR crash categories by the type of location where the crash occurred. According to the percent columns, both FARS and GES indicate that the location category with the highest SR percentage is the interchange area. However, the crash location with the second highest percentage is "Not related to junction," which describes crashes occurring on roadway sections between intersections or interchanges. This category contains 10 times the number of interchange crashes in the GES data and 23 times the number of interchange crashes in the FARS data, implying that while attention should be given to speeding crashes in interchange areas, a higher priority should be given to strategies for reducing SR crashes at between–junction roadway sections.
Table 6. Frequency and number/percentage of SR crashes regarding crash location.
Thus, in determining high–priority categories for treatment consideration, both the SR percentage and the SR frequency should be considered.
The CART results must also be viewed carefully. While CART defines those subsets of crashes (i.e., subsets based on categories within multiple variables) with high proportions of SR crashes, even if one only goes to the third– or fourth–level branch, the total numbers of SR crashes and the proportion of total SR crashes is often less than 5 percent.
This report does not offer treatment–related conclusions. The goal of this report was to produce information that could be used by others in developing such conclusions. That information is noted appropriately throughout the report.