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-RD-98-133
Date: October 1998
Accident Models for Two-Lane Rural Roads: Segment and Intersections
3. Data Collection
Limitations on Data Quality
Minnesota and Washington State data were constrained to lie on rural two-lane roads with segment length 0.1 miles or longer with both segments and intersections having reasonable bounds on ADT. Other reasonable constraints were also imposed, including relatively complete and consistent data for the time periods of interest. Many observations from the original populations were lost when these constraints were imposed, but good-sized samples remained. The Washington intersection samples, "opportunity" samples, were smaller than the other samples and it is not known how representative they are of the population of Washington State intersections.
Data collected include: accident counts, exposure and ADT, lane and shoulder widths, Roadside Hazard Rating, number of driveways, horizontal and vertical alignments, commercial traffic percentage, weather (in Minnesota), intersection angles and channelization, and speed limits. These data are often estimates based on averages and are subject to some uncertainties in location and time. ADTs are based on observations at selected sites, interpolation, and/or extrapolation, and are particularly crude estimates in the case of intersections. In view of the importance of ADT in the modeling, the crudity of these estimates should serve as a caution.
Driver and vehicle characteristics were not collected, nor were such design variables as sight distances and minor road alignments.
Despite shortcomings in quality and completeness, the data obtained provide a relatively diverse and comprehensive basis for analysis and modeling.