Accident Prediction Factors for Rural Highway Segments in Developing Countries

The focus of this research was to conduct a statistical analysis of the effect of selected predictor variables on total accidents on rural highway segments using data from sixteen different state highways in Karnataka State, India. The variables considered in the study were carriageway width, shoulder width, traffic volume, traffic composition, spot speed and pedestrian volume. Multi-regime models of Poisson and negative binomial structures were constructed for the prediction of total accidents for different ranges of pedestrian volume using data from eight state highways in five districts of Karnataka State. Final model selection was based on stringent criteria and the finally selected set of models for prediction of total accidents were cross validated by performing the Chi-squared test, z-test and mean absolute deviation test using data from eight other state highways in four other districts of Karnataka State. Based on the developed multi-regime models, accident prediction factors were developed to predict the total accidents in three years for different ranges of pedestrian volume under different conditions of roadway width, traffic volume and composition. The accident prediction factors developed in the present research can be advantageously applied for the prediction of accidents by the measurement of a single parameter – Speed; the product of the accident prediction factor and the average spot speed giving the expected number of accidents at the particular location.

• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Robert, Vivian R
  • Veeraragavan, Amrithalingam
• Conference:
  • Transportation Research Board 86th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2007-1-21 to 2007-1-25
• Date: 2007

Language

• English

Media Info

• Media Type: CD-ROM
• Features: References; Tables;
• Pagination: 18p
• Monograph Title: TRB 86th Annual Meeting Compendium of Papers CD-ROM

Subject/Index Terms

• TRT Terms: Crash causes; Crash rates; Highway safety; Mathematical prediction; Pedestrians; Regression analysis; Rural highways; Speed; Statistical analysis; Traffic volume; Width
• Geographic Terms: India
• Subject Areas: Data and Information Technology; Highways; Pedestrians and Bicyclists; Safety and Human Factors; I81: Accident Statistics;

Filing Info

• Accession Number: 01044874
• Record Type: Publication
• Report/Paper Numbers: 07-1325
• Files: BTRIS, TRIS, TRB
• Created Date: Mar 30 2007 6:59AM