Statistical Modeling of Grievous Road Traffic Accidents

Certain accident impact factors are often hidden in statistical models due to over-dispersion of traffic accident data. Given the limitations of the traditional statistical model to solve over-dispersion data, a random-parameters negative binomial (NB) regression model was applied to capture the heterogeneity of data distribution so as to identify factors that explain grievous road traffic accidents. Using 7-year data from grievous accidents in China, this paper carried out a simulated calculation of the proposed model. The empirical results show that compared to a fixed-parameters NB regression model, random-parameters models are quite enough to characterize the uncertainty of the accident and can greatly improve the validity of the empirical results.


  • English

Media Info

  • Media Type: Web
  • Features: Figures; References; Tables;
  • Pagination: pp 2249-2256
  • Monograph Title: CICTP 2012: Multimodal Transportation Systems—Convenient, Safe, Cost-Effective, Efficient

Subject/Index Terms

Filing Info

  • Accession Number: 01505226
  • Record Type: Publication
  • ISBN: 9780784412442
  • Files: TRIS, ASCE
  • Created Date: Jan 27 2014 11:02AM