Analysis and testing of Koornstra-type induced exposure models

Details

• Creators:
  Mengert, Peter
• Corporate Creators:
  John A. Volpe National Transportation Systems Center (U.S.)
• Subject/TRT Terms:
  Accident Data Accident Rates Computer Models
• Publication/ Report Number:
  FR-45-U-FAA-86-01
• Resource Type:
  Tech Report
• OCLC Number:
  173447379
• Corporate Publisher:
  United States. National Highway Traffic Safety Administration
• NTL Classification:
  NTL-SAFETY AND SECURITY-SAFETY AND SECURITY ; NTL-SAFETY AND SECURITY-Human Factors
• Abstract:
  Induced exposure models postulate a structure for accident data which permits the
  
  estimation of two factors: exposure and proneness. Since information on exposure
  
  is needed in order to assess the accident risk of different driver, vehicle, and
  
  environmental situations and since reliable exposure data is expensive to collect,
  
  induced exposure models hold the promise of a rich exposure data source which is
  
  perfectly matched to the accident data to be analyzed. This paper assesses the
  
  validity of the postulated structure of one induced exposure model: the Koornstra
  
  Model. The Koornstra Model was chosen for analysis and testing because it
  
  appeared to have the most potential for usefulness based on :
  
  a) previously reported favorable results (in limited testing);
  
  b) universal applicability;
  
  c) damaging criticism of certain other models; and
  
  d) a rich and well posed model structure.
  
  This paper analyses and tests the Koornstra Model from three different points of
  
  view:
  
  1) Is it based on reasonable assumptions?
  
  2) Does the model provide a significantly better fit to accident data than
  
  a simpler model which does not permit exposure or proneness to be
  
  estimated?
  
  3) How do the exposure estimates provided by the model compare with
  
  those from externally collected data?
  
  
• Format:
  PDF
• Alternate URL:
  http://ntl.bts.gov/lib/33000/33300/33361/33361.pdf
• Collection(s):
  Volpe Center
• Main Document Checksum:
  urn:sha256:57ac999053a9574226783f114e7129af45a34d1b8833efdc49a9d83951b8e8c4
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/5837/dot_5837_DS1.pdf
• File Type:
  [PDF - 1.91 MB ]