A MODEL FOR THE NUMBER OF ROAD ACCIDENTS AT RANDOMLY SELECTED LOCATIONS. AN ANALYSIS OF THE EFFECTS OF COUNTERMEASURES IMPLEMENTED IN STUDIES WITH TEST GROUPS AND CONTROL GROUPS

MODELL FOER ANTAL TRAFIKOLYCKOR PAA SLUMPVIS UTVALDA PLATSER. ANALYS AV AATGAERDSEFFEKTER VID STUDIER MED FOERSOEKS- OCH KONTROLLGRUPPER

This report deals with models describing road accidents at randomly selected locations. On the assumption that the number of accidents at a given location has a Poisson distribution, it is shown that the number of accidents at a randomly selected location does not follow a Poisson distribution (provided that the average number of accidents varies between the locations). An accident distribution is obtained which has an "over-dispersion", i.e. the variance of the number of accidents is greater than the average number of accidents. It is found that the variation between the average accident levels can be described with a gamma distribution. The observed numbers of accidents from both Sweden and the UK correspond very closely to the theoretical model. The model is used to estimate the effect on the number of accidents of a countermeasure, which is studied by means of an experiment with a test group and control group of locations. The effect of the countermeasure is estimated with the aid of the maximum likelihood method. Using the ml-estimator, two different methods are applied to obtain approximate confidence intervals for the effect of the countermeasure. Finally, an attempt is made to test the effect of the countermeasure when the size of the test material and the control material are limited. (Author/TRRL)

• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/issn/03476049
• Corporate Authors:

  Swedish National Road and Transport Research Institute (VTI)

  Linköping,   Sweden  SE-581 95
• Authors:
  • Danielsson, S
• Publication Date: 1983

Language

• Swedish

Media Info

• Features: References; Tables;
• Pagination: 24 p.
• Serial:
  • VTI meddelande
  • Issue Number: 355
  • Publisher: Swedish National Road and Transport Research Institute (VTI)
  • ISSN: 0347-6049

Subject/Index Terms

• TRT Terms: Confidence intervals; Correlation analysis; Countermeasures; Crash rates; Crashes; Distributions (Statistics); Forecasting; Location; Mathematical models; Measures of effectiveness; Poisson distributions; Prevention; Random sampling; Safety; Statistics; Variance
• Uncontrolled Terms: Effectiveness
• Old TRIS Terms: Confidence level; Poissons distribution
• ITRD Terms: 1643: Accident; 1661: Accident prevention; 1612: Accident rate; 6527: Correlation (math, stat); 6572: Distribution (stat); 132: Forecast; 6473: Mathematical model; 6555: Statistics
• Subject Areas: Data and Information Technology; Highways; Safety and Human Factors; Security and Emergencies; I81: Accident Statistics;

Filing Info

• Accession Number: 00380019
• Record Type: Publication
• Source Agency: Swedish National Road and Transport Research Institute (VTI)
• Report/Paper Numbers: Monograph
• Files: ITRD, TRIS
• Created Date: Jan 30 1984 12:00AM