BEFORE AND AFTER ACCIDENT RESEARCH AND "FREE" TESTS

VORHER-NACHHER-UNFALLUNTERSUCHUNGEN UND "FREIE" TESTS

ON THE BASIS OF FREE TESTS IT IS POSSIBLE TO ARRIVE AT A STATISTICALLY VERIFIABLE DECISION WHETHER THE INCREASE OR DECREASE IN THE NUMBER OF ACCIDENTS DURING THE YEAR FOLLOWING THE INTRODUCTION OF A MEASURE CONSIDERED SIGNIFICANT FOR TRAFFIC SAFETY IS COINCIDENTAL OR NOT. WITHIN THE FRAMEWORK OF THE REQUIRED ACCURACY THE PROBABILITY TEST BASED ON THE NORMAL DISTRIBUTION IS UNIVERSALLY APPLICABLE. THIS HAS THE ADVANTAGE THAT THE X - VALUE VALID FOR (X) = 0,95 (NAMELY X = 1,64485) IS APPLICABLE IN EACH CASE. THIS MEANS THAT IN CONTRAST TO THE POISSON TEST NUMERICAL TABLES ARE NOT REQUIRED. IN ADDITION TO THE STATISTICAL 95% CERTAINTY INTERVAL, THE INTERVAL EXTENDING DOWN TO 90% IS OF SIGNIFICANCE IN ANTICIPATING POSSIBLE CONFIRMATION OF REJECTION (DECREASE OF ACCIDENTS) OR ACCEPTANCE (INCREASE OF ACCIDENTS) OF THE HYPOTHESIS H(0), WHEN IN THE FUTURE ADDITIONAL "AFTER" YEARS OF OBSERVATIONS ARE TO BE INCLUDED IN THE ANALYSIS. IN THE CASE OF MORE THAN ONE "AFTER" OBSERVATION YEAR HOWEVER IT IS SUGGESTED THE MORE COMPREHENSIVE SIGNIFICANCE TESTS BE USED THAN THESE "FREE" TESTS.

  • Corporate Authors:

    Transpress VEB Verlag fuer Verkehrswesen

    Franzoesische Strasse 13/14, Postfach 1235
    108 Berlin,   Germany 
  • Authors:
    • HARPE, K
  • Publication Date: 1977

Language

  • German

Media Info

  • Pagination: 56-9
  • Serial:
    • Strasse
    • Volume: 17
    • Issue Number: 2

Subject/Index Terms

Filing Info

  • Accession Number: 01293052
  • Record Type: Publication
  • Source Agency: Forschungsgesellschaft für Straßen- und Verkehrswesen (FGSV)
  • Files: ITRD
  • Created Date: Nov 21 2010 3:50AM