Accession Number : ADA134737

Title :   Statistical Modeling of Railroad Safety Performance.

Descriptive Note : Memorandum rept.,


Personal Author(s) : Bodt,B A ; Thomas,J

PDF Url : ADA134737

Report Date : Sep 1983

Pagination or Media Count : 36

Abstract : In recent years the Department of Transportation and the Association of American Railroads have become concerned over the safety performance of this nation's railroads. Recent statistics showed that the frequency of accidents was alarmingly high. Acting on this concern, the D.O.T. asked the Ballistic Research Laboratory to determine how railroads could be made safer. One approach to this problem would be to develop a mathematical model which would allow a railroad's safety to be expressed as a function of safety program characteristics. In this manner, guidelines for spending which could improve a railroad's safety record would be established. Several statistical techniques, including Cluster analysis, Discriminant Analysis, and Multiple Linear Regression, were employed in the development of this mathematical model. The model is a good predictor for the years 1976-1979, and should assist railroads in developing guidelines for spending which should improve their safety performance.

Descriptors :   *Railroads, *Safety, *Mathematical models, *Statistical analysis, Accidents, Multivariate analysis, Discriminate analysis, Cost analysis, Wounds and injuries, Validation, Regression analysis

Subject Categories : Surface Transportation and Equipment
      Safety Engineering

Distribution Statement : APPROVED FOR PUBLIC RELEASE