The characteristics of scientific research are briefly discussed, its breadth and variety is illustrated in a table (which presents several examples by short title; notes the position of the work on a scale from pure to applied; notes whether it is descriptive or hypothesis testing; shows the intended use of the research), and comments on the use of traffic records. A court study in Denver is described, the validity of which was strongly dependent on the completeness and lack of bias of the records in matter of accidents and violations. In California the records system was used to identify problem drivers and to evaluate the effectiveness of warning letters. Further cases are quoted of the use of the system in the resolution of other problems. Achievement of valid inferences from valid data, and hazards in inferring causal relations are discussed. The more typical in traffic records studies are those related to regression effects, instability, history or external validity. Recommendations are made with regard to data collection. Two procedures to test the validity of data bases are: examine the feasibility and value of using an independent nonpolice agency to produce data for all accidents in a given locality; use national statistical and multidisciplinary accident investigation team data to provide a frame of reference against which to measure or examine the validity of data assembled in various jurisdictions.

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00097401
  • Record Type: Publication
  • Report/Paper Numbers: Proceeding
  • Files: TRIS, TRB
  • Created Date: Aug 13 1975 12:00AM