DEVELOPMENT OF PREDICTIVE MODELS TO IDENTIFY PERSONS AT HIGH RISK OF ALCOHOL RELATED CRACH INVOLVEMENT

The basic approach employed on this effort was to first identify from the literature and existing data sources several groups likely to be at high risk of alcohol related (A/R) crashes and then for each group to separately develop a predictive model to identify those smaller subgroups of individuals at an even further heightened risk of A/R crash involvement. The high risk groups which were selected for modeling were two age groups of young males (16-20 and 21-24), persons previously convicted of driving under the influence of alcohol, persons with three or more moving traffic violations, persons recently divorced and persons recently released from prison. Additionally, a model was constructed on a one-tenth sample of the genral driving population. The objective of the statistical analysis was the developement of a predictor model for A/R crashes using as independent variables those data elements which are most strongly associated with A/R crashes. The true value of models developed in this application cannot be assessed until the results of the prospective validity tests are in. The results of this study indicate that it may indeed be possible to develop better predictors of A/R crash involved drivers utilizing information available to real-world program administrators. The crash risk of drivers so identified is much greater than that of the normal driver, and, while there remains a large proportion of false positives in the high subgroups, a promising potential use for this type of predictive modoling approach is to use the models for targeting high risk subpopulations for more intensive countermeasure activities. /Author/

• Supplemental Notes:
  • Presented at the 7th International Conference an Alcohol, Drugs and Traffic Safety, Melbourne, Australia, January 23-28, 1977.
• Corporate Authors:

  University of North Carolina, Chapel Hill

  Highway Safety Research Center
  Chapel Hill, NC  United States  27599
• Authors:
  • Lacey, J H
  • Stewart, J R
  • Council, F M
• Publication Date: 1977-1

Media Info

• Features: References; Tables;
• Pagination: 29 p.

Subject/Index Terms

• TRT Terms: Adolescents; Driver records; Drivers; Drunk driving; Forecasting; Mathematical models; Statistical analysis; Teenage drivers; Traffic crashes
• Uncontrolled Terms: Models
• Subject Areas: Data and Information Technology; Highways; Safety and Human Factors;

Filing Info

• Accession Number: 00168044
• Record Type: Publication
• Source Agency: National Safety Council Safety Research Info Serv
• Files: TRIS
• Created Date: Mar 29 1978 12:00AM