MEASURING THE OUTCOMES OF DRIVER TRAINING: UNIVERSITY OF SOUTHERN CALIFORNIA DRIVER PERFORMANCE TEST

A new test for evaluating driver performance is described. The objective was to develop a test that is reliable, valid, and feasible for routine administration in the high schools. This test requires 30 min and is scored by a trained coder. The scoring is simplified to permit the coder to focus on observing and judging driver behavior. Standards for performance are learned by the coders during a 40-h training program. The test requires driver interaction with moderately heavy traffic and is intended to test the limits of driver performance. Intercoder agreement was about 80 percent, even though there were different seating positions. Some changes in scoring assignments and training are expected to improve the reliability of the test. This pilot study was carried out with 197 students at the end of their driver training course. A number of part scores and subtotals were used so that faults could be diagnosed. In general, these novice drivers did poorly in visual scanning. However, there was a wide range in driver proficiency. A hypothesis, the validation of which needs a large-sample study, is proposed to test the reason for the poor scanning. /Author/

Media Info

  • Media Type: Print
  • Features: Figures; References; Tables;
  • Pagination: pp 63-67
  • Monograph Title: Pedestrian controls, bicycle facilities, driver research, and system safety
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00173109
  • Record Type: Publication
  • ISBN: 0309026571
  • Files: TRIS, TRB
  • Created Date: May 18 1978 12:00AM