A reliability and usability study of TRACEr-RAV: the technique for the retrospective analysis of cognitive errors--for rail, Australian version

Appl Ergon. 2011 Nov;42(6):852-9. doi: 10.1016/j.apergo.2011.01.009. Epub 2011 Feb 26.

Abstract

The aim of this study was to compare the usability and reliability of two human error identification tools: TRACEr-Rail (developed by the Rail Safety and Standards Board in the UK) and TRACEr-RAV (an Australian specific version of the tool). Following an attempt to modify TRACEr-Rail to more appropriately suit the Australian rail context, it was predicted that TRACEr-RAV would be rated as more usable and be applied more consistently by Australian users than TRACEr-Rail. In Experiment 1, twenty-five rail employees used either TRACEr-Rail or TRACEr-RAV1 to extract and classify errors from six Australian rail incident reports. In Experiment 2, eleven university students used both TRACEr-Rail and TRACEr-RAV2 to extract and classify errors from three incident summaries. The results revealed that although modification of TRACEr-Rail to become TRACEr-RAV1 and TRACEr-RAV2 did not result in improved inter-rater reliability, modification resulted in improved ratings of usability in Experiment 2. Most participants in Experiment 2 preferred TRACEr-RAV2 to TRACEr-Rail. The poor inter-rater reliability observed was most likely the result of inadequate training, limited practice in using the tools, and insufficient human factors knowledge.

Publication types

  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Accidents / classification
  • Accidents / psychology*
  • Adult
  • Australia
  • Cognition*
  • Female
  • Humans
  • Male
  • Observer Variation
  • Railroads* / standards
  • Reproducibility of Results
  • Task Performance and Analysis
  • Workforce