What You See is What You Get? Correspondence of Video and Interview Data on Secondary Task Engagement While Driving-A Naturalistic Driving Study

Abstract

Numerous studies use questionnaires or interviews to investigate the prevalence of secondary task engagement while driving. This data may be subject to memory distortion. This study aims at investigating the extent to which interviews are valid instruments to assess secondary tasks. Therefore, we evaluated interviews and video data allowing the observation of secondary task engagement from a Naturalistic Driving Study. We equipped the vehicles of 94 subjects with cameras filming the driver's vehicle cabin. Video and interview data were collected twice within the study period of 3 days. We then determined hit rate, misses, false alarms, correct rejections, sensitivity, as well as specificity for 15 secondary tasks. We found 594 secondary tasks in the videos. In 10% of all comparisons (Nall=2.187 for 15 tasks) the interview correctly identified task engagement (hit). In 17% of the comparisons drivers missed to report a task. In 9% of the comparisons there was a false alarm and in 64% we found correct rejections. More conscious and longlasting tasks (hands-free phoning, smoking) were remembered best. The interview method seems to be a valuable and valid tool to assess rather consciously conducted and legally prohibited secondary tasks while driving.