Citation

Lawson JJ. Accid. Anal. Prev. 1982; 14(5): 371-380.

Copyright

(Copyright © 1982, Elsevier Publishing)

DOI

unavailable

PMID

unavailable

Abstract

For comparison with regularly-available road accident data, the Department's Road Safety Branch attempted to obtain aggregate exposure measures in its National Driving Survey of 1978-1979. Distance travelled and travel time were the major target variables, to be classified by age, sex and experience of drivers, model year and type of vehicles, roadway type and surface condition, and time of day, week and month. Attempts were also made to simultaneously obtain data on alcohol use and vehicle fuel consumption. The method chosen was to sample drivers by geographic area and by month for one year, obtaining from each respondent a seven-day log of trips driven in vehicles under 10,000 lb gross weight. The logs included odometer readings to reveal trip distances. Two personal interviews were conducted with each respondent: first to obtain background data on drivers and vehicles, and introduce the trip "diary"; and subsequently to review the diary for accuracy. A pilot survey revealed that drivers could be persuaded to provide complete trip records, but that the survey methods did not produce accurate data on exposure by class of road, nor reliable data on either alcohol use or fuel consumption. The main survey took place from May 1978 to June 1979, obtaining usable responses from almost 10,000 drivers on approximately 200,000 trips. Response rates to different stages of the sampling and survey procedure are presented, the most significant being that 62% of drivers provided apparently-complete diaries. A follow-up investigation of non-respondents produced some success, and some reassurance about non-response bias.