RESUMO
Over 8,800 returns of a questionnaire, distributed to Canadian pilots, were analyzed. Items were chiefly about life events, hobbies, humour, and characteristics which might bear on susceptibility to "pilot error" accidents. Sets of items were compiled which discriminated whether or not respondents had had an aircraft accident in the past 10 years. Around 70% of the respondents could be classified correctly according to the self-reported accident criterion. Somewhat over half the items were significant accident markers in at least one of four pilot-license categories examined. There was surprisingly little evidence of a "macho" syndrome associated with aircraft accidents; for example, participation in relatively risky hobbies was often negatively related to accident history. This report, like much past research, is entirely retrospective, so that some significant accident markers may only reflect accident aftereffects. Since the questionnaire was not anonymous, however, a follow-up study on the prediction of future accidents is feasible and is being considered.