PREDICTING PILOT ERROR INCIDENTS OF US AIRLINE PILOTS USING LOGISTIC-REGRESSION

In a population of 70164 airline pilots obtained from the Federal Avia tion Administration, 475 males and 22 females had pilot-error incident s in the years 1986-1992, A simple chi-squared test revealed that fema le pilots employed by major airlines had a significantly greater likel ihood of pilot-error incidents than their male colleagues, In order to control for age, experience (total flying hours), risk exposure (rece nt flying hours) and employer (major/non-major airline) simultaneously , the author built a model of male pilot-error incidents using logisti c regression, The regression analysis indicated that youth, inexperien ce and non-major airline employer were independent contributors to the increased risk of pilot-error incidents, The results also provide fur ther support to the literature that pilot performance does not differ significantly between male and female airline pilots. (C) 1997 Elsevie r Science Ltd.