摘要: |
The Federal Aviation Administration (FAA) is commencing a massive hiring of air traffic control specialists using a new selection procedure, the Air Traffic Selection and Training (AT-SAT) computerized test battery. Before AT-SAT could be used for hiring purposes, however, the issue of its potential for adverse impact (potential unfair discrimination) had to be addressed. A previous project (Wise, Tsacoumis, Waugh, Putka, & Hom, 2001) reweighted the subtests and adjusted the overall constant to mitigate potential group differences that could result in adverse impact, without unduly compromising validity. A subsequent study (Dattel & King, 2006) used research participants and found that this effort appeared to have achieved its goal of mitigating group differences that could result in adverse impact. The present study endeavors to: 1) describe how AT-SAT functions as an operational selection method with respect to the several applicant pools, and 2) determine how the reweighting effort fares with actual applicants in the goal of reducing/eliminating group differences that could result in adverse impact. Of the 854 applicants who have taken AT-SAT as part of a job application process (rather than as according to a research protocol), 219 applicants (25.64%) voluntarily disclosed their race; gender was known for 253 (29.63%). The results suggest that the reweighting effort is paying dividends as group differences that could result in adverse impact are not in evidence. While the initial numbers reported here are relatively small, the issue of group differences that could result in adverse impact will be continually monitored. Longitudinal validation, comparing AT-SAT results to training and on-the-job performance, is a research priority due to concerns about the overall passing rate of 93.33%, which is higher than the expected passing rate of 67%. |