PROBLEM
Air traffic controllers have in the past few years come into great public prominence. Although there are only some 10,000 of these workers in centers and towers, it is now apparent to most people that even a slight inefficiency in the air traffic control system can lead to a dramatic disruption of American life. The Federal Aviation Administration (FAA) has been conducting extensive research into all types of factors affecting controller performance. Among this wide-ranging program have been a few attempts t o relate personality variables to job performance. Trites, Kurek and Cobb ( 6 ) studied the effects of over-and under-achieving during the training program on job performance, as well as their relation to the scales of the California Personality Inventory (a), and reported several significant relationships. Karson(5), using peer and supervisory ratings of job performance, related these to the personality factors measured by the 16 PF test(2) by means of correlational and factor analytic methods. I n general, it was found that peer and supervisory ratings of air traffic controllers (ATCs) were highly related to each other (P = +.55), but little relationship was found between peer and supervisory ratings and any of the personality variables, although the ratings showed correlations of -.18 and -.17 respectively with age. I n the factor analysis associated with the correlations, the two performance measures were alone on their own private factor and did not load on any of the personality factors.
This finding was surprising, especially in view of the fact that Trites, et uZ.,(*) had previously found relationships between performance and personality traits, and also because the senior author has found the 16 PF to be effective both in screening out potential psychiatric problems and in detecting high anxiety in controllers that was also found in clinical interview. Because of this, it was decided to attempt again to find relationships between the 16 PF scales and job performance measures, but using another estimate of job performance, the FAA's "Employee Appraisal Record for Nonsupervisory Employees" (EAR) This is an official U. S. Civil Service form, used for the annual evaluation of all employees of the appropriate rank in the FAA.
METHOD
The 16 PF, Form A, 1962 edition has, since 1966, been administered t o all operational controllers as part of the FAA's Air Traffic Controller Health Program. All 16 PFs were machine scored for the usual 16 scales, plus a "Motivational Distortion" scale, which attempts to determine response bias (0.
The basic criterion measure, the EAR, was made available by the FAA Civil Aeromedical Institute2 on many of the Ss who had taken the 16 PF. I n the light of the senior author's previous disillusionment with criterion measures, this form was examined carefully for evidence that it actually had a t least some validity.
The EAR consists of five sections, of which only three have direct relevance to performance. Part I1 (Career Potential) contains ratings by a superior on broad performance areas, such as "ability to reach decisions" and "creative ability.'' Part IV (Performance) has ratings on six areas of very specific relevance to the