Predictive Validity of the Air Force Officer Qualifying Test (AFOQT) for Pilot Performance

Abstract

Abstract. The Air Force Qualifying Test (AFOQT) has been the primary selection test battery for officer candidates in the US Air Force since 1953. Despite a wealth of literature on the validity of the AFOQT in predicting pilot performance, there is less evidence on its validity generalization. This study investigated the predictive validity of 16 AFOQT subtests and its Pilot composite via psychometric meta-analytic procedures. Based on 32 independent samples from 26 studies, results indicated that pilot performance is best predicted by subtests indicative of perceptual speed, aviation-related aptitude and knowledge, and quantitative ability constructs, and least predicted by subtests indicative of verbal ability construct. Evidence for validity generalization of AFOQT subtests is presented, and implications for practical use are discussed.

A meta-analysis on air traffic controllers selection: cognitive and non-cognitive predictors

“…The same results are observed in pilots (ALMamari & Traynor, 2020). In Martinussen et META-ANALYSIS ON ATCO SELECTION 33 al.…”

Aviation psychology: development stage in science and education

“…Modern aviation psychology is in possession of a methodology for psychological support of aircraft-specific work, that takes into account flight duration, the number of crew members, departmental affiliation, the type of aircraft, and the specificity of extreme situations [17; 31; 32]. Special techniques for diagnosing psychophysiological and psychological qualities have been developed and introduced into the practice of professional-aviation staff selection [22]. The development of aviation technology in the twentieth century has led to the creation of simulators capable of forming not only the motor skills of the airborne staff but also contributing to prudence, motivation, self-control, intellectual orientation with foresight elements, the "feeling" of the airplane and flight vision, operational mindset [7; 25; 28].…”

Predicting Commercial Pilot Training Performance

“…The meta-analysis of the AFOQT included a total of 32 samples from 26 studies and investigated the predictive validity of single measures and of the total pilot score for this specific test battery. The mean (uncorrected) correlations with pilot training were .17 for the Pilot Composite and ranged between .00 (word knowledge) to .17 (instrument comprehension and scale reading; ALMamari & Traynor, 2020) for single measures. An earlier meta-analysis of the different constructs measured by the AFOQT indicated a mean predictive validity of .15 for verbal ability and .24 for perceptual speed based on five studies (Johnson et al, 2017).…”

“…Based on both meta-analyses and primary studies, there is strong evidence that cognitive and psychomotor tests predict future pilot performance; however, some variation between studies has been detected (ALMamari & Traynor, 2019, 2020; Hunter & Burke, 1994; Martinussen, 1996).…”

“…Two more recent meta-analyses had more limited scopes: one summarized the predictive validity of multiple test batteries (ALMamari & Traynor, 2019) and the other looked at one specific test battery used for military selection in the United States (the Air Force Officer Qualification Test, AFOQT; ALMamari & Traynor, 2020). The meta-analysis of multiple test batteries by ALMamari and Traynor (2019) was based on 118 independent samples consisting of either military (93 studies) or civilian pilots including university programs (25 studies), and found mean predictive validities (uncorrected) for the five different test categories ranging from .10 (controlled attention) to .34 (work sample).…”