Predictive Validity of Knowledge Tests for Pilot Training Outcome

Abstract

Tests of subject-specific knowledge (mathematics, physics, and English as a foreign language for non-English-speaking countries) are commonly used in test batteries for the selection of ab initio pilots – for example, the Air Force Officer Qualifying Test (AFOQT) or the German Aerospace Center (DLR) assessment. However, in validity research, knowledge tests are often underrepresented. This study evaluated the predictive validity of knowledge tests compared with cognitive ability tests and school grades. The validity criterion was the outcome of pilot training (pass/fail) of a preselected group of applicants (N = 402) who completed a 2-year flight training program. The predictive validity of the entire test battery was r = .55. Cognitive ability tests, knowledge tests, and school grades emerged as comparably valid predictors. These findings are discussed in the framework of Cattell’s theory of fluid and crystallized intelligence, suggesting that knowledge tests are predictively valid because they are indicators of motivation and of being a good learner.