A psychometrician's dream: Test validities in an unrestricted sample of future Swiss Air Force pilots

Abstract

Introduction. In the context of personnel selection the predictive validities of the selection tests are usually based on the restricted sample of recommended applicants. The resulting validity coefficients can be corrected by the restriction-of-range formula, which is based on a number of assumptions especially in multivariate settings. Research question. Which tests show the highest predictive validity for flying performance in an unrestricted sample of student pilots? Method. N=135 Swiss Air Force applicants were examined with the DLR selection battery. Criterion information about flying performance was collected after ten flying lessons. Results. A number of substantial predictor-criterion correlations were found. Discussion. The highest predictive validities were found for psychomotor tests under multitasking conditions, spatial orientation, and mechanical comprehension. Conclusion. The findings confirm results from earlier studies showing that complex psychomotor tests are the best predictors for the first stages of training to become a successful pilot.