Test of mechanical principles as a suppressor variable for the prediction of effectiveness on a mechanical repair job

Abstract

Psychological measures for prediction of job effectiveness in a skilled mechanical repair job were assessed using the usual validation-cross-validation paradigm. 63 industrial mechanical repairmen doing similar work were studied. The most predictive and operational selection system was determined through use of multiple-regression analysis. Test administration was simplified using a suppressor variable to obtain acceptable predictive efficiency with the smallest number of predictors. The prediction system employed a school achievement type of mechanical comprehension test to improve criterion prediction by suppressing noncriterion variance in another mechanical aptitude test. The suppressor relationship between the 2 tests supported a job description that specified mechanical knowledge and experience rather than knowledge of formally taught mechanical principles. (PsycInfo Database Record (c) 2025 APA, all rights reserved)