Global Versus Decomposed Estimates of Cross-Job Retraining Time

Cross-job retraining is becoming a viable option for coping with increasingly rapid technological changes in the workplace. In this study, we used data col- lected from 836 supervisors in 43 U.S. Air Force enlisted jobs to compare global versus decomposed estimates of cross-job retraining time in terms of interrater reliability and convergent validity. Convergent validities of retrain- ing time estimates were assessed in terms of their correlations with each other and with two additional determinants of retraining ease: learning difficulty of the new job, and old-versus-new job differences in aptitude requirements. In general, the reliabilities for the global and decomposed judgments were com- parable. Additional correlational results supported the convergent validities of both the global and decomposed retraining time estimates.