Selecting predictors for maximizing the classification efficiency of a battery

Determined the effect that variables in a personnel classification system have on classification efficiency by using a simulation approach that relies on synthetic predictor scores and predicted performance criterion scores for multiple jobs. Classification efficiency was measured as mean predicted performance determined after optimal assignment to jobs. Population parameters required to generate synthetic scores and to compute evaluation equations for simulations were obtained from the US Army's Project A (1990) study. The present study provides a comparison of differential assignment theory (DAT) with general aptitude theory and validity generalization. The theoretical value and practical usefulness of DAT is supported by the finding that both longer test batteries and the use of P. S. Horst's (1954) differential validity index to select tests increase potential classification efficiency.