From dust bowl empiricism to rational constructs in biographical data

A number of authors have noted the possibility of using rationally derived biographical scales in prediction. The present report was an attempt to use empirically, factorially, and rationally derived biographical scales in a concurrent validation to predict service orientation. Over 867 service empl oyees, along with a large number of applicants, completed a 137-item biogra phical questionnaire constructed to cover the 15 aspects of service deliver y identified through an extensive job analysis. Criteria included superviso ry ratings of service orientation and overall performance along with three personnel measures, days absent, times absent, and times tardy. Empirical, factorial, a priori rational, and post-hoc rational scales were developed w ith a two-thirds sample and cross-validated using a one-third hold-out samp le. The empirical scales had the highest validity and the greatest shrinkag e, rendering this approach ineffective in predicting any of the criteria. T he factorial and rational scales had positive validity with four of the fiv e criteria, with the post-hoc rational scales having cross-validated correl ations in the 0.30's with three of the criteria including the rating of ser vice orientation. The results were promising in terms of both the predictio n and understanding of a criterion of increasing importance to many organiz ations, customer service orientation.