Abstract

The purpose of the present study was to examine the validity of empirically keyed biodata. 65 biographical items were administered to 5,922 insurance saleswomen. The criterion measure of job success was obtained from company records. The empirical keying method based upon the WAB was employed to score the biodata. The initial result showed that scoring weights developed on a total sample were somewhat unstable. Thus, the sample was divided into old versus young subgroups and the scoring keys were developed separately. The result of cross-validation revealed scoring weights of both subgroups were stable as well as having high predictability. Also, the result of discriminant analysis showed practical significance of biodata as a selection tool. Finally, problems and future directions of the empirical keying method were discussed.