| Abstract: | Asample of 103 enlisted naval jobs was evaluated on a 13-factor experimental job evaluation system. For a criterion measure, the jobs were independently ranked on overall difficulty and responsibility. Using the Wherry-Doolittle test selection method on a subsample of 58 jobs, five factors were selected that gave a shrunken multiple R of .968 with the criterion. On a hold-out sample, the prediction equation gave criterion scale values that correlated .937 with the actual criterion scale values. It is suggested that, in the absence of any 'true' criterion of naval job values, overall judgments may well be accepted as a criterion of use in job evaluation research. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
