| Abstract: | Through differential weighting of application-blank information (N = 152), 2 validations against voluntary female clerical-turnover criteria were accomplished. Concurrent (16 scoreable items with N = 72) and predictive (13 scoreable items with N = 60) cross validities are reported on holdout and follow-up samples, respectively. Statistically significant relationships are demontrated in both the concurrent and the predictive studies despite intervening variables, preceding the predictive study, sufficient to destroy validity. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
