Abstract
The 81 items on the adjective check list of the Activity Vector Analysis (AVA) were analyzed for discrimination between a sample of 100 male prisoners and 5000 applicants and employees representative of the sex and occupation distributions in business and industry. T-score norms were established for the resulting scoring stencil of 41 items, using a sample of 400 male and 100 female non-prisoner AVAs. Additional samples of AVA forms were drawn randomly from data files for testing sex difference and discriminant cross validation. A significant (p = .001) sex difference on the experimental measure was found among non-prisoners (102 males vs 99 females) and 113 male and 76 female prisoners. Separate discriminant cross-validities among these subsamples were obtained for males and females. Correlations between the new measure and the Vectors of AVA and the factor scores of the 16 PF Test (n = 104 male junior college students) provided some evidence of construct validity and meaning. Predictive validation research has been initiated.