Abstract
The effects of employing Goldberg's index of item ambiguity, Ambdex, in the construction of personality scales was investigated with 200 male and 200 female adults, aged (M) 40.1 yr. and 37.1 yr. respectively. The basic procedure involved the development of scales of MMPI items maximally correlating with WAIS subtest performance. The cross-validation correlations of scales developed from the total item pool were compared with the cross-validation correlations of scales constructed from only those items previously identified as having low Ambdex scores. The size of the cross-validation correlations of the two groups of scales was essentially the same even though the low Ambdex scales were approximately half the length of the scales developed from the total item pool.
Get full access to this article
View all access options for this article.