ABSTRACT
The empirical keying or contrasted-groups methodology used by Hathaway in developing the eight basic clinical scales of the MMPI represented a significant innovation in personality inventory construction. Previous inventories relied on intuitive judgment to produce items related to a trait of interest and to combine them with other items thought to be related to the same trait to compose a scale. The items and scales thus composed tended to be highly face valid (i.e., the items tended to be obviously related to the trait to be measured). Unfortunately, they were often unsatisfactorily related to external criteria and were overly vulnerable to test-taking attitudes of the examinee that would bias responses and potentially distort the results of testing.
