A Comparison of Alternative Proposed Subscale Structures for MMPI Scale 2

Scale 2 of the MMPI was subjected to confirmatory (Oblique Multiple Groups) factor analysis, following the procedures employed by Bernstein and Eveland (1982) in a related context. Various proposed subscale structures (those of Comrey, 1957; Harris & Lingoes, 1955; Weiner, 1948) were evaluated in two different populations: Job Applicants and Psychotherapy Outpatients. Particular effort was directed toward seeing how differences in item endorsement probability (response level) affect the item level structure of Scale 2 and, therefore, the proposed subscales. It was found that none of the proposed subscales explained the item level structure very well. The conclusion was that there is nothing to be gained by use of the proposed subscale systems compared to the use of Scale 2 as a whole. Although there is an apparently large set of factors present in Scale 2, these factors seem to reflect nonsubstantive aspects of item interrelationships.