MMPI cannot say scores: normative data and degree of profile distortion.

Abstract

Few previous studies have investigated MMPI Cannot Say scores (the number of items people omit in completing the MMPI). Study 1 examined the frequency of occurrence of omitted items in three different populations. A significantly higher proportion of complete MMPI answer sheets (no omitted items) was found among job applicants than among two psychiatric groups, and the vast majority of test subjects in all three groups responded to all items. Study 2 investigated the degree of profile distortion introduced by randomly omitting various numbers of MMPI items from complete answer sheets. As more MMPI items were omitted, there was a progressive reduction in the elevation of the MMPI profile, but several different analyses failed to reveal a definite point at which the number of omitted items clearly became excessive. Changes in the profile's high-point pair were sometimes produced by numbers of omitted items previously considered to be expected in a "normal" MMPI record. This type of profile distortion could alter significantly the interpretation of a profile. Changes were therefore suggested in the T-score values for the Cannot Say scale given on the standard profile.