Abstract

The Purpose of this study is to construct Korean scales of MMPI and to analyze its reliability and validity. The subject were 690 psychiatric inpatients, 183 juvenile delinquents from confinement services, and 656 normal adults. MMPI was administered to all of the subjects and item analysis was conducted based on data obtained from the subjects. Test-retest reliability was calculated from 168 normal data and factor analysis was done on the data obtained from 4,295 normal data. In constructing Korean version MMPI, quite a number of items contained in the scales of NF(38 items), NSc(30 items), NHy(26 items) need to be corrected. 14(D) to 10(Pa) items were found to be inappropriate in ND, NPd, NMf-f, NPa, NPt, NMa Scales while less than 10 items were identified to be corrected in NL, NK, NHs, NMf-m, NSi Scales,. Internal consistency and test-retest reliability of Korean version of MMPI were proved to be acceptable. However, for the NMf Scale the reliability was very low. It is speculated that the items of Mf scale of the original MMPI were constructed with data obtained from homosexual patients. Since the homosexuality is not considered as serious mental problem in Korean culture, NMf scale needs a further attention in cultural aspects. Korean version of MMPI was classified into 3 factors which could be named as psychopathology, depression/introversion, defensiveness-anxiety. Even though these factors were not fully consistent in their sequential order, they exist consistently across sex and age. The results of discriminant analysis on the existing MMPI revealed relatively high predictive power for diagnostic classification, while the Korean version of MMPI is much higher than the former in most patient groups. The predictive power of diagnostic classification was elevated approximately 7% for manic patient group. In particular the results suggest that Korean scales of MMPI considering cultural variables is superior to existing MMPI in construct validity as well as external validity.