E-ISSN : 2733-4538
The most of the previous researches suggested that a combination of some MMPI scales are useful to identify random response set of the MMPI profiles. Greene (1978) recommended the TR > 4, Ca > 4, F > 80 for a clinical decision. In order to identify the validity of this criteria, 182 MMPI profiles obtained from psychiatric patients, 142 profiles of the students who visited to the university counseling center and 150 computer generated random response profiles were analyzed. The results show that the Greene's criteria does not classify random response set profiles successfully because most of the items of F scale do not satisfy the original selection criterion. As an alternative to this identification criteria, a new scale which would replace the F scale was constructed and applied. The consequencies are that a new scale is more successful in identifying the random response set profiles than the originally suggested criteria. A discriminant analysis using the TR, Ca and the new F scale as prediction variables showed that 97.4 % is successful in classifying the profiles while 97.0 % is successful in classifying the profiles used for cross validation.