Abstract

The purpose of this research is to improve the validity of MMPI as a diagnostic tool through examine the three validity scales of L, F and K of the test in relation to their effect of test-taking attitudes as a suppressor variable upon personality inventory. Two problems were taken for this research as follows: (1) Grasping the test-taking attitudes of each group, which shows different profiles, by examine the three hypotheses about the response tendency on. validity scales. (2) Testing the item validity about all the items contained in the three validity scales. The subjects were 500 normals and 200hospitalized psychotic patients. Among them 250 normals by random sampling were used for a tentative norm. According to two dimensions i. e. whether a subject is a normal or patient, and his MMPI profile is normal or abnormal, four experimental groups were constructed with 100 normals and 80 patients. Three hypotheses which established to solve the problem are as these: 1. High Lscores indicate the response tendency in socially desirable way. Therefore, the Group Pn should obtain the highest L scores among the four groups. 2. High F scores would be obtained when a subject would misrespond to the items because of his carelessness or thoughtlessness or ignorance about the MMPI items or excessive emotional disorders at the test situation. So, the Group Pab should reveal the highest F scores among the four groups. 3. The K scale measures the test-taking attitudes of subjects. And first, high K scores should be produced by the Group Pn owing to their conscious effort to fake good. Next, low K scores should be obtained by the Group Pab for their response tendency to bake bad or to reveal their problems in excessively frank due to their self-criticism. The methods used to solve the Problem [ was that Mean, Median and Standard Deviation of each validity scale scores were calculated for the four groups separately and then Analysis of Variance was taken. Next, on problem 2, item response frequencies for 109 items of the three validity scales were counted. And then, the item validities were examined by combinations of three conditions, item response frequency of each item, scoring direction of it and the results of problem 1. The conclusions of this research are as follows: 1. Subjects, who are different with scores of L, F and K scales of the MMPI, have signficantly different test-taking attitudes according to accepting the three hypotheses of the problem 1. 2. As a result of problem 2, 10 items among 109 items of the three validity scales are found invalid.