Schizophrenic patients have been reported to show characteristic pattern in Rorschach responses. But existing studies for testing and confirming the schizophrenic Rorschach indices have tended to fail in establishing reliable indices becaue of the methodological flaws and analytical limitations. This study was intended to make Rorschach criteria of schizophrenia which has both differential power and cross-validity. The study consisted of 2 parts.
The first study aimed to identify Rorschachvariables with discriminating power among the schizophrenics, the neurotics and the normals. The subjects were ?8 schizophrenics, 44 neurotics and 88 normals. The clinical subjects were inpatients in Hanyang University Hospital from 1981 to 1985. And the normals, matched to the clinical subjects in sex, age and educational level, were before-hand screened by MMPI.
The Rorschach test was administered and scored by 5 experienced clinical psychologists. Seventy-six Rorschach variables were selected and scored using Exner system. The inter-score reliability was .88.
Analysis of variance indicated 9 variables characteristic to schizophrenic group: frequent failure, Dd, Unusual Verbalization (disordered thinking) ; increased C, Sum C, Experienced .Actual (inadequate controls) ineptness) ; low F+%, X+~lo (distorted perception).
These nine variables were finally determined as discriminant variables. The discriminant function is as follows.
f (x)=O. 47x (failure) +0. 105 x (Dd) +0.258 x (Unusual Verbalizations) + 0. 286 X (C)+0.077 x (EA) 0.0046 x (F+ %) 0.043 x (X+ %) + 0.0099 x (M%) +0. 111 x (Sum Q+2.713
The mean discriminant scores were 1. 2988 for the schizophrenics, -0.2447 for the neuro tics and -0.9442 for the normals.
The correctly discriminated percentage was 68% for the schizophrenics, 580/o for the neurotics and 78% for the normals.
In the 2nd part, new subject groups (35 schizophrenics, 23 neurotics and 45 normals) were sampled for testing the cross-validity of the discriminant function. The results proved that the function has not only considerable differential power but also could be reliable criteria of schizophrenia; the correctly predicted percentage was 71/o for the schizophrenics, 53% for the neurotics and 67% for the normals.
Considering the cultural differences, the above investigation results were compared with Exner´s. Despite the differences in discriminant variables (failure, H,M.P), it was proved that there were no significant differences between the discriminating power of Exner´s and author´s function. The possiblecultural factors for differences in ,variables and the reasons why failed in revealing any significant differences in discriminating power were discussed.
In addition, qualitative analysis was applied to the mispredicted cases for detecting the cause of misprediction and for improving the better discriminating power.
Finally, limitations and implications of this study were discussed.
|