柯氏性格量表之精神病質因素量尺及正常因素量尺的同時性效度研究

Psychoticism scale and normalcy scale of KMHQ with the SAPS and SANS: The relationship study

柯氏等人（1987）為求柯氏性格量表的效度，將此量表進行因素分析，結果共得六項因素。本研究以精神分裂症為樣本，進一步探求六項因素中的精神病質因素量尺與正常因素量尺的內容效度。有關精神分裂症的症狀評量中，Andreasen（1982）發展的活性症狀評量表（scale for the Assessment of Positive Symptoms，簡稱為SAPS）與退化性症狀評量表（scale for the Assessment of Negative Symptoms，簡稱為SANS），是兩套具有良好的信度的量表。因此探討這兩份量表與精神病質因素與正常因素兩量尺的關係，可以說明精神病質因素與正常因素兩量尺的內容成分。本研究的結果如下：(1)精神病質因素的內容較偏重於測量精神分裂症患者的活性症狀，而正常因素僅與SANS量表的「缺乏意志——缺乏情感」及「缺乏生趣——缺乏社交」兩向度呈顯著的負相關。(2)精神分裂症患者在精神病質因素的得分顯著地比心理健康組的得分高，但是與心理不健康組的得分則無顯著的差異。在正常因素量尺的得分上，精神分裂症組與心理健康組的區辨效度受性別的影響，而精神分裂症組與心理不健康組的得分則無明顯的差異。(3)本研究的精神分裂症患者在SAPS與SANS兩份量表的反應結果，並不支持Andreasen（1984）的觀點——SAPS與SANS兩份量表是兩套具有負相關內容的量表。

In the study of construct validity of KMHQ, six factors were derived by factor analysis (Ko et al., 1987). This research further investigates the content validity of psychoticism scale and normalcy scale, the two of the six factors of KMHQ. Scale for the assessment of positive symptoms (SAPS for short) and scale for the assessment of negative symptoms (SANS for short) wewe developed by Anderson (1982), both woth high reliability to assess the symptoms of the schizophrenia. To analyze the relationship of the SAPS and SANS with the psychoticism scale and normalcy scale may illustrate the components of the latters. The results were: (1) Most of psychoticism scale's content assess the positive symptoms of the schizophrenia. Normalcy scale and SANS have the significantly negative correlation in the dimensions of ''avolition-apathy'' and ''ahedonia-asociability''. (2) In the psychoticism scale, the scores of the schizophrenia were significantly higher than the mental healthy, but no significant difference with the mental ill. In the normalcy scale, the discrimination between the schizophrenia and the mental healthy was interacted by sex. The scores of the schizophrenia and the mental ill had no significant difference. (3) Under this research, the scores of the schizophrenia in SAPS and SANS had no negative correlation. The result didn't support the idea of Anderson (1984).