中国版SCL-90的效度研究

目的：检测中国版症状自评量表（The symptom Check-List-90,SCL-90)的效度。方法：624名正常人和268名心理门诊病人用SCL-90测量。结果：⑴SCL-90的各因子内部一致性较好，α系数在0.78-0.90之间；⑵判别函数分析发现SCL-90能很好区分病人和正常人，正判断为80.6%；⑶因素分析表现出单个因素能代表SCL-90大多数的因子；⑷SCL-90的九个因子之间存在高的相关，相关系数范围在0.39-0.79之间。结论：⑴SCL-90作为精神症状测验和研究量表是可接受的；⑵SCL-90不能作为不同精神症状区分和诊断量表；⑶SCL-90不是一个多维度的量表。

Study on the validity of the Symptom Check-List-90 of Chinese version

Objective To investigate the validity of the Chinese version of the Symptom Check-List-90 (the SCL-90) in Chinese population. Methods A total of 624 normal person and 268 out patients were examined with the SCL-90. Results (1)The internal consistency of the original subscales was found to be good and Cronbach's coefficient alpha ranged from 0.78 to 0.90; (2)Discrimination function analysis based on the 9 original subscales showed that the power of the SCL-90 in discriminating the patients and the community was good with a corrective rate of 80.6%; (3)Factor analysis on the items of the questionnaire yielded a very strong un-rotated first factor which could represent other factors; (4)High inter-correlations were found between the 9 original subscales of the SCL-90 with a correlation coefficient ranging from 0.39 to 0.79. Conclusion (1) The SCL-90 is acceptable as a research setting and psychiatric screening inventory; (2) The SCL-90 does not have potential as an inventory to distinguish different diagnostic groups; (3) The SCL-90 is not a multidimensional inventory.