Clark A, Friedman M J
J Pers Assess. 1983 Aug;47(4):396-404. doi: 10.1207/s15327752jpa4704_10.
This study tested whether the symptom dimension subscales derived for the Symptom Checklist-90 (SCL-90) by factor analysis could be replicated in a new sample. It was administered to 442 veterans undergoing psychiatric treatment at a VA psychiatric outpatient clinic. Factor analysis using the principal-components method yielded a factor structure different from those previously reported for other populations. The first factor extracted, Depression, explained 37% of the variance of the entire instrument, or more than eight times the variance explained by the second factor. Only five of the nine reported SCL-90 symptom subscales emerged in this study; the Anxiety and Psychoticism subscales disappeared, and the Paranoia and Interpersonal Sensitivity subscales merged. In this population the instrument seems to measure a single global distress factor instead of nine independent symptom subscales as reported previously. This finding was corroborated by the fact that depressed, anxious, and schizophrenic patients showed no differences in SCL-90 symptom profile shapes, although they did differ in overall symptom intensity.
本研究检验了通过因子分析从症状自评量表90（SCL - 90）得出的症状维度分量表是否能在一个新样本中得到重复。该量表施用于在退伍军人事务部（VA）精神科门诊接受治疗的442名退伍军人。使用主成分法进行因子分析得到的因子结构与先前针对其他人群报告的不同。提取的第一个因子“抑郁”解释了整个量表37%的方差，是第二个因子所解释方差的八倍多。本研究中仅出现了九个报告的SCL - 90症状分量表中的五个；“焦虑”和“精神病性”分量表消失了，“偏执”和“人际敏感”分量表合并了。在这个群体中，该量表似乎测量的是一个单一的总体痛苦因子，而非如先前报告的九个独立症状分量表。抑郁、焦虑和精神分裂症患者在SCL - 90症状剖面图形状上没有差异这一事实证实了这一发现，尽管他们在总体症状强度上存在差异。
