Arrindell W A, Barelds Dick P H, Janssen Irene C M, Buwalda Femke M, van der Ende Jan
Heymans Institute, University of Groningen, The Netherlands.
Br J Clin Psychol. 2006 Sep;45(Pt 3):377-91. doi: 10.1348/014466505x68924.
There are no studies available that have examined the factorial invariance of dimensions underlying the Symptom Checklist-90-Revised (SCL-90-R) across at least three distinct samples. In the following study, we wished to determine whether a dimensional model comprising eight primary factors previously identified in psychiatric out-patients, phobics and the general population (Arrindell & Ettema, 2003) could be extended to a homogeneous sample of pain patients comprising females suffering from peri partum pelvic pain (PPPP) syndrome (N = 413). The internal consistency and discriminant validity of the dimensions were also examined.
The SCL-90-R and measures of disability, pain-related fear, pain intensity and fatigue were administered to the participants. The multiple group method was used to determine factorial invariance. Pearson correlations were determined between the SCL-90-R and aforementioned measures.
The factorial invariance of an 8-dimensional model of primary factors underlying the SCL-90-R, namely, agoraphobia, anxiety, depression, somatization, cognitive-performance deficits, interpersonal sensitivity-mistrust, acting-out hostility and sleep difficulties, was extended with success to the present sample of PPPP patients. In spite of substantial correlations between the internally consistent SCL-90-R symptom dimensions, some evidence of discriminant validity was reported in that specific subscales showed different patterns of correlations with measures of disability, pain-related fear, pain intensity and fatigue.
The 8-dimensional system based on the work of Arrindell and Ettema (2003) was invariant across psychiatric patients, phobics, the general population and pain patients. The invariance of the SCL-90-R hostility dimensions may have implications for a re-formulation of Watson and Clark's tripartite model of general distress, specific anxiety and specific depression.
目前尚无研究考察过症状自评量表-90修订版（SCL-90-R）各维度在至少三个不同样本中的因素不变性。在以下研究中，我们希望确定一个由先前在精神科门诊患者、恐惧症患者和普通人群中识别出的八个主要因素组成的维度模型（阿林德尔和埃特马，2003年）是否可以扩展到一个由患有产后骨盆疼痛（PPPP）综合征的女性组成的疼痛患者同质样本（N = 413）。还对各维度的内部一致性和区分效度进行了考察。
对参与者施测SCL-90-R以及残疾、疼痛相关恐惧、疼痛强度和疲劳的测量工具。采用多组方法确定因素不变性。确定SCL-90-R与上述测量工具之间的皮尔逊相关性。
SCL-90-R所基于的主要因素的八维度模型，即广场恐惧症、焦虑、抑郁、躯体化、认知-表现缺陷、人际敏感-不信任、外显敌意和睡眠困难，成功扩展到了目前的PPPP患者样本。尽管内部一致的SCL-90-R症状维度之间存在显著相关性，但报告了一些区分效度的证据，即特定子量表与残疾、疼痛相关恐惧、疼痛强度和疲劳测量工具显示出不同的相关模式。
基于阿林德尔和埃特马（2003年）研究的八维度系统在精神科患者、恐惧症患者、普通人群和疼痛患者中具有不变性。SCL-90-R敌意维度的不变性可能对重新构建沃森和克拉克的一般痛苦、特定焦虑和特定抑郁的三方模型有影响。
