Hales Derek P, Dishman Rod K, Motl Robert W, Addy Cheryl L, Pfeiffer Karin A, Pate Russell R
Department of Exercise Science; The University of Georgia; Ramsey Student Center, 300 River Road; Athens, GA 30602-6554, USA.
Ethn Dis. 2006 Winter;16(1):1-8.
Meaningful comparison of depression symptoms requires that the measurement instrument has equivalent measurement properties among racial and ethnic groups. We tested the factorial validity and invariance of the Center for Epidemiologic Studies-Depression (CES-D) Scale among Black (n=610) and White (n=452) adolescent girls in the 12th grade. The invariance analyses were conducted by using LISREL 8.70 with maximum likelihood estimation and Satorra-Bentler scaled chi-square statistics and standard errors. The hypothesized second-order model (first-order factors: depressed affect, somatic and retarded activity, interpersonal, and positive affect; second-order factor: depression) demonstrated good overall fit in both groups. Comparison of nested models supported the between-group invariance of the overall factor structure, first- and second-order factor loadings, first-order factor variances, and the second-order factor variance. Item uniquenesses were not invariant. Our results support the hypothesis that a meaningful comparison of composite CES-D scores can be made between Black and White girls in the 12th grade.
对抑郁症状进行有意义的比较，要求测量工具在不同种族和族裔群体中具有等效的测量属性。我们在12年级的黑人（n = 610）和白人（n = 452）少女中测试了流行病学研究中心抑郁量表（CES-D）的因子效度和不变性。不变性分析使用LISREL 8.70进行，采用最大似然估计以及Satorra-Bentler尺度卡方统计量和标准误差。假设的二阶模型（一阶因子：抑郁情绪、躯体和迟缓活动、人际关系以及积极情绪；二阶因子：抑郁）在两组中均显示出良好的整体拟合度。嵌套模型的比较支持了整体因子结构、一阶和二阶因子载荷、一阶因子方差以及二阶因子方差在组间的不变性。项目独特性并非不变。我们的结果支持这样一个假设，即可以在12年级的黑人和白人女孩之间对综合CES-D分数进行有意义的比较。
