Department of Psychology, University of Toledo.
J Trauma Stress. 2011 Aug;24(4):435-43. doi: 10.1002/jts.20653. Epub 2011 Aug 10.
The present study used factor mixture modeling to explore empirically defined subgroups of psychological trauma victims based on confirmatory factor analysis (CFA) and latent class analysis of posttraumatic stress disorder (PTSD) symptoms. We sampled 310 medical patients with a history of trauma exposure. Confirmatory factor analysis revealed that the 4-factor emotional numbing PTSD model yielded the best model fit. Using latent factor means derived from this model and the 4-factor dysphoria PTSD model (indexing severity on PTSD factors), 3 latent classes of participants were identified using factor mixture modeling. The 3-class model fit the data very well and was validated against external measures of anxiety and rumination.
本研究采用因子混合建模，通过验证性因子分析（CFA）和创伤后应激障碍（PTSD）症状的潜在类别分析，对心理创伤受害者进行实证定义的亚组探索。我们对 310 名有创伤暴露史的医疗患者进行了抽样。验证性因子分析表明，4 因素情绪麻木 PTSD 模型具有最佳的模型拟合度。使用该模型的潜在因子均值和 4 因素烦躁 PTSD 模型（对 PTSD 因子的严重程度进行索引），使用因子混合建模确定了 3 个参与者的潜在类别。3 类模型非常好地拟合了数据，并针对焦虑和沉思的外部测量进行了验证。
