Detecting mixtures from structural model differences using latent variable mixture modeling: A comparison of relative model fit statistics
Detecting mixtures from structural model differences using latent variable mixture modeling: A comparison of relative model fit statistics
复制标题
DOI：
10.1080/10705510709336744
复制
发表时间：
2007-01-01
影响因子：
6
通讯作者：
Kim, Kevin H.
中科院分区：
文献类型：
--
作者：
Henson, James M.;Reise, Steven P.;Kim, Kevin H.
采用3（样本量）× 3（外源性潜在平均差异）× 3（内源性潜在平均差异）× 3（因子间相关性）× 3（混合比例）析因设计探索潜在变量混合建模中结构模型参数估计值的准确性。此外，几种基于似然性的统计方法的有效性（Akaike信息准则[AIC]、贝叶斯信息检验[BIC]、样本量调整BIC [ssBIC]、一致AIC [CAIC]、Vuong-Le-Mendell-Rubin调整似然比检验[aVLMR]）、基于分类的统计（CLC [分类似然信息准则]，ICL-BIC [综合分类似然]，归一化熵准则[NEC]，熵），和分布统计（多变量偏斜和峰度检验）来确定哪种统计最好地恢复正确的组分数目。结果表明，结构参数得到了恢复，但模型拟合统计数据并不非常准确。ssBIC统计量是最准确的统计量，CLC、ICL-BIC和aVLMR显示出有限的实用性。然而，这些统计数据对于小样本（n = 500）均不准确。
The accuracy of structural model parameter estimates in latent variable mixture modeling was explored with a 3 (sample size) x 3 (exogenous latent mean difference) x 3 (endogenous latent mean difference) x 3 (correlation between factors) x 3 (mixture proportions) factorial design. In addition, the efficacy of several likelihood-based statistics (Akaike's Information Criterion [AIC], Bayesian Information Ctriterion [BIC], the sample-size adjusted BIC [ssBIC], the consistent AIC [CAIC], the Vuong-Le-Mendell-Rubin adjusted likelihood ratio test [aVLMR]), classification-based statistics (CLC [classification likelihood information criterion], ICL-BIC [integrated classification likelihood], normalized entropy criterion [NEC], entropy), and distributional statistics (multivariate skew and kurtosis test) were examined to determine which statistics best recover the correct number of components. Results indicate that the structural parameters were recovered, but the model fit statistics were not exceedingly accurate. The ssBIC statistic was the most accurate statistic, and the CLC, ICL-BIC, and aVLMR showed limited utility. However, none of these statistics were accurate for small samples (n = 500).