Cho Sun-Joo, De Boeck Paul, Embretson Susan, Rabe-Hesketh Sophia
Vanderbilt University, Nashville, USA,
Psychometrika. 2014 Jan;79(1):84-104. doi: 10.1007/s11336-013-9360-2. Epub 2013 Dec 12.
An additive multilevel item structure (AMIS) model with random residuals is proposed. The model includes multilevel latent regressions of item discrimination and item difficulty parameters on covariates at both item and item category levels with random residuals at both levels. The AMIS model is useful for explanation purposes and also for prediction purposes as in an item generation context. The parameters can be estimated with an alternating imputation posterior algorithm that makes use of adaptive quadrature, and the performance of this algorithm is evaluated in a simulation study.
提出了一种具有随机残差的加性多级项目结构（AMIS）模型。该模型包括在项目和项目类别层面上，项目区分度和项目难度参数对协变量的多级潜在回归，且两个层面均具有随机残差。AMIS模型不仅有助于解释，在项目生成情境中也可用于预测。可以使用一种利用自适应求积的交替插补后验算法来估计参数，并在模拟研究中评估该算法的性能。
