Imputation for incomplete high-dimensional multivariate normal data using a common factor model
Imputation for incomplete high-dimensional multivariate normal data using a common factor model
复制标题
DOI：
10.1002/sim.1867
复制
发表时间：
2004-09-30
影响因子：
2
通讯作者：
Belin, TR
中科院分区：
文献类型：
--
作者：
Song, JW;Belin, TR
在应用研究中，通常会在少量的案例中测量大量的变量。即使单个变量的缺失率很低，这些数据集也可能有大量不完整的案例。在这里，我们提出了一种新的方法来处理失踪的连续标度项目在多元数据中，提取共同的因素，以减少在多元正态模型的协方差参数估计的数量的基础上。该技术在几个模拟设置可用的情况下进行分析，并与岭前的多元正态模型进行比较。该方法还说明了一项研究，超过100个变量评估急诊室干预自杀未遂的青少年。版权所有（C）2004约翰威利父子有限公司。
It is common in applied research to have large numbers of variables measured on a modest number of cases. Even with low rates of missingness on individual variables, such data sets can have a large number of incomplete cases. Here we present a new method for handling missing continuously scaled items in multivariate data, based on extracting common factors to reduce the number of covariance parameters to be estimated in a multivariate normal model. The technique is compared in several simulation settings to available-case analysis and to a multivariate normal model with a ridge prior. The method is also illustrated on a study with over 100 variables evaluating an emergency room intervention for adolescents who attempted suicide. Copyright (C) 2004 John Wiley Sons, Ltd.