Multiple imputation for missing data: Making the most of what you know
Multiple imputation for missing data: Making the most of what you know
复制标题
DOI：
10.1177/1094428103255532
复制
发表时间：
2003-07-01
影响因子：
9.5
通讯作者：
Cummings, JN
中科院分区：
文献类型：
--
作者：
Fichman, M;Cummings, JN
缺失数据是组织研究中常见的问题。缺失数据可能是由于纵向研究中的损耗或实验室或现场环境中对问卷项目的不回答。对缺失数据的不当处理（例如，列表删除、平均值填补）可导致使用完整病例分析统计技术的有偏倚的统计推断。本文介绍了一个模拟和数据分析的案例研究，使用的方法来处理缺失的数据，多重插补，允许有效的统计推断与完整的案例统计分析。用于在多变量正态模型下实现多重插补的软件是免费且广泛可用的（例如，NORM、SAS、SOLAS）。应常规考虑插补缺失数据。作者使用HomeNet项目的数据说明了这种技术的应用。
Missing data are a common problem in organizational research. Missing data can occur due to attrition in a longitudinal study or nonresponse to questionnaire items in a laboratory or field setting. Improper treatments of missing data (e.g., listwise deletion, mean imputation) can lead to biased statistical inference using complete case analysis statistical techniques. This article presents a simulation and data analysis case study using a method for dealing with missing data, multiple imputation, that allows for valid statistical inference with complete case statistical analysis. Software for implementing multiple imputation under a multivariate normal model is freely and widely available (e.g., NORM, SAS, SOLAS). It should be routinely considered for imputing missing data. The authors illustrate the application of this technique using data from the HomeNet project.