ISSN 1671-3710
CN 11-4766/R
主办：中国科学院心理研究所
出版：科学出版社

心理科学进展 ›› 2014, Vol. 22 ›› Issue (12): 1985-1994.doi: 10.3724/SP.J.1042.2014.01985

• 研究方法 • 上一篇    

追踪研究中缺失数据处理方法及应用现状分析

叶素静1;唐文清1,2;张敏强1;曹魏聪1   

  1. (1华南师范大学心理学院, 心理应用研究中心, 广州 510631) (2广西大学教育学院, 南宁 530004)
  • 收稿日期:2014-01-01 出版日期:2014-12-15 发布日期:2014-12-15
  • 通讯作者: 张敏强, E-mail: zhangmq1117@qq.com
  • 基金资助:

    国家社会科学基金“十二五”规划教育学一般课题(BHA130053)、广东省哲学社会科学规划项目(09sxk1q001)、广州市基础教育学业质量监测系统项目(GZIT2013-ZB0465)、“国家基础科学人才培养基金‘华南师范大学心理学基地’(J1030729)”。

Techniques for Missing Data in Longitudinal Studies and Its Application

YE Sujing1; TANG Wenqing1,2; ZHANG Minqiang1; CAO Weicong1   

  1. (1 Center for Studies of Psychological Application, School of Psychology, South China Normal University, Guangzhou 510631, China) (2 School of Education, Guangxi University, Nanning 530004, China)
  • Received:2014-01-01 Online:2014-12-15 Published:2014-12-15
  • Contact: ZHANG Minqiang, E-mail: zhangmq1117@qq.com

PDF (PC)

5122

被引次数

11

摘要：

追踪研究中普遍存在缺失数据, 缺失数据处理方法的选择影响统计推断的精度及研究结果的有效性。首先, 阐述缺失机制及判断方法, 比较追踪研究中主要的缺失数据处理方法的特点、及实际应用中的缺失处理方法的选择和软件实现。其次, 对国内心理学中92篇追踪研究文献进行分析, 发现有59篇(64.13%)报告不同程度缺失, 其中仅39篇报告了处理方法且均为删除法。未来研究应深入探讨现有缺失数据处理方法的有效性, 进一步规范应用研究中缺失数据的处理。

关键词: 追踪研究, 缺失数据, 缺失机制, 缺失数据处理方法

Abstract:

Missing data are not uncommon in longitudinal studies. Different techniques for handling missing data affect accuracy of the results and validity of statistical inference. Firstly, we will elaborate on missingness mechanism and how to judge them. Then we make a summary of missing data techniques that mainly used in longitudinal study, and how to choose an appropriate missing data technique as well as software for analysis. Secondly, based on a literature review of psychology research in China, among 92 studies, we found that 59 contain a certain degree of missing data. Among these, 39 studies reported using deletion method. The validity of missing data techniques needs further study, and the reporting of missing data in published research also needs to be better established.

Key words: longitudinal study, missing data, missingness mechanism, missing data technique