What Is Meant by "Missing at Random"?

被引:135
作者
Seaman, Shaun [1 ]
Galati, John [2 ,3 ]
Jackson, Dan [1 ]
Carlin, John [4 ,5 ]
机构
[1] MRC Biostat Unit, Cambridge, England
[2] La Trobe Univ, Murdoch Childrens Res Inst, Clin Epidemiol & Biostat Unit, Bundoora, Vic 3086, Australia
[3] La Trobe Univ, Dept Math & Stat, Bundoora, Vic 3086, Australia
[4] Murdoch Childrens Res Inst, Clin Epidemiol & Biostat Unit, Parkville, Vic, Australia
[5] Univ Melbourne, Melbourne, Vic 3010, Australia
基金
英国医学研究理事会;
关键词
Ignorability; direct-likelihood inference; frequentist inference; repeated sampling; missing completely at random; IGNORABILITY; INFERENCE;
D O I
10.1214/13-STS415
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
The concept of missing at random is central in the literature on statistical analysis with missing data. In general, inference using incomplete data should be based not only on observed data values but should also take account of the pattern of missing values. However, it is often said that if data are missing at random, valid inference using likelihood approaches (including Bayesian) can be obtained ignoring the missingness mechanism. Unfortunately, the term "missing at random" has been used inconsistently and not always clearly; there has also been a lack of clarity around the meaning of "valid inference using likelihood". These issues have created potential for confusion about the exact conditions under which the missingness mechanism can be ignored, and perhaps fed confusion around the meaning of "analysis ignoring the missingness mechanism". Here we provide standardised precise definitions of "missing at random" and "missing completely at random", in order to promote unification of the theory. Using these definitions we clarify the conditions that suffice for "valid inference" to be obtained under a variety of inferential paradigms.
引用
收藏
页码:257 / 268
页数:12
相关论文
共 40 条
[1]  
[Anonymous], 2002, STAT ANAL MISSING DA, DOI [DOI 10.1002/9781119013563, 10.1002/9781119013563]
[2]  
[Anonymous], 1997, CHAPMAN HALL SERIES, DOI DOI 10.1201/9781439821862
[3]  
[Anonymous], 2009, BR MED J
[4]  
[Anonymous], 1974, THEORETICAL STAT
[5]   NORMAL LIKELIHOOD FUNCTIONS [J].
ANSCOMBE, FJ .
ANNALS OF THE INSTITUTE OF STATISTICAL MATHEMATICS, 1964, 16 (1-2) :1-19
[6]  
Clayton D., 1993, STAT MODELS EPIDEMIO
[7]   Analysis of longitudinal data with drop-out: objectives, assumptions and a proposal [J].
Diggle, Peter ;
Farewell, Daniel ;
Henderson, Robin .
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS, 2007, 56 :499-529
[8]  
DIGGLE PJ, 1994, BIOMETRICS, V50, P580
[9]  
EDWARDS A. W. F., 1970, J R STAT SOC B, V32, P196
[10]  
Fisher R.A., 1956, Statistical methods and scientific inference, V3rd