Asymptotic Inference with Incomplete Data
Authors:
- 5 April 2013
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Theory and Methods
- Vol. 42 (17) , 3174-3190
Abstract
This article discusses asymptotic theory for the maximum likelihood estimator based on incomplete data. Although much literature has implicitly assumed the basic properties of the estimator, such as consistency and asymptotic normality, it is hard to find their precise and comprehensive proofs. In this article, we first show that under MAR an estimator based on the likelihood function ignoring the missing-data mechanism is strongly consistent. The estimator is then shown to be asymptotically normal. When the data are NMAR and when the data are MAR without parameter distinctness, the consistency and the asymptotic normality are shown. Several examples are provided.Keywords
This publication has 15 references indexed in Scilit:
- Normal distribution based pseudo ML for missing data: With applications to mean and covariance structure analysisJournal of Multivariate Analysis, 2009
- Missing Data in Clinical Studies2007
- Analysis of Incomplete Multivariate Data1997
- A Course in Large Sample Theory1996
- Multiple Imputation for Nonresponse in Surveys1987
- Inference and missing dataBiometrika, 1976
- Linear Statistical Inference and its Applications1973
- Asymptotic Properties of Non-Linear Least Squares EstimatorsThe Annals of Mathematical Statistics, 1969
- Maximum Likelihood Estimates for a Multivariate Normal Distribution when Some Observations are MissingJournal of the American Statistical Association, 1957
- Note on the Consistency of the Maximum Likelihood EstimateThe Annals of Mathematical Statistics, 1949