具体信息如下：
————————————————————————————————
Title (题目)：Subsample Ignorable Likelihood Methods for Regression with Missing Data

Speaker (报告人)：Roderick J. Little
                Department of Biostatistics, University of Michigan

Time (时间)：2:00PM,  Aug. 25th, 2010 （８月２５日周三，下午两点）

Place (地点)：Room 1303 Science Building #1, Peking University （北京大学理科一号楼1303教室）

Abstract (摘 要)：
We consider multivariate regression of one or more outcomes Y on covariates X, when there are missing values in X and also possibly Y. Two common approaches are (a) complete-case (CC) analysis, which discards the incomplete cases, and (b) ignorable likelihood (IL) methods, which base inference on the likelihood for the observed data, assuming the missing data are missing at random. Specific IL approaches include maximum likelihood, Bayesian inference, or multiple imputation of the missing values based on a Bayesian model (Little and Rubin 2002). IL approaches retain all the data, but CC analysis yields valid inferences when missingness depends on the missing X's but not Y, a missing not at random mechanism where IL methods are subject to bias. We propose subsample ignorable likelihood (SSIL) methods, which are ignorable likelihood methods applied to a subsample of observations that are complete on a subset of the incomplete regressor variables. Plausible conditions on the missing data mechanism are presented under which SSIL gives consistent estimates, but both CC analysis and IL methods are inconsistent. These conditions dictate the choice of the subsample. Extensions are outlined.

References：
Little, R.J.A. & Rubin, D.B. (2002). Statistical Analysis with Missing Data, 2nd Edition. New York: John Wiley.
Little, R.J. & Zhang, N. (2010). Subsample Ignorable Likelihood for Regression Analysis with Missing Data. Submitted for publication.

About the speaker（报告人介绍）：
Roderick J. Little教授, 美国密歇根大学公共卫生学院生物统计系教授、主任。同时担任密歇根大学统计系教授和社会研究院高级研究员。
Roderick J. Little教授的研究领域集中在三个领域：缺失数据的统计分析、复杂抽样设计与数据分析、以及统计应用（包括社会科学、医疗卫生、大型调查等领域）。他是当代统计学与应用统计学领域杰出的学者之一，尤其是缺失数据分析领域的国际顶尖学者，多次获得美国统计学会奖励。其主要著作是与D.B. Rubin合著的“Statistical Analysis with Missing Data”。Little教授是一位长期从事教学的国际大师，他教学深入浅出，注重应用，培养了许多当代著名学者，颇受学生的欢迎。