3 p.m. - 4 p.m. Location: FN 2.106
Ursula U. Mueller
Texas A&M University
Efficient estimation in regression with missing responses
I will first review some of my results on efficient estimation in semiparametric regression with responses Y "missing at random" based on imputation. Then I will demonstrate that characteristics of the conditional distribution of Y given the covariate vector X can be estimated efficiently using complete case analysis, i.e. one can simply omit incomplete cases and work with an appropriate efficient estimator which remains efficient.
The "efficiency transfer" is a general result and holds true for all regression models for which the distribution of Y given X and the marginal distribution of X do not share common parameters. The derivation uses the "transfer principle" for obtaining limiting distributions of complete case statistics (for general missing data models) from corresponding results in the complete data model. For an illustration we consider estimation of regression parameters and of functionals of the error distribution.
Sponsored by the Department of Mathematical Sciences