IMPUTING RESPONSES THAT ARE NOT MISSING

Wolfgang WEFELMEYER
University of Köln
Köln, Germany

Ursula U. MUELLER
Bremen, Germany

Anton SCHICK
Binghamton, Germany


ABSTRACT

We consider estimation of linear functionals of the joint law of regression models in which responses are missing at random. The usual approach is to work with the fully observed data, and to replace unobserved quantities by estimators of appropriate conditional expectations. Another approach is to replace all quantities by such estimators. We show that the second method is usually better than the first. Similar results hold for time series regression.



Back to the Workshop's Program