Discussion of Paper "Fitting regression models to case-control data by maximum likelihood" by Scott and Wild (1997)
2010-04-23 12:30 - 14:30
Room 405, Mathematics Research Center Building (ori. New Math. Bldg.)
This paper considers prospective regression models for a categorical response with case-control data supplemented by known population totals in each response category. Under the supplemented case-control sampling, the authors proposed an approach for deriving the maximum likelihood estimator of the parameters of interest. The main contribution of their estimation procedure is to produce maximum likelihood estimates by iteratively updating the obtained pseudo-likelihood estimates.