Abstract: An approach to organizing the calculation of the efficient score for a finite dimensional parameter of interest in the presence of an infinite dimensional nuisance parameter is presented. The approach involves choosing a set of submodels for the nuisance parameter. The condition that the scores for the submodels be orthogonal to the efficient score takes the form of an equation whose solution appears in a representation of the efficient score. The approach is illustrated with several examples.
Key words and phrases: Case-control, censoring, interval censored data, missing covariates, projection, tangent space.