Sample Design for Analysis using High-Influence Probability Sampling

Abstract Sample designs are typically developed to estimate summary statistics such as means, proportions and prevalences. Analytical outputs may also be a priority but there fewer methods results on how efficiently design samples for the fitting estimation of statistical models. This paper develops general approach determining efficient sampling probability-weighted maximum likelihood estimators considers application generalized linear We allow non-ignorable sampling, including outcome-dependent sampling. The new have probabilities selection closely related influence dfbeta Cook's distance. is shown perform well in simulation based data from New Zealand Health Survey.