Estimating the proportion of a categorical variable with probit regression

Abstract

This article discusses the estimation of a population proportion, using the auxiliary information available, which is incorporated into the estimation procedure by a probit model fit. Three probit regression estimators are considered, using model-based and model-assisted approaches. The theoretical properties of the proposed estimators are derived and discussed. Monte Carlo experiments were carried out for simulated data and for real data taken from a database of confirmed dengue cases in Mexico. The probit estimates give valuable results in comparison to alternative estimators. Finally, the proposed methodology is applied to data obtained from an immigration survey.