Estimation in zero-inflated binomial regression with missing covariates

Alpha Oumar Diallo (1, 2) , Aliou Diop (1) , Jean-François Dupuy (2)

1 LERSTAD - laboratoire d'Etudes et de recherches en Statistiques et Développement
2 IRMAR - Institut de Recherche Mathématique de Rennes

The zero-inflated binomial (ZIB) regression model was recently proposed to account for excess zeros in binomial regression. Since then, the model has been applied in various domains, such as dental epidemiology and health economics. In practice, it often arises that some covariates involved in ZIB regression have missing values. Assuming that the missingness probability can be estimated parametrically, we propose an inverse-probability-weighted estimator of the parameters of a ZIB model with missing-at-random covariates. Consistency and asymptotic normality of the proposed estimator are established. A consistent estimator of the asymptotic variance-covariance matrix is also provided. The finite-sample behavior of the estimator is assessed via simulations.

Mots clés: count data; excess of zeros; Asymptotics; inverse-probability-weighting

Domaines: Méthodologie [stat.ME]

Alpha Oumar Diallo, Aliou Diop, Jean-François Dupuy. Estimation in zero-inflated binomial regression with missing covariates. Statistics, 2019, 53 (4), pp.839-865. ⟨10.1080/02331888.2019.1619741⟩. ⟨hal-01585220⟩