ABSTRACT We investigate inverse-probability-weighted (IPW) maximum likelihood estimation in zero-inflated binomial regression with missing-at-random covariates. Large sample properties (consistency, asymptotic normality) of the IPW estimator are established. Finite sample properties are assessed via simulations. The methodology is illustrated on a real data set.

中文翻译：

---

缺少协变量的零膨胀二项式回归中的估计

摘要 我们研究了具有随机缺失协变量的零膨胀二项式回归中的逆概率加权 (IPW) 最大似然估计。建立了 IPW 估计量的大样本属性（一致性、渐近正态性）。通过模拟评估有限样本属性。该方法在真实数据集上进行了说明。