Estimation and testing for partially linear additive varying-coefficient quantile regression with missing data

Résumé fourni par la source

This paper studies estimation and testing for partially linear additive varying-coefficient quantile regression with missing data at random. We use B-splines to estimate the non-parametric varying-coefficient functions, and employ the inverse probability weighting method to deal with the bias caused by missing data. Under some mild conditions, we establish the asymptotic properties of the proposed estimators, which provide theoretical support for the reliability and accuracy of the proposed estimation method. Moreover, we propose a rank score test for hypothesis testing, including the significance test of linear coefficients and the constancy test of varying-coefficient functions. Finally, the proposed estimation method is further illustrated by simulation studies, which demonstrate the finite sample performance of the method, and is applied to an empirical analysis.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.