Time: 2023/06/20(Tue.), 13:30 – 15:00(Beijing Time)

Title:    Specification Tests for Regression Models with Measurement Errors

Venue: Zhonghui Buidling 106

About the speaker:

宋晓军，北京大学光华管理学院商务统计与经济计量系副教授，博士生导师，西班牙马德里卡洛斯三世大学经济学博士。主要研究兴趣是理论计量经济学，包括非参数/半参数方法，假设检验和自助法，以及计量经济学的应用等。论文发表在Econometric Theory，Journal of Applied Econometrics，Journal of Business & Economic Statistics和 Journal of Econometrics 等国际期刊。主持和参加自然科学基金面上项目和国家重点专项等。自2020年1月起担任Economic Modelling副主编。

Abstract:

We propose a novel class of specification tests for regression models with measurement errors. The highlight of our method is the introduction of a projection device to eliminate the parametric estimation effect, which is a common issue in nonsmoothing type tests. The resulting projection-based tests have asymptotic null distributions that are invariant to the estimator used. Besides, the proposed method allows for a slower convergence rate of the parametric estimator. We suggest a convenient multiplier bootstrap method to simulate the critical values as accurately as desired. Finally, we evaluate the finite sample performance of the proposed tests through Monte Carlo simulations.