回归模型中变量遗漏偏差补救的面板数据方法

A Panel Data Method of Correcting the Omitted Variable Bias in Regression Models

建立计量经济模型时,忽视那些难于量化或无法获得观测数据的解释变量是一种最常见的处理方法,但这可能造成模型参数OLS估计量的变量遗漏偏差.一般的教材或研究文献没有谈及对其如何补救,该文提出一种采用面板数据和固定效应回归来对这种偏差进行补救的方法.

When forming econometric models, those variables which cannot be observed or measured are often omitted. It can lead to the omitted variable bias in the estimation of model parameter OLS. By using panel data and fixed effects regression models, we have developed in this paper a new method which can be used to correct the omitted variable bias in some regression models.