A PORTMANTEAU TEST FOR SERIALLY CORRELATED ERRORS IN FIXED EFFECTS MODELS
提出一种用于固定效应模型误差项序列相关的混合检验，该检验可解释为拉格朗日乘子检验或沃尔德检验，蒙特卡洛实验显示其具有良好的大小和功效性质。
We propose a portmanteau test for serial correlation of the error term in a fixed effects model. The test is derived as a Lagrange multiplier test, but it also has a straightforward Wald test interpretation. In Monte Carlo experiments, the test displays good size and power properties.The authors thank the co-editor, the referee, David Drukker, Christian Hansen, and Jeffrey Wooldridge for their helpful comments.