部分线性面板数据模型中个体效应的Bootstrap Hausman检验

对于面板数据模型而言,为了提高估计和预报的精度,区分个体固定效应和个体随机效应非常重要。 针对部分线性面板数据模型,在小样本下采用参数 Hausman 检验方法来识别模型中的个体效应,并通过 Bootstrap 抽样方法求得统计量的上分位点进而构造假设检验的拒绝域。 模拟结果显示,此检验的稳健性和可靠性比现有的非参数检验高,且在计算量和计算时间上有较大优势。

The Bootstrap Hausman test in partially linear panel data models with individual effects

In panel data models, in order to improve the accuracy of estimation and prediction, it is very important to test the null hypothesis of random effects versus fixed effects. This paper proposes a parametric Hausman test for partially linear panel data models with individual effects, and adopts the bootstrap-sampling-method to obtain thequantiles of the test statistics and subsequently construct the rejection regions. Simulations indicate that the Bootstrap Hausman test performs better than the nonparametric Hausman test proposed by Henderson, and furthermore the Bootstrap Hausman test has a great advantage in terms of computation.