偏正态非平衡面板单因素随机效应模型的Bootstrap推断  

Bootstrap Inference for Skew-Normal Unbalanced Panel Model with One-Way Error Component

作　　者：叶仁道[1] 杜微晓 YE Rendao;DU Weixiao(School of Economics,Hangzhou Dianzi University,Hangzhou 310018)

机构地区：[1]杭州电子科技大学经济学院,杭州310018

出　　处：《系统科学与数学》2023年第1期244-270,共27页Journal of Systems Science and Mathematical Sciences

基　　金：国家社会科学基金(21BTJ068)资助课题。

摘　　要：针对偏正态非平衡面板单因素随机效应模型,文章研究了回归系数和方差分量函数的假设检验和区间估计问题.首先,基于矩阵分解技术,给出回归系数的精确检验方法.其次,利用Bootstrap方法和广义方法,构造单个方差分量、方差分量之和的检验统计量和置信区间.再次,建立方差分量之比的精确检验和近似检验.文章证明了所给检验方法和置信区间的变换不变性等理论性质.Monte Carlo结果表明,对于所设参数和样本量,文章所给方法在犯第一类错误的概率和功效意义下,具有统计优良性.最后,将上述方法应用于汽油消耗量的案例分析.For skew-normal unbalanced panel model with one-way error component,the hypothesis testing and interval estimation problems for the regression coefficients and variance component functions are considered.Firstly,based on the matrix decomposition technique,the exact test statistic for the regression coefficients is obtained.Secondly,the test statistics and confidence intervals for the single variance component and sum of variance components are constructed by the Bootstrap approach and generalized approach.Next,the exact test and approximate test for the ratio of variance components are established.We prove the theoretical properties such as the transformation invariance of these testing approaches and confidence intervals.The Monte Carlo simulation results indicate that for the set parameter configurations and sample sizes,the proposed approaches have excellent statistical properties in terms of the size and power.Finally,the above approaches are illustrated with a real example of gasoline consumption.

关 键 词：偏正态非平衡面板单因素随机效应模型 回归系数 方差分量函数 BOOTSTRAP 广义方法 

分 类 号：O212.1[理学—概率论与数理统计]