The Hypothesis Testing For Regression Coefficients In Panel Data Models

In this article, the hypothesis testing of regression coefficients in balanced and unbalanced one-way error component regression models are considered.In first chapter, the research status and required theoretical results of the Panel data models are presented.The second chapter, the problem of hypothesis testing for regression coefficients in balanced Panel data model are discussed. For these problems, a parametric bootstrap (PB) approach is proposed. Simulation results indicate that the PB test, regardless of the sample sizes, can control Type I error rates and powers satisfactorily, whereas the generalized variable test may far exceed the intended level when the sample sizes are small or moderate.The third chapter, the problem of hypothesis testing for regression coefficients in unbalanced Panel data model are studied. For these problems, a parametric bootstrap (PB) approach is proposed based on the ANOVA estimation and Bootstrap theories. Simulation results indicate that the PB test can perform well in terms of maintaining the intended level and power..