Parameter Estimation of Multiple Linier Regression by Comparing Bootstrap and Jackknife Methods

Abstract

This research applies parameter estimation of multiple regression analysis using bootstrap and jackknife resampling methods. bootstrap resampling method is a resampling procedure that draws samples repeatedly randomly with returns. While the jackknife resampling method is a resampling procedure by performing calculations that remove one or more observations from the specified sample. Based on the data in this study, the bootstrap resampling method is better than the jackknife resampling method. This can be seen from the small bias and standard error values in the bootstrap resampling method compared to the jackknife resampling method. Similarly, the bootstrap resampling method's confidence interval is narrower than the jackknife resampling method. In addition, the parameter estimation coefficients between the least squares method and the bootstrap resampling method are almost similar. This shows that the bootstrap resampling method is better than the jackknife resampling method in this study.