Comparisons between Resampling Techniques in Linear Regression: A Simulation Study

The classic methods used in estimating the parameters linear regression need to fulfill some assumptions. If assumptions are not fulfilled, conclusion is questionable. Resampling one of ways avoid such problems. study aims compare resampling techniques regression. original data clean, without any influential observations, outliers and leverage points. ordinary least square method was as primary estimate then compared with techniques. variance, p-value, bias, standard error a scale best among random bootstrap, residual bootstrap delete-one Jackknife. After all analysis took place, it found that did perform well while Jackknife works quite well. Random better than square. Is parameter small sample. At same time, suggested use when sample size big because more accessible apply big.