Bootstrap Computation using R

By Lyra Filiola

Bootstrap is a computational method based on computer. Development of computer technology makes bootstrapping data becomes much easier than ever. Bootstrap is useful to estimate of statistics such as confidence intervals and the significance level of these statistics is depending on the number of bootstrap replications. Bootstrap method is especially useful for non parametric statistics because bootstrap method does not require any assumption about distribution of sample. This tutorial demonstrates how to use bootstrap using free statistical software package R. I provide several scripts in R programing language that use Bootstrap method to estimate the accuracy of mean of sample, and to obtain confidence interval for regression estimator using correlation model or bootstrap paired for both simple and general linear regression.

If you are new to Bootstrap, it is recommended that you first read Kardi Teknomo's Bootstrap tutorial. The content of this tutorial is as follows

• Very brief Introduction to R
• Bootstrap Statistics for Mean
• Bootstrap for Simple Linear regression
• Bootstrap for General Linear regression

This tutorial is copyrighted.

Preferable reference for this tutorial is

Filiola, L., (2006) Bootstrap Computation using R, http://people.revoledu.com/kardi/tutorial/Bootstrap/Lyra/index.html