Introduction to Bootstrap

Abstract

A fundamental issue in statistics is about assessing the variability of an estimate known as the standard error (SE) of the statistics of interest. Drawing different samples from the same population will yield different estimates; this hypothetical set of estimates represents the sampling distribution of the statistics. However, in real life, it is not possible to repeat the study many times. Before the widespread availability of fast computers, sampling distribution of a statistics was derived mathematically. The mathematical derivations usually require assumptions about the distribution of the data and apply only to certain statistics such as mean.