Bootstrapping, or the bootstrap, is a statistical methodology that is frequently used in situations where standard distributional assumptions, such as normality, do not hold. In addition, the bootstrap can be used to estimate standard errors and confidence intervals for parameter estimates. It is particularly useful where there is not a known sampling distribution for the statistic of interest, [Page 218]thereby making calculation of standard errors difficult or impossible. There are a number of variations in the bootstrap that make it useful in a wide variety of situations. Regardless of context or application, the bootstrap is based upon a basic framework of resampling with replacement from the original sample. This entry discusses the basic nonparametric bootstrap, bootstrap confidence intervals, variations in the bootstrap, and when to ...
Looks like you do not have access to this content.