|
Abstract:
|
This talk is intended for people who have completed at least one course in statistics, from introductory statistics on up. I'll give an overview of the basic ideas and methods of bootstrapping and permutation tests, with an emphasis on how statisticians can use these methods for communicating with others---teachers to help students understand concepts such as sampling distributions, standard errors, confidence intervals, and P-values and consultants to help clients understand statistical results. I'll show how to use the bootstrap for statistics such as an x intercept in regression or a difference in trimmed means where classical methods are hard. And I'll show examples where the bootstrap is more accurate than classical methods because it doesn't require assuming normality or equal variances; in one example, classical t-tests give P-values that are off by a factor of four!
|
