Joshua Hansen

Research often starts with asking a question and gathering data related tothat question. Once we have data, what can we say about the parameters ofa general population? This is the question Bradley Efron tried to answer ashe formulated the bootstrap method, which is a resampling method thatgives confidence intervals of parameters from a population. While thismethod can be used algebraically with no simulation, its most common useincludes simulation as this leverages the power of the computer, a toolwhich was just becoming readily available in Efron’s time. The bootstrapmethod is quite useful, but there are situations where it does not providesatisfactory confidence intervals. Efron set out to mitigate theseinconsistencies by creating the bias-corrected bootstrap method. This talkwill address the development of the bootstrap methods and also look to thefuture at further improvements that could be made.