5. Discover the Bootstrap technique

The bootstrap method is useful for evaluating confidence intervals for estimated parameters, particularly when the parent population is not normal. This method consists of simulating new possible samples from an initial sample, by drawing randomly with a discount, $n$ values in $n$ values available. This operation is repeated a large number of times (usually between 50 and 200), and each time the parameter of interest is calculated (mean or standard deviation, for example, see Example of results using the Bootstrap method). The result is a so-called "bootstrap" sample, whose histogram is an estimate of the true distribution of the estimator, enabling us to determine a confidence interval for the...