In statistical analyses, the term “bootstrapping” refers to methodology that involves randomly sampling a population with replacement.  In this module, we’ll discuss some examples of bootstrapping and other stochastic sampling methods I frequently apply in my own analyses.

One of my most frequent uses of bootstrapping and stochastic sampling methods is when doing cross-validation of a statistical model, such as a linear regression model.

When fitting a model to data, an important, but too often overlooked issue is cross-validation; ensuring that the model that you fit to one data set not only fits that data well, but also has good predictive abilities for an equivalent, but separate (ie; independent) data set.  However, if all you have is the one data set, cross-validating the model poses a problem.

Linear regression

Best fit

AIC

AIC is a statistic

Example calculating AIC with R