Il metodo del bootstrap nella propagazione degli errori

Abstract

This note illustrates, through a very simple example, the power and flexibility of the bootstrap method to assess in a straightforward and unambiguous way how the random errors in direct measurements propagate themselves into the derived quantities of interest. Furthermore, a version of the bootstrap with the initial data set generated by Monte Carlo has proved to be suitable in situations where the standard deviations of the directly measured quantities may be considered as known, but the effect of various experimental errors on the accuracy of final results is quite obscure.