The bootstrap was introduced by Bradley Efron (1979) more than two decades ago, mainly to calculate confidence intervals for a parameters in situations where standard methods were not applicable (Efron and Gong, 1983). An example of this would be a situation where the number of observations is so small that asymptotic results are unacceptably inaccurate. Since its invention, the bootstrap has seen many more applications and has been used to solve problems which would be too complicates to be solved analytically (Hall, 1992; Efron and Tibshirani, 1993).