Evaluation of the power of Goodness-of-Fit tests for the comparison of data distributions

A comprehensive study has been performed to provide general guidelines for the practical choice of the most suitable goodness-of-fit test in real data analysis. Quantitative comparisons among the goodness-of-fit tests contained in the Statistical Toolkit are presented; the case of samples drawn from smooth theoretical functions is taken into account. This study is the most complete and general approach so far available to characterize the power of goodness-of-fit tests for the comparison of two data distributions. The results provide guidance to the user to identify the most appropriate test for his/her analysis on an objective basis.