Normality Test Explorer

Live Script exploring tests of an empirical distribution for normality.

This Live Script explores methods to test if an empirical distribution such as a residuals distribution is consistent with normality. The script fits a simulated two-component Gaussian empirical distribution to various model Gaussian mixtures and examines use of the Akaike Information Criterion (AIC) and the Bayes Information Criterion (BIC) to select the best model. It also performs an Anderson-Darling test and a Kolmogorov-Smirnov test for normality.

This script may interest students and educators in physics and other STEM fields. Sliders are provided to adjust the empirical distribution parameters and examine the sensitivity of the tests in characterizing normality. 'Try this' suggestions are included for further exploration.