The article compares how well six different tests can tell if data comes from a normal distribution. They used computer simulations with four types of skewed normal distributions to see which tests were best at detecting non-normality. The results show which tests are most powerful for this task.