Avoid lost discoveries, because of violations of standard assumptions, by using modern robust statistical methods

Modern robust methods can make a practical difference in detecting and describing differences between groups and associations between variables. Such procedures should be applied more frequently when analyzing trial-based data.