We often encounter that data generate does not fulfill the assumptions of a parametric test, for example, ANOVA. In that case, we need to transform data by using different methods like square root, inverse, log10, etc. While doing so, it is tedious to try out all methods until data is approximately normal (Or if no transformation methods normalizes then conduct equivalent non-parametric approaches). Therefore, it would be easier if we can follow a systematic method. Hence, I want to request your esteemed self to provide your invaluable input. I would be grateful if relevant literature is provided. Thank you.

Similar questions and discussions