Assumption of normality means that you should make sure your data roughly fits a bell curve shape (Normal Distrubution) before running certain statistical tests or regression