Here is how a typical scientific journal study works

1 Choose a alpha value and if it will be a one or two tail t test

2 Collect data for two groups

3 Calculate mean, standard deviation and sample size for both groups

4 based on 1 & 3 decide if groups are statistically significantly different, PRESUMING they are both normally distributed

5 Never publish raw data showing they are normally distributed

This is sneaky because the normal distribution makes people look more "equal" than Pareto