Previously I simulated skewed distributions for normality testing using a uniform distribution and normality testing performed badly in detecting normal and non-normal data. I have now changed this to a normal distribution of skew values to see how this affects testing performance.

Looking at data to see what your next read should be!