I have a table of data.
Column 1 Before Column 2 After
In some cases the After values are larger than the Before and vice versa.
If I calculate the difference between each Before/After pair some values are negative some are positive. I then calculate the mean difference and get a small value. Much smaller than I would expect looking at the difference between Before and After values. The problem is the positives and negatives are canceling each other out, it's the mean after all.
So I make a new column and calculate the absolute difference for each Before/After pair. I then calculate the mean difference and get a much more sensible (in my mind) value which is more representative of the actual difference.
Is there a better way I can express the variation between before and after values?
