The only way to reliably test for comparison is to test Avira free and Pro side-by-side at the very same time using the very same files. That would take two separate systems and testing performed on each product simultaneously.
There are very minor differences between free and Pro, that result in differences in detection that are not statistically relevant. A 1 or 2 % difference is essentially meaningless.