What research actually asks of you

Two studies can run the identical test on the identical data and only one of them survives review. The difference is never the arithmetic.