That procedure might make the math work, but really it's fudging the numbers of the original return to get a desired result. Is this really a recommended procedure?
