I'm more inclined to report reported differences to bad reporting of inputs. How many studies are conducted under controlled conditions? People lie about everything all the time, especially to themselves.