High Impact = High Statistical Standards? Not Necessarily So