Evidence against “absolute” scaling