On the Perception of Small Differences