I think it's because stats is usually not personal. When you actually know someone whom, say, homeopathy helped - now that's a good reason to believe.