interesting article about the falibility of many scientific studies

http://arstechnica.com/science/news/2010/03/were-so-good-at-medical-studies-that-most-of-them-are-wrong.ars