Evaluating credibility of scientific studies