Better incentives could reduce the alarming number of studies that turn out to be wrong when repeated

— Read more on ScientificAmerican.com