A new analysis from the Journal of the American Medical Associations suggests that studies that have very large findings tend to be the most unreliable. They’re small and usually the first to measure the result that’s often not replicated in follow-up studies.
Unlimited access to BostonGlobe.com for only 99 cents for the first 4 weeks.Sign up
Are you a home delivery subscriber?
Get FREE access as part of your print subscriptionStart Here
Contact us for help