Half of social-science studies fail replication test in years-long projectnature.com7 points·nradov··1 commentOpen articleSaveView on HN