The decline effect and the scientific method : The New Yorker

The decline effect in empirical testing, why some studies can’t be reproduced.