Very, very interested in this article about the "decline effect" in (social and natural) science: phenomena that seem impressive in their first publication seem to become weaker when measured again in later experiments.