A recent medicine study called attention to exaggerations of findings in abstracts. Do we have the same problem in soc?