In 2016, Brian Wansink wrote a blog post that prompted scientific sleuths to investigate his work. They found evidence of data manipulation, and, after several news articles and two investigations by his institution, he would found to have committed misconduct, as defined by Cornell University. His work had been used to inform US policy around food, much of which has now been thrown into question. Cases like this are rare, but science is not immune to misconduct. The rise in 'paper mills' — organisations that produce questionable or fake papers that they sell authorships on — has led some to worry that misconduct is on the rise and that a proportion of the scientific literature cannot be trusted. In episode two of Self Correction, we explore how researchers are responding to the problem of research misconduct. We discuss how difficult it is to determine the prevalence of misconduct, and how sleuths, journalists and research integrity institutions are fighting back. This episode was written and produced by Nick Petrić Howe. Dan Fox was the editor. The music was provided by Triple Scoop Music. Subscribe to Nature Briefing, an unmissable daily round-up of science news, opinion and analysis free in your inbox every weekday. Hosted on Acast. See acast.com/privacy for more information.