Scientific Uncertainty in the Press: How Newspapers Describe Initial Biomedical Findings:

Newspapers preferentially cover initial biomedical findings although they are often disconfirmed by subsequent studies. We analyzed 426 newspaper articles covering 40 initial biomedical studies associating a risk factor with 12 pathologies and published between 1988 and 2009. Most articles presented the study as initial but only 21% mentioned that it must be confirmed by replication. Headlines of articles with a replication statement were hyped less often than those without. Replication statements have tended to disappear after 2000, whereas hyped headlines have become more frequent. Thus, the public is increasingly poorly informed about the uncertainty inherent in initial biomedical findings.