"Internet Publicity of Data Problems in the Bioscience Literature Correlates with Enhanced Corrective Action"
Paul S. Brookes has published "Internet Publicity of Data Problems in the Bioscience Literature Correlates with Enhanced Corrective Action" in PeerJ.
Here's an excerpt:
Several online forums exist to facilitate open and/or anonymous discussion of the peer-reviewed scientific literature. Data integrity is a common discussion topic, and it is widely assumed that publicity surrounding such matters will accelerate correction of the scientific record. This study aimed to test this assumption by examining a collection of 497 papers for which data integrity had been questioned either in public or in private. As such, the papers were divided into two sub-sets: a public set of 274 papers discussed online, and the remainder a private set of 223 papers not publicized. The sources of alleged data problems, as well as criteria for defining problem data, and communication of problems to journals and appropriate institutions, were similar between the sets. The number of laboratory groups represented in each set was also similar (75 in public, 62 in private), as was the number of problem papers per laboratory group (3.65 in public, 3.54 in private). Over a study period of 18 months, public papers were retracted 6.5-fold more, and corrected 7.7-fold more, than those in the private set. Parsing the results by laboratory group, 28 laboratory groups in the public set had papers which received corrective action, versus 6 laboratory groups in the private set. For those laboratory groups in the public set with corrected/retracted papers, the fraction of their papers acted on was 62% of those initially flagged, whereas in the private set this fraction was 27%. Such clustering of actions suggests a pattern in which correction/retraction of one paper from a group correlates with more corrections/retractions from the same group, with this pattern being stronger in the public set. It is therefore concluded that online discussion enhances levels of corrective action in the scientific literature. Nevertheless, anecdotal discussion reveals substantial room for improvement in handling of such matters.