Post publication discussion and corrections
On October 25th, Retraction Watch released a database of 18,000 articles and conference papers which had been retracted since the 1970s. An analysis of the database was published in the journal Science. This concludes that the increase in retractions in recent years might reflect better editorial practices. Not all retractions involve misconduct, 40% of the retractions in the database are due to errors, problems with reproducibility, and other issues. Around 50% of the retractions are due to “fabrication, falsification, or plagiarism” with an additional 10 percent due to other unethical practices.
An Editorial in JAMA examines the issues of fabrication, falsification, and plagiarism in the reporting of medical research, and the roles and responsibilities of Editors and Institutions in dealing with these challenging problems.
Professor Iekuni Ichikawa, interprets the Retraction Watch Leaderboard, and the fact that some countries, journals and fields appearing more regularly on it might in fact reflect their more vigorous attempts to uncover misconduct.
And finally on the subject of misconduct, two articles from Nature look at blacklists. The Chinese government has proposed a blacklist of journals to deter researchers from misconduct. It has not released its selection criteria as yet. A second article discusses the pros and cons of such blacklists.
Ethical Oversight
A new Code of Conduct for Research Integrity was put into place in the Netherlands on 1st October. It replaces the previous Code which has been in use since 2004 and requires the Institution to foster research integrity through empowering scientists to maintain the principles and standards of responsible research practice. A Nature article discusses this in some detail.
Also on the subject of how institutions can improve their ethical oversight, this article looks at what institutions learn from the aviation industry to ensure that they examine their structures, processes and cultures identify and change problematic behaviours which lead to bullying.
Conflicts of Interest
In this article in JAMA, Jeffery Botkin argues that Institutions should strengthen their COI policies by considering that failure to disclose significant financial relationships relevant to the conduct of research is research misconduct.
Also on Conflicts of Interest, a study by the Nordic Cochrane Centre in Copenhagen suggests that data analysis of clinical trials is often carried out by the funders of the research, and this involvement is not disclosed.
Journal Management
In early October, the Coalition for Diversity and Inclusion in Scholarly Communications (C4DISC) was launched to discuss and address issues of diversity and inclusion within scholarly communication. A discussion of the origins, supporters, principles and plans of the organisation is here.
Also launched in October was the 2018 STM report which gives a comprehensive overview of the scholarly publishing industry worldwide.
Data and replicability
Vlaeminck and Podkrajac examine whether economics journals enforce their data policies. Their analysis of papers published in 2013 and 2014 found that not many did at that point. They encourage editors and reviewers to be stricter in enforcing their journal data policies.
A first analysis of articles which register protocols before research is conducted suggests that they are more likely to report null findings and thus help to correct the publication bias towards positive findings.
COPE Council member Deborah Kahn
Read November Digest: Post-Publication Discussions and Corrections