Retractions of scientific publications: responsibility and accountability.

Vedran Katavić

doi:10.11613/bm.2014.024

Abstract

This evidence-based opinion piece gives a short overview of the increase in retractions of publications in scientific journals and discusses various reasons for that increase. Also discussed are some of the recent prominent cases of scientific misconduct, the number of authors with multiple retractions, and problems with reproducibility of published research. Finally, some of the effects of faulty research on science and society, as well as possible solutions are discussed.

Highlights

I‘d like to start this opinion piece with a disclaimer – I am going to do something not entirely common in a scientific journal, and NOT cite those articles I mention in this manuscript! That is, I will not disclose the titles nor cite the articles that have been retracted or are being investigated for possible fraud
Most of the burden of keeping the record straight eventually falls onto the journal editors’ backs [6], and, the editors are not the scientific community’s policemen [7], they have responsibilities and roles, as well as tools at their disposal. These are best defined by the Council of Editors’ White paper on Publication ethics [8], which notes that editors can correct the public record by publishing either corrections, which “identify a correction to a small, isolated portion of an otherwise reliable article” or retractions, that “refer to an article in its entirety that is the result of pervasive error, nonreproducible research, scientific misconduct, or duplicate publication”
While the frequency of corrections has been constant throughout the various scientific fields [11], the frequency of retracted publications has dramatically increased even after correcting for the increase in total scientific publication output [9], while at the same time the “time-to-retraction” has significantly decreased [13]

Summary

Corrections and retractions in scientific journals

Corrections and retractions are an integral part of scientific publication because they constitute what can be described as the scientific method and ethical publishing – the burden of evidence dictates that what was thought to be right, or what was purported to be right has to yield to what is right, and, needs to be corrected in the public domain’s record keeping. Most of the burden of keeping the record straight eventually falls onto the journal editors’ backs [6], and, the editors are not the scientific community’s policemen [7], they have responsibilities and roles, as well as tools at their disposal These are best defined by the Council of Editors’ White paper on Publication ethics [8], which notes that editors can correct the public record by publishing either corrections (errata or corrigenda), which “identify a correction to a small, isolated portion of an otherwise reliable article” or retractions, that “refer to an article in its entirety that is the result of pervasive error, nonreproducible research, scientific misconduct, or duplicate publication”. The depressing take home message is, that it is safe to assume that the fraudulent research far outweighs the research that is retracted [18]

The stigma of correction and retraction

Authors with multiple retractions

Number of retracted publications

The solution?