When reading a research paper, can you be certain that the institution the author claims to be affiliated with is actually the institution that was responsible for supporting the research? In this post Vivienne C. Bachelet presents findings from a recent study suggesting that a significant proportion of author affiliations are unverifiable. Highlighting how a lack of editorial guidance in this area has presented an opening for institutional malpractice, she argues that now is the time to more clearly regulate how authors attribute their research to different research organisations.
When reading a scholarly article, have you ever wondered whether the reported affiliations in the author byline were real or not? Most likely you have not, as doubting an author’s affiliation would breach the trust inherent to academia and scholarly communication. However, as research has become increasingly globalised over the past 30 years, the landscape of author affiliation has become more complex, with ‘multiple affiliations’ becoming the new normal.
Multiple affiliations occur when an author belongs to more than one organization that has contributed to a research project. According to the 6th edition of the publication manual of the American Psychological Association, a dual affiliation may be included if “the two institutions contributed substantial support to the study”. So, let’s say author A is reporting the results of a collaborative research project that involves a university and a research centre. She will, of course, report both affiliations because both have contributed substantially.
However, what if authors – or more accurately a proportion of authors – are systematically reporting affiliations to universities that have not contributed to the research? Should we be surprised, or concerned? You might argue that these situations are what journal editors are for; to check the truthfulness of the metadata associated to a scholarly publication, just like they send the body of the manuscript out for peer review. Well, no. There are no guidance documents or recommendations for editors on how to report affiliations in the author byline, and editors do not consider themselves enforcement officers over what authors submit in the same way they acknowledge their role in upholding the scientific record.
It was in the context of this discussion, that we came across evidence of an author who was publishing in high-impact journals with the name of a Chilean university in exchange for a significant fee without having any ties to that university. Taking Chile as a starting point, we decided to explore the prevalence of non-verifiable Chilean institutional affiliations amongst multiple-affiliated authors. To do this, we conducted an exploratory case study on the whole population of authors with multiple affiliations, one of which was to a Chilean university, included in the Scopus database for 2016. We have published the protocol of the study and, more recently, the full results. We used Google to find the institutional websites that would allow us to corroborate whether reported affiliations were in effect real. We found, shockingly, that 38% of the authors had no publicly available record that could establish a link with the reported university, affecting 40% of the included articles. Private, for-profit universities had a greater proportion of potentially misrepresented author affiliations (40%) compared to private, not-for-profit (28%) and public, State-owned (26%), with unverifiable affiliations occurring mainly in the health sciences and physical sciences; the social sciences were the least affected by this phenomenon. If this potential misconduct is found to be present in more countries, the implications may be far-reaching.
Firstly, university ranking systems use productivity, to a greater or lesser extent, as one of the indicators to rank universities. Consequently, if universities are tweaking the indicator, the result is a distortion of a system used worldwide to compare and assess universities. The effects are probably most pronounced in the middle of the ranks, where it is difficult to stand out from the pack and where there are institutions without the research facilities, scientists, and external grants that take years to set up, but with huge cash flows. These institutions may be tempted to effectively ‘buy’ their way up the ranks by offering honoraria, thereby adding papers to their name, a proportion of which may have resulted from the efforts of other universities, and thus free riding on the efforts, traditions and research systems of others.
Secondly, many universities are mandated to meet national accreditation standards, which, of course, include productivity as one of the key indicators. Cash-flow intensive universities can respond with relative ease to quality indicators on infrastructure but may need to boost indicators such as productivity by using bypass mechanisms like the one we describe in our paper. Furthermore, at least in the case of Chile, accreditation is a prerequisite for state-guaranteed student loans, essential when trying to massify higher education by including first-generation students from lower and middle classes, a way of expanding the market for private, for-profit universities.
Thirdly, many countries provide public funds to universities based on the number of high-impact papers they publish. Once again, misrepresented affiliations will benefit universities who are piggybacking on the efforts of others. Likewise, when experience is considered, grants are also awarded based on the number of publications, among other factors.
So, our exploratory research did confirm that we may be facing a big problem of institutional affiliation misrepresentation, though we cannot conclude that all 978 of 2,583 unique authors of the population of our study are misrepresented. We can say that either a) universities are not doing the job regarding transparency and accuracy of their websites or b) there is an emerging market of universities trying to spuriously pump up their productivity.
What is most alarming is that up to now, and to the best of our knowledge, no one has called out the emperor’s new clothes. Journals should be made aware that potentially one-third of reported affiliations are not truthful, a finding to be confirmed with further multi-country studies. To address this, ranking houses could punish universities who breach research integrity standards by tweaking indicators, much like how journals found harbouring citation cartels were punished by being stripped of their impact factor. International neutral bodies like COPE (Committee on Publication Ethics) or the ICMJE (International Committee of Medical Journal Editors) could also act and start working on guidance documents and specific recommendations on the proper and ethical reporting of author affiliations. Scientific communication relies on public trust so, when this trust is questioned, we need to find solutions without delay.
This post draws on the author’s co-authored article, Misrepresentation of institutional affiliations: The results from an exploratory case study of Chilean authors, published in Learned Publishing.
About the author
Vivienne C. Bachelet is an MD with an MSc in Clinical Epidemiology. She is Associate Professor in the School of Medicine of the Universidad de Santiago de Chile (USACH), where she teaches evidence-based medicine and does research on publication ethics, research integrity, reporting guidelines, methods and health policies. Vivienne is also founder and editor-in-chief of Medwave, a medical journal based in Santiago, Chile. Vivienne just ended a three-year term in COPE council in September, 2019.
Note: This article gives the views of the author, and not the position of the LSE Impact Blog, nor of the London School of Economics. Please review our comments policy if you have any concerns on posting a comment below.