An affiliation tells us something, but how much?
Open almost any research paper and you will find the authors’ names followed by their institutional affiliations. Some authors list one university, others list two or three, and occasionally you will find researchers associated with even more institutions.
We tend to accept these affiliations without giving them much thought. If somebody lists a university, we generally assume that they have some formal relationship with that institution. But what exactly is that relationship, and what does it tell us about the research?
Consider a researcher who lists four universities. They might be employed full-time by one, teach part-time at another, hold an adjunct appointment at a third and be an emeritus professor at the fourth. All four relationships might be entirely genuine, but somebody reading the paper probably has no way of knowing what those affiliations actually represent.
Does that matter? I think it might, particularly when we consider how much importance universities, publishers and researchers place on the accuracy and transparency of the scholarly record.
What does an affiliation actually mean?
There are many legitimate reasons why somebody might have more than one institutional affiliation.
An academic might hold a full-time position at one university while contributing to teaching, research or doctoral supervision at another. They might have an honorary or visiting appointment, maintain an emeritus relationship with a former employer, or be working across institutions as part of a collaborative research project.
There is also the question of timing. Research can take several years, and an academic might move institutions while a project is underway. Listing both universities could be entirely appropriate, particularly if they both contributed to the research.
None of these arrangements is inherently problematic. Indeed, universities benefit enormously from the expertise of researchers who are not necessarily employed by them on a full-time basis. An adjunct academic might supervise doctoral students, contribute to research projects, mentor staff, teach specialist courses or provide access to international research networks. An emeritus professor may continue to be one of the institution’s most productive researchers.
It would be wrong to suggest that an affiliation is only legitimate when somebody is a full-time employee. However, these examples highlight the underlying problem. An institutional affiliation, in most cases, tells us very little about the nature of that association.
Is having an appointment enough?
Suppose I am employed full-time by University A and have an adjunct appointment at University B.
I undertake a research project entirely at University A. The research is funded by University A, its facilities are used, and the work is undertaken with colleagues employed there. University B is not involved in the research.
When I submit the resulting paper, should I list both universities as my affiliations?
There are several possibilities. I might argue that I hold legitimate appointments at both institutions and should therefore be entitled to list them. University B might even encourage, or perhaps require, its adjunct academics to acknowledge their affiliation whenever they publish.
But should simply holding an appointment be sufficient? What if I have adjunct appointments at five universities? Should I list all five on every paper, even when some of those institutions have had no involvement in the research being reported?
I am not suggesting that somebody who does this is necessarily behaving dishonestly. There may be contractual expectations, longstanding research relationships or other legitimate explanations. Nevertheless, I think there is an important distinction between having a genuine institutional appointment and having a justification for including that institution on a particular publication.
An affiliation can be factually correct without necessarily giving readers an accurate impression of the institution’s involvement in the research.
What do publishers and research integrity guidelines say?
There is already some recognition of this issue.
The International Committee of Medical Journal Editors (ICMJE) states that funding sources and affiliations which have not contributed to the work should not be listed because doing so would be misleading.
That is quite a significant statement. It suggests that having a relationship with an institution is not, by itself, sufficient justification for including it on a publication.
Nature Portfolio’s authorship policy also provides guidance, stating that the primary affiliation should be the institution where the majority of the researcher’s work was undertaken. Meanwhile, Elsevier’s publishing ethics policy expects institutional names to be reported in a standard form so that the institutions can be independently verified.
These are useful principles, although they do not fully resolve the question I am raising. A university might be able to confirm that somebody holds an adjunct appointment. That establishes the legitimacy of the relationship, but it does not necessarily explain what the appointment involves or whether the institution contributed to the research being published.
There is also evidence that verifying affiliations can be difficult.
In a 2019 study published in Learned Publishing, researchers examined Scopus-indexed publications from 2016 involving authors who reported multiple institutional affiliations, including at least one Chilean university.
The study examined 4,961 author records involving multiple affiliations. The researchers were unable to verify 38% of the affiliations to Chilean universities using information available on institutional websites. This affected 40% of the articles examined.
Those findings need to be interpreted carefully. The research concerned publications from a particular year and national context, and an affiliation that cannot be verified through a university website is not necessarily incorrect. Nevertheless, the study raises an important question about how easily institutional affiliations can be checked and whether the current arrangements provide sufficient transparency.
We already expect other forms of transparency
Academic publishing has become increasingly concerned with transparency. Authors are expected to disclose funding sources and potential conflicts of interest. Many journals also now require authors to declare their use of artificial intelligence, although the requirements vary between publishers.
The CRediT contributor taxonomy provides a standardised way of identifying different contributions to research, including conceptualisation, methodology, data analysis, supervision and writing. These developments recognise that simply listing somebody as an author does not tell readers everything they might reasonably want to know about their involvement.
Could we apply a similar principle to institutional affiliations?
If it is important to explain what an author contributed to a paper, is it not also reasonable to explain what their relationship is with the institutions they list?
Of course, these are different matters. Authorship declarations describe contributions to the research, while affiliations identify institutional relationships. We should not confuse the two, but perhaps that distinction strengthens the argument for greater transparency rather than weakening it.
Would an affiliation declaration help?
I am not proposing that every researcher should have to publish their employment contract, salary arrangements or detailed institutional responsibilities. Nor do I think we should create a system that regards full-time employment as somehow more legitimate than an adjunct or honorary appointment.
Instead, I wonder whether a short affiliation declaration could provide useful additional information.
For example, an author might indicate that University A was their full-time employer when the research was undertaken, while University B was an adjunct appointment through which they contributed to a collaborative research project.
An emeritus professor might identify their emeritus status and explain that they continue to participate in the university’s research activities. A researcher who changed institutions might explain that the research was undertaken at one university while the manuscript was completed at another.
This would not require a lengthy statement. It could potentially be part of the manuscript submission process through a small number of standard categories, with an option to provide additional information where appropriate.
Importantly, I think two questions would need to be addressed. What is the researcher’s relationship with the institution, and what was that institution’s connection to the research being reported?
The answers would not always be straightforward, and we would need to avoid introducing unnecessary complexity into an already demanding publication process.
There is also the question of verification. An author could declare that they hold an adjunct appointment, but who would confirm that the appointment exists? Should publishers seek confirmation from universities, or would an author declaration be sufficient? Additional transparency would be useful, but it would not automatically eliminate misleading affiliations.
Is more information always better?
There is a reasonable counterargument to everything I have suggested.
Academic publishing already places substantial administrative demands on researchers. Every additional declaration, form and reporting requirement takes time, and the benefits should justify that effort.
Perhaps readers do not need to know whether an author is full-time, part-time, adjunct or emeritus. Perhaps all that matters is that the affiliation is genuine and consistent with the journal’s requirements. I can see some merit in that argument, particularly if the alternative is another complicated reporting process that provides little useful information.
However, there is a difference between information being unnecessary and not being available.
Institutional affiliations form part of the published scholarly record. They are used to connect researchers with organisations, establish research relationships and attribute publications to institutions. Given that importance, I think it is reasonable to question whether the information currently provided is sufficient.
We should certainly avoid assuming that a researcher with several affiliations is doing anything improper. But equally, we should not assume that every affiliation conveys the same kind of institutional relationship. Perhaps we have become so accustomed to seeing university names beneath authors’ names that we have stopped asking what those names actually mean.
The issue is not whether researchers should be allowed to have multiple affiliations. I think there are excellent reasons why they should, and there are many circumstances where those affiliations reflect genuine and valuable academic relationships.
The issue is whether the scholarly record should tell us more about the relationships than it currently does.
When researchers list a university as an affiliation, should they also be required to explain the nature of their relationship with that university?
I publish two free newsletters each week
Governance, Leadership and Research
A weekly brief for higher education leaders and researchers examining governance, leadership and research in practice.
https://buff.ly/GhsZqVl
Publishing with Integrity
Examining scholarly publishing through an ethical lens, challenging assumptions and supporting academic career development.
https://buff.ly/WiG1yFb
If you subscribe, you will be alerted whenever I publish a new edition.
About the author
Professor Graham Kendall has held senior leadership positions in higher education, including Vice-Chancellor, Provost/CEO, Deputy Vice-Chancellor, Pro-Vice-Chancellor and Vice-Provost roles. He is an Emeritus Professor at the University of Nottingham and has published more than 300 peer-reviewed papers. His interests include university leadership and governance, research strategy, artificial intelligence and research integrity.