Abstract

Journals require authors to report their current affiliation, and any affiliation relevant to the submitted work. Institutional research productivity is usually associated to increased funding, higher ranking positions and greater admission applications. On the author side, publications are required in the ‘publish or perish’ dictum, and many universities provide extra honoraria for each published article that includes the affiliation. The American Psychological Association states that ‘the institutional affiliation identifies the location where the author or authors were when the research was conducted, which is usually an institution.’ It then goes on to recommend that a dual affiliation may be included ‘only if two institutions contributed substantial support to the study.’ To the best of our knowledge, this is the only known definition of affiliation. Neither the ICMJE nor COPE have issued recommendations or guidelines on what constitutes a properly reported affiliation.

We and others have observed that there are cases of incorrect or false reporting of institutional affiliations that might introduce a systemic bias in the current indicators that are used to rank universities and to rate researcher productivity. To determine the scope of this problem we conducted an exploratory case study in Chile on author affiliation misrepresentation for Scopus-indexed articles in 2016, the protocol of which we published recently.¹ In yet unpublished results from our study, we found that 38% of authors have multiple affiliations in at least one article that was not verifiable, and 40% of articles have at least one author in which it was not possible to verify the reported affiliation to a Chilean university. In 30% of author/article records for the year, we could not corroborate the reported affiliation to a Chilean university. The significant magnitude of the prevalence of unsubstantiated affiliations in this exploratory study underpins the need for a consensus on the definition of affiliation together with recommendations for both authors and journal editors. Aim To develop a consensus statement on the definition of author affiliation that will guide authors during the submission of their manuscripts and journal editors to identify potential breaches of research integrity. Method The development of this recommendation will be based on the Delphi method. We will first identify and define the key concepts of what an affiliation is following a broad literature review. Then we will develop an initial draft of the document that will be sent out for expert consultation and input (eg, ICMJE and COPE). Lastly, an iterative international survey with a broad-ranging roster of participants from the STEMM disciplines will be conducted until consensus is reached. The resulting document will be published and disseminated to key stakeholders, including university ranking systems, funders, editors, and research integrity officers.