Credit for all

Group proposes new method for classifying each author’s role in a scientific article

032-035_Autoria_221-1Nelson ProvaziBy early 2015, a group of American and British researchers will propose a model for more precisely identifying how each author has contributed to a scientific article. Under development for the past two years, the initiative seeks to devise a classification system that can inform each author’s role, even in papers signed by hundreds of researchers. The outline of this “taxonomy,” which would be applied at the moment a manuscript is submitted online to a science journal, is currently under evaluation. It defines 14 forms of participation in article preparation (see table), such as the intellectual design of the study and the various stages of conducting the investigation and preparing the manuscript. The most controversial categories are those that assign credit for tasks devoid of any intellectual contribution to article production, such as project administration, data curation (managing collected information sources and making them accessible for subsequent use), and funding acquisition.

The preliminary outline was tested by the corresponding authors of 230 articles from the field of the life sciences that appeared in publications of the Nature, Elsevier, and PLOS groups or in the periodicals Science and eLife. About 85% of respondents found the taxonomy easy to use. In the opinion of 45%, author identification was more accurate than with the methods currently in use, whereas 37% felt there was no difference. Some authors suggested that certain categories should be further sub-divided, while others proposed that non-intellectual contributions to a paper be included elsewhere. The researchers working on the initiative, who have ties to Harvard University and the Wellcome Trust, a British foundation dedicated to supporting biomedical research, promise to refine the classification in the coming months, tailoring it to the needs of other fields of knowledge as well. Some categories may be added to the outline and others discarded. A workshop to discuss the changes has been scheduled for late 2014. Liz Allen, head of evaluation at the Wellcome Trust and one of the researchers involved in the proposal, reported on the endeavor in an article published in the magazine Nature in April 2014. “There is certainly more work to do,” she says.

The initiative was motivated primarily by the rise in the number of multiple author papers, in turn a product of the increasingly collaborative nature of research. According to data presented by Allen, from 2006 to 2010, the average number of authors signing articles linked to Wellcome Trust-sponsored studies rose from 10.21 to 28.82 in the field of genetics and from 6.28 to 8.32 overall. “In multicentric studies, research entails contributions by dozens of researchers, and authorship is quite diffuse. In some cases, the various research groups involved make quite distinct contributions, but the author list fails to reflect this,” says Abel Packer, director of FAPESP’s Scientific Electronic Library Online program (SciELO/FAPESP). Greater transparency, he points out, is vital so that research funders and the scientific community can identify who did what in a multi-author paper. The taxonomy is intended to overcome other problems as well, like the lack of a standard order for listing an article’s authors, with some disciplines putting the lead researcher first, others last, and still others using alphabetical order.

032-035_Autoria_221-2NElson ProvaziEditors of scientific journals have grown steadily more concerned with authorship issues and for some time now many have required that each author’s contribution be specified on papers submitted for publication. According to Sigmar de Mello Rode, chair of the Brazilian Association of Science Editors (ABEC) and executive editor of the journal Brazilian Oral Research, the pressure to publish has prompted cases of researchers claiming to author more than 30 articles a year, which, Rode says, suggests that either they do not really take part in most of the projects or they do so only superficially. “Many people confuse the formation of a research group – which is a good thing if you want to work towards a goal – with an assembly line, where everyone does just one thing and enhances the prestige of the others in the group. This really does boost each participant’s scientific output, since a number of studies are published in the time it takes to carry out one. There are even cases of multiple authors on reviews of the literature, where there is no justification for more than one or two authors,” says Rode, professor at the São José dos Campos School of Dentistry, of São Paulo State University (Unesp). “The attribution of authorship has become a source of ethical dilemmas in science communication, either because authors who have contributed little or nothing are included or because ghost authors are not,” says Packer.

According to the criteria of the International Committee of Medical Journal Editors, there are three prerequisites for being an author of a study: the researcher must make a substantial contribution to conception and design and to data acquisition, interpretation, and analysis for the study; he must participate in drafting the manuscript and revising it critically, making an authentic contribution to its intellectual content; and he must give his final approval of the content to be published. “Anyone who does not qualify as an author should be cited in the acknowledgements, including participation in the work itself, its translation, funding acquisition, technical and statistical analyses, material loans, and so on,” says Rode. The Code of Best Scientific Practices released by FAPESP in 2011 sets out similar guidelines: “A scientific study must name as authors all those researchers, and only those researchers, who have expressly agreed to be named and who have made direct, substantial intellectual contributions to the design or performance of the research whose results are presented in the study,” the code advises. The code further stipulates that “specifically, providing infrastructural or financial resources for the performance of a study, such as laboratories, equipment, consumables, materials, human resources, institutional support, etc. does not suffice for being named as an author on a paper produced as a result of this study.”

032-035_Autoria_221-3NElson ProvaziIn the view of Joaquim Nóbrega, editor of the Journal of the Brazilian Chemical Society, although science is an increasingly cooperative, collective effort, certain criteria must be safeguarded. “Even if we understand that producing a manuscript involves a number of stages and tasks, we cannot exonerate ourselves from the ethical concerns that have historically afforded science broad social recognition,” he states. The rules adopted by this publication include the stipulation that the author who submits a manuscript is responsible for listing all researchers who were effectively responsible for the study. “We crosscheck the authorship on the revised version with that on the submitted manuscript. If any change has occurred or if new authors have been added, we request a justification signed by all authors, along with a detailed description of the editing process and each author’s contribution to the manuscript. If the situation is not duly explained, the revised manuscript will be rejected,” states Nóbrega.

The new taxonomy proposed by the Harvard and Wellcome Trust team is more comprehensive and to some extent replaces the concept of “authorship” with “contribution.” “It provides more precise recognition of the authors’ individual contributions and establishes categories of contribution not included in current criteria,” says Packer. In his opinion, adopting a new classification system is feasible as well as desirable because it is consonant with the effort to lend greater transparency to the process of producing and communicating science. But Packer observes that some rough edges will have to be smoothed out. One of these relates to copyright, which would be extended to a larger group of researchers than today if the proposed outline prevails and is accepted by scientific journals. Another challenge would be to establish indicators that take the new taxonomy into account. “There’s no point adopting a new classification system and then continuing to use the h-index as a parameter,” states Packer, referring to the measure that represents the relation between the quantity and quality of an author’s scientific output, based on the number of published articles and their citations, but that does not take into account whether the researcher was the lead author or played a secondary role.

According to those responsible for the taxonomy, journal editors would benefit from it because the new classification could save the time currently spent checking each author’s participation and managing disputes between authors. Rafael Loyola, editor-in-chief of the journal Natureza & Conservação, published by the Brazilian Association of Ecological Science and Conservation (ABECO), feels that how much good it will do editors is a relative question. “We have no way of verifying whether the information we’re given is true. We trust that the authors are acting in good faith,” he says. Loyola thinks that researchers themselves stand to gain the most. “The task of organizing a list of signatures can be simplified with the taxonomy. Similarly, it would be easier for the group of authors to see how each individual contributed, and this might avert disputes when negotiating everyone’s ranking on the list,” he states. Loyola also says that the editors of Natureza & Conservação do not often encounter trouble with attribution of authorship, but it does happen at times. “There was one occasion when the authors of an article that had already been revised several times and was about to be accepted for publication made an unusual request: they wanted another name added to the author list. We asked how the new author had contributed and they told us he was the head of the laboratory and they’d made the mistake of submitting the article without his knowledge. We only accepted the article after we’d made sure the new author had indeed taken part in the research, and we required that he add his comments to the manuscript and agree with the final version,” he explains.

Charles Pessanha, editor emeritus of the journal Dados, a publication of the Institute of Social and Political Studies (IESP), of the Rio de Janeiro State University (UERJ), had a similar experience. An article that had already gone through peer-review and had been returned to the author with a request for changes was re-submitted – with one more author. “We advised the author that this was an irregular procedure and that the article wouldn’t be published like that. But she successfully showed that the second author had contributed a lot to the revision of the article. We came to the conclusion that she was being honest and shouldn’t be punished for it. And we published the article,” relates Pessanha, professor at the Economics Institute of Rio de Janeiro Federal University (UFRJ).

In Pessanha’s opinion, the proposed new taxonomy is following an inevitable tendency, which is to credit everyone who takes part. “The solution is very similar to the one adopted by movie studios. There are so many people involved in a film that specific credit has to be given to all of them at the end of it,” he says. “The process of scientific production requires an increasing range of skills and, concomitantly, the participation of new actors. It’s vital that they all receive credit, because each one needs to be recognized for what he did.” Pessanha does believe, however, that crediting everyone should not distort the concept of authorship, which is much narrower. “We have to find a way to recognize everyone’s contribution. But someone who didn’t participate in designing the study, laying out the research, and interpreting and analyzing the data can’t be considered an author. Just as in film credits, producers are recognized, and even receive their own awards, like other professionals. The fact that someone acquired funding for a study doesn’t automatically make that researcher an author. An author has to take part in the design and academic preparation of the paper.”