A Democratic Method and Index to Rationalize Authorship Decisions and Determine the Order of Authors

Disputes on authorships or the order of authors on publications are among the most common conflicts in science, particularly where multi-author papers have become the norm. Here, I propose an adaptable and generalizable method to rationalize and democratize decisions about authorships, which can be applied in diverse scientific areas. In the process, the authors of a manuscript collect, discuss, score, and weight their contributions to the final manuscript. The decision process is assisted by a simple spreadsheet-based “authorship matrix”, which helps to break down the complexity of different contributions and their importance. The matrix generates a score s describing the percentage of the author’s total contributions and can be used to determine authorship and their order. The transparency and simplicity score s can reduce conflicts and thus increase productivity in collaborations. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 9 October 2018 doi:10.20944/preprints201810.0181.v1 © 2018 by the author(s). Distributed under a Creative Commons CC BY license.


Introduction
Publications increase the visibility, funding prospects, prestige, and career opportunities of scientists.In this context, bibliometric parameters such as the number of their publications, citations, the h-index 1 or the impact factor of scientific journals have gained a lot of importance.3][4] Many journals now require specifying author contributions (such as conception, experiments, writing etc.), but again, the list of authors for a specific contribution is often long.In addition, an informal code behind the authorship order exists that can vary between scientific disciplines or countries. 2Often, the first author did most of the practical work and the last author is the principle investigator and corresponding author.Where this is the case, the first and last positions are more prestigious than positions in the middle and are often disputed between colleagues. 5Such authorship order conflicts can become very emotional and have the potential to disrupt the group climate and collaborations.When the first authorship dispute occurred in my group, I was looking for a way settle it in a rational way.The point systems proposed by Schmidt 4 or Kosslyn 6 are primarily designed to determine who becomes an author or not on a paper, and seemed too little flexible as the categories for possible contributions (e.g.conception, experiments or writing) are fixed and fixed points are allocated to each category.However, the importance of each category may vary substantially from paper to paper.Moreover, the decision process could be perceived as undemocratic or not transparent, if the PI makes such decisions and the decision can always be questioned due to different opinions on the importance of a specific contribution.In fact, almost 70 % of the coauthors of multi-author manuscripts disagree with the corresponding author's perception of individual contributions. 7This discrepancy may partly from the challenging and complex cognitive task of selfreporting. 8To increase the acceptance, transparency, and quality of authorship decisions and their interpretation, it would therefore be desirable to formalize and democratize the decision process.Here I propose a decision making protocol assisted by an adaptable "authorship matrix", which generates a single number, the contribution score s, which describes an author's total contributions to a manuscript in percentage.

Generating the Score s
To finalize the list of authors and their order on a publication, all, or at least the most important authors of a study (for example all potential first and last authors), come together, discuss their contributions and create an authorship matrix (Fig. 1).Authors with minor contributions can also be included to provide an external opinion, and collaborators can be added to the discussion by phone or a video conference.scientists in publications, 9 and encourages collaborations.The discussion and decision process is relatively fast; in our experience, it takes about 30 minutes.
As the authorships are only determined before submitting a manuscript, everyone is likely to stay motivated throughout the writing process.It makes of course sense to assign someone as the lead for a new project, but sometimes students or postdoc leave before a project is finished, or new people enter a project and give it an unforeseen new direction.
The matrix-assisted method proposed here helps to break down the complexity, and structure the decision process, and could therefore be a valuable tool for ombudspersons.Finally, journals or funding agencies awarding fellowships may in future not only ask to specify author contributions, but also their contribution scores s. to increase transparency.I realize that many will object to this method and the score s as it could inevitably end the common practice to include authors that did actually not intellectually contribute to a paper.Maybe some have also developed other ways to deal with authorship conflicts, or it does not make sense in their discipline or for their particular manuscript.In any case, democratizing and rationalizing authorship decisions using the score s could be very useful to those who want to settle or reduce authorship conflicts in their groups, creating a more collaborative, professional and productive working environment.

SUPPLEMENTARY FILES
Example of an authorship matrix (Microsoft excel file)

Fig. 1 .
Fig. 1.Example authorship matrices for three authors (X, Y, Z).An empty matrix is created (steps 1-2) and duplicated.Then all authors fill in points (step 3) and weigh categories (step 4) in their own matrix M to generate the contribution score s (step 5).In this example, author Y could become first author, as she has the highest score s.