A researcher presenting work, from https://www.flickr.com/photos/wcupa/31253642307/

Today we’re releasing tenzing, a web-based app that makes it easier for researchers to indicate who did what in their manuscripts. Tenzing helps researchers record early in a project the expectations around the different roles of researchers in a project, and it makes life easier when it comes time to submit the paper to a journal. It uses a standardised format called CRediT, the Contributor Roles Taxonomy.

Most science today is conducted by groups of people, not solitary individuals. Different members of the team often have different roles. Yet until a decade or so ago, journals still operated as if there was no need to provide any information other than a list of names — the author list. Some information could be tentatively inferred from the order of names in the list, but how order is determined reflects often-unwritten practices around authorship that can be obscure to people outside a subfield and can differ substantially between labs.

In the last few decades, many journals have begun to encourage, and some to require, that teams give some indication of who did what in the work reported by a paper. In some journals, this is done in a brief “author note” or “author information” section. Thanks to this development, researchers are more likely to get the specific recognition they deserve. Funders of scientists and those that hire scientists can allocate resources better, assembling teams with the right combination of skills.

At least, that’s the theory. Holding much of this back has been a lack of standardization. Without a consistent vocabulary for describing what each researcher did in a project, and without a structured format for that information, it is difficult to aggregate across papers the type of contributions a researcher makes. For institutions and meta-scientists interested in hiring the right combinations of people, it is difficult to tally the sorts of contributions typically involved in different sorts of projects.

To address this, standardized taxonomies are needed for the types of contributions made to scientific projects. One such taxonomy, developed in 2014, is CRediT, the Contributor Role Taxonomy, which has since been utilized by dozens of publishers (see adopters) and hundreds of journals. While the CRediT categories don’t fit well with every research project, they are useful in many cases, and it will soon become an ANSI standard. It can be particularly helpful for collaborations with many authors.

Edmund Hillary and Tenzing Norgay in 1953. Photograph: AP

What Tenzing Does

Tenzing helps teams use CRediT to indicate what area of a project each person contributed to. Tenzing can be helpful even if you’ll be submitting your manuscript to a journal that doesn’t require or endorse CRediT.

Tenzing users enter information into a Google Sheet template. There is a column for each of CRediT’s 14 research contribution categories, and a row for each project contributor’s name.

Researchers tick the columns to indicate which areas they contributed to (or that they plan to contribute to, if this is being done at the planning stage).

After the researchers upload the filled-out spreadsheet, tenzing can generate various outputs, such as the CRediT information in prose sentence form, suitable for the Author Information section of a journal article.

The annotated list of authors with affiliations is also generated, in a format suitable for the first page of one’s manuscript.

Tenzing also generates the XML code for article webpages to encode CRediT information in machine-readable form. We’ve had some interest from publishers in integrating this with their workflow — a submitting author would upload the tenzing-created XML file to the manuscript submission portal. At present, however, no journals are able to handle author-generated XML.

Why use tenzing?

Let’s say you’re starting a research project, mostly with researchers you’ve worked with before. Many members of the group may have some expectation of being included in the eventual paper, but not all, and even some of your frequent collaborators may not have a clear idea of what their exact role is.

It’s best if teams agree on what everyone’s role is, and whether that role deserves authorship, towards the beginning of a project. Doing it at the end is a formula for disputes. Such disputes form the majority of cases brought to research integrity offices. Most disputes don’t go that far, but still may leave some people bitter at being excluded, or resentful that some people were included on an authorship list without apparently deserving it.

To avoid such issues, you send a link to a copy of tenzing’s Google Sheet template (you can alternatively use an Excel template) to your collaborators, asking them to indicate what areas they expect to contribute to, along with their institutional affiliations. This process might prompt more discussion of who will do what. You then examine the spreadsheet to assess whether each person’s planned contributions would qualify them for authorship — this will depend on the journal’s authorship guidelines.

The project proceeds and when writing the manuscript you go back to the spreadsheet, check that it’s still accurate, download it to your computer, and upload it to tenzing, which generates an annotated author list that may be useful for the title page of your manuscript. You also have it generate a paragraph describing the CRediT contributions of the authors in concise prose.

Imagine your plan is to submit your manuscript to the open access journal Collabra: Psychology. Collabra’s publisher has not (yet) implemented CRediT in their platform, but its editor does encourage researchers to provide CRediT information in the Author Contributions section included in each article. Or perhaps you’ll submit your manuscript to a journal that doesn’t have a policy about CRediT but does encourage researchers to indicate who did what in an “Author Information” section - and tenzing will write that prose for you.

For those researchers who write their manuscripts in R Markdown and use the R package papaja to generate APA-formatted manuscripts, you can now include CRediT info along with the other author information format in the header of your manuscript and papaja will use tenzing’s code to include the corresponding prose in the Author Information section of the manuscript.

Read more about tenzing and contributorship in our paper or in the official documentation of the underlying R package.

The future

If you have questions about tenzing or suggestions, message us or post an issue on Github. If you’re associated with a journal and are thinking of encouraging authors to use tenzing, contact us. We’re also interested in hearing from publishers interested in integrating into their workflow the XML that tenzing produces. We think tools like tenzing should be expanded to use other taxonomies as they become available, but CRediT is the only standard that is widely used today.

-Alex Holcombe, Marton Kovacs, Frederik Aust, and Balazs Aczel

If you use tenzing, please consider citing our paper:

Holcombe, A. O., Kovacs, M., Aust, F., & Aczel, B. (2020). Documenting contributions to scholarly articles using CRediT and tenzing. PLOS ONE, 15(12), e0244611. https://doi.org/10.1371/journal.pone.0244611

Science, Experimental Psych. Reforms. @PsyOpenAccess