While watching the end of a movie last night to see where exactly it was filmed, I paid attention to the credits. Coming from a publishing perspective, I tried matching what they list to what we capture about the author, researcher, and creator in a bibliographic citation. What are the matches? The film lists the animation team, the animal handlers, the costume designers, the makeup artists, the photographers, the accounting team, and on and on and on. There are hundreds of people behind the scenes, and they are all listed in the credits. What do we list for a team of 3000 researchers working on a big data project? We cite a few authors with a spare affiliation line and an acknowledgement to the funders. What about the rest of the contributors? Are there others that should be cited?
In Boston on May 16 at Harvard University, The Wellcome Trust hosted the International Workshop on Contributorship and Scholarly Attribution. Several presenters outlined other roles we should probably begin to acknowledge and what those roles are.
Cassandra Extavour talked about the issues of authorship practice and policy in experimental science research labs. She outlined both the roles and the rewards inherent in each from the particular perspective of an academic researcher.
- The principal investigators will publish, and should demonstrate the ability to lead productive research teams. Their role indicates an intellectual contribution that is critical to the projects. They have to obtain funding to support research, and their reward is to get tenure, promotion, and other honors. They also help grad students graduate, and help post-docs get jobs.
- Post-docs are looking for career stability and for opportunities to publish. They need to demonstrate the ability to carry out institutional research and independent research so they can, depending on career goals, obtain post-doc jobs or actual research and professor positions.
- Graduate students want to graduate so they can get a job either in industry or in academia. They will publish to add to their credentials and show they can do research. It may be important to them to be acknowledged for their role in research, but they often are not cited.
- Research assistants want to keep a good job and obtain a subsequent career opportunity when their current one runs out. They contribute to research, but the opportunity to publish is a bonus.
- Undergraduate students certainly have their eye on graduation and getting a letter of reference so they can enter the job market or obtain postgraduate employment. They often contribute to the research but are seldom listed as an author or acknowledged in the publications.
Dr. Extavour was the first in a panel of three. She was the first on the panel to point out the differentiators in the order of authors as listed on the publication. Her field is biochemistry, in which field the author order is a definite indication of contribution to the paper. The first author did the experimental and analytical work, and is usually a grad student or postdoc. The last author, a senior author, is the principal investigator and a significant contributor of both intellectual content and funding. The second author did some of the research and is usually a student or tech. For the other authors, it is unclear how they contributed.
The author role is merited for those who did at least some of the experimental design, carried out experiments, analyzed the data, wrote the manuscript, or obtained the funding. But there is a constant problem. Who decided the merit? There is conflict in how much contribution is required. Many people naturally under- or over-value how much they contributed. The PI makes the decision but should clarify expectations / lab publication policy early in the project, and then continue to guide team members in effective communication / negotiation of authorship goals.
The corresponding author takes the responsibility for the work. Most likely, a graduate or post-doc last author is the senior author. Authors can add a section to their paper, “author contribution”, explaining who did which analysis, and who actually wrote the paper.
Liz Allen of the Wellcome Trust was the second panelist. She outlined her thoughts in “Contribution and attribution in research: funder perspectives.”
The Wellcome Trust currently is funding about 3000 researchers in 30 countries. Because they need to measure effectiveness of the funding to help determine future funding, who has done what is key to them. They have about 60 kinds of funding, including things like investigator awards, fellowships, strategic awards, and transitional awards.
They want the funding to deliver discoveries, applications of research, products, and inventions. They want to fund research leaders and the research environment, and exert influence for the public good.
This leads to the need for measures to find out the impact their funding made on the research outcomes. The existing measures by publication don’t do it very well. For example, big science has collaborative teams, big teams. In a large list of authors, some of whom Wellcome funded, how do they know what funding there was, whether it helped, who did the writing, and who did the research? There may be an acknowledgment thanking them for the funding, but how was it applied? And there is often a disconnect between author list and acknowledgments.
To pursue the practicalities of ensuring accuracy in measuring outcomes, they measured the average number of authors on Wellcome Trust associated papers. The statistics are enlightening. Genetics papers increased three-fold per author between 2006 and 2010. Some of this is because of the collaborative nature of big science, with papers including up to 3000 authors. But in a list of 3000, where the first and final authors have known roles, what did everyone else do?
Christine Laine presented “Authorship: the journal perspective.” She provided this definition: An author is an individual who has made substantial intellectual contributions to a published work.
Authorship really matters. It has significant academic, social, and fiscal implications. Readers want to know who did the work. The authorship identifies who is accountable for the integrity of the work. Conflict of interest statements to identify any potential bias are very important. It is a fundamental principle that all persons designated as authors should qualify for authorship, and that all who qualify should be listed.
The International Committee of Medical Journal Editors (ICMJE) guidelines for determining authorship state that the authorship should meet three criteria:
1. Substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data
2. Drafting the article or revising it critically for important intellectual content
3. Final approval of the version to be published.
Aside from text contributions, things that may qualify people as contributors include acquisition of funding, collection of data, referral of patients, provision of study samples, and oversight of the lab. There may be acknowledgements listing contributors and the nature of their contributions.
The Council of Science Editors has each author and contributor personally affirm his or her role and disclose it publically to readers. Since editors cannot determine roles, they need the contributors to do it. Whenever there are author battles, it indicates a poorly organized lab. Some of these issues can be partially resolved by following the Committee on Publication Ethics (COPE) guidelines. Editors cannot police the author contributor listings for every submission, but they sometimes need to clarify suspicions that an author list is incomplete or inflated.
There is an increasing number of contributors, due to the changes in the way science is now being done. Many journals are collecting and publishing information about author and other contributions. Most journals do not police but simply report what authors say, although some ask for the identification of a guarantor, like the corresponding author.
There are many types of author contributions, including concept and design; analysis and interpretation of data; provision of study materials; collection and assembly; and drafting of the article. Any contributor needs to be able to say “I am willing to stand and defend the content of this paper.” Contributors must be accountable. They also need to know which other colleagues are responsible for other sections of the data.
What’s the difference between author and writer? What level of contribution makes someone an author? These are the perennial questions for the editor and for all contributors.
Judith Singer, Senior Vice Provost – Harvard finished up the set by giving an overview assessing the contributions from the academic perspective.
Author order is difficult to determine – alpha order is not fair to those later in the alphabet, but it is very commonly used in the humanities and social sciences.
In peer review, the author names are not present. Funding decisions might be made based on a body of work rather than an individual paper. It is a question of track record. It is more problematic for a woman and for minorities. The question is who really did the work.
Harvard did not have a tenure track until the 1990s. To consider promoting to tenure, the organization needs to collect a lot of information. It’s a very high stakes decision, and very expensive for the institution, especially if they get it wrong. It is much higher stakes than deciding whether to publish or to fund someone. For a parallel guideline, consider that venture capitalists figure less than half of what they fund will succeed. If tenure turns out to be a mistake, the professor is still there for life because no one else wants him or her.
Guidelines on first and last, etc. author position varies greatly by discipline. Multidisciplinary teams are increasingly published with less reliable author placement. Citations impact is difficult to determine and self-citing is a problem. In history fields, people don’t cite each other. We really need to think about the researcher’s impact on the field. We are rethinking the way we ask questions about research effectiveness. When people take the work from one field, grapple with it, massage for another application or to apply in a new field, where does the attribution lie? Suppose you take techniques for one field and apply them in a new field. What might be prosaic in the old field might be arbitraged in the new field as a potentially disruptive technology. We need to look for independence of thought and the research trajectory. While one might be skeptical of detailed acknowledgement fields in papers, they do need to be done with balance. Talks, presentations, are a place where people show their intellectual chops. Talks don’t count for tenure but are important for evaluating how well someone does know the field. It remains a conundrum.
So there you have it. Four different takes, like the blind men and the elephant trying to get a picture on how contributions should be made and understood. There is clearly a need for clear guidelines. It also appears that there are a few awards, either fiscal or social, beyond the awarding of tenure, which then cause a twisted focus on how publishing is done. Perhaps emulating the movie credits, we could give awards for best engineering design, best data analysis model etc. Maybe we even need a taxonomy of contributor roles.
There is much to think about, and there will be several more papers to present to you in the coming weeks.
Marjorie M.K. Hlava
President, Access Innovations