The Development of a Worksheet for Authorship of Scientific Articles

One of my most unique articles published was not on science, but on the process of reporting science. In 1985, I was a graduate student at UC Davis. I was asked by a faculty member to help two researchers from another university with a project I was familiar with. Easy enough. A few months later, in a pleasant surprise, I received a draft manuscript, with the two researchers listed as authors one and two, me as third, and the faculty member as fourth. I provided some important methodological revisions, and returned the manuscript. Later, I found that the fourth author had only a single comment... replacing me with him as third author!  

As any graduate student knows, authorship is a big deal. I stewed over my "demotion" for a few weeks, then set down my thoughts on authorship (below). I submitted it to a journal in my professional area (Wildlife Society Bulletin), and it was rejected ("no real contribution," reviewer number 2 wrote). I resubmitted it to a non-peer reviewed journal, Bulletin of the Ecological Society of America). It was accepted for publication.

Now, prior to Google Scholar and online databases, the way many scientists reviewed journal contents was through a weekly publication titled Current Contents. I found this description of Current Contents at that time:

Current Contents was first published in paper format, in a single edition devoted only to biology and medicine... Initially, it consisted simply of a reproduction of the title pages from several hundred major peer-reviewed scientific journals, and was published weekly, with the issues containing title pages from journal issues only a few weeks previously, a shorter time lag than any service then available. There was an author index and a crude keyword subject index only. Author addresses were provided so readers could send reprint requests for copies of the actual articles.

Current Contents had an international audience, and every week there were a few articles highlighted and summarized. In another pleasant surprise, the editors selected my article on authorship to highlight. The reprint requests poured in. Take that, WSB editor!

Decades later, I note that my original article is cited in many newer articles on authorship in science, and even used in graduate workshops on research methodology. According to Google Scholar, as of today that original article was cited by 9. My article on authorship has been cited by 35. I reprint it below in hopes it is useful to other graduate students.

***

A WORKSHEET FOR AUTHORSHIP OF SCIENTIFIC ARTICLES

 

Inclusion as an author in a scientific publication is important to many ecologists for reasons of prestige and advancement. Publications are a key factor in deciding on promotions for many ecologists at universities (Jackson and Prados 1983, Croll 1984). The order of listed authors on a paper is assumed to be an indication of the relative contribution of each of the included authors.

 

Day (1983:15-19), Croll (1984), Kennedy (1985), and Jackson (1986) reviewed contemporary difficulties with decision-making in assigning authorship. Dickson et al. (1978) proposed guidelines for determining inclusion and ranking in authorship of a scientific publication. They divided research investigations into five areas: conception (including funding), design, data collection, data analysis, and manuscript preparation, and recommended that authors need to make, at a minimum, a significant contribution in manuscript preparation and in at least one other area. Authorship order was determined by a ranking of the number of areas in which significant contributions were made.

 

This paper details a method for assisting in (1) deciding who is to be listed as an author on a paper, and (2) the ordinal ranking of authors listed on a paper. Of course, the best procedure for dealing with potential problems in assigning authorship is to deal with the issue at the beginning of a study. However, even preassigned roles can have complications, especially when personnel on a project change, or when responsibilities are transferred. In addition, people often underestimate the inputs required, especially time, for the various contributions, making initial agreements, in retrospect, seem unfair. The trend toward multiauthored papers may indicate how research is becoming increasingly interdisciplinary. In these situations, a method for defining authorship roles becomes useful. This simple technique should be a useful decision-making aid, especially for projects with many researchers involved.

 

A general framework for a decision-making worksheet, with an example, is given in Table1. For each of the five parts of the research investigation (as defined by Dickson), the relative contribution of each participant is assessed. For each part, total contributions should equal 100%. When all contributions have been assigned, row values are added, resulting in a "score" of between 1 and 500.The relative contribution of all participants can then be assessed, and a natural break between subsets of scores on the lower end of contributions can be used as a cutoff to delineate inclusion as an author. Scores can then be ranked for order of authorship.

 

This technique has a number of assumptions. First, it assumes that each of the five parts of a research investigation are weighted equally. In some situations, this may not be the case. For example, a study may require minimal funding, the infrastructure of a principal investigator's laboratory may be essential to a successful project, or the dataset may be collected over several years. This situation is easily dealt with by weighting the unbalanced part with a multiplier. For example, all values in the "data collection" column can be multiplied by 1.2, if data collection is judged to have been 20% more important than the other areas.

 

Secondly, this technique assumes that all contributions can be judged fairly and accurately. This may not always be the case; indeed, it may be that this technique would only be necessary for papers where it is difficult to assess contributions. Two points are suggested for resolving this. It must first be recognized that each contribution score is usually an estimate, and, as such, has some corresponding error associated with it. Therefore, the difference of only a few points between participant's scores is probably not sufficient to rate relative contributions, and other methods must be utilized to determine authorship ranking (perhaps even the flip of a coin). As the second point, a consensus type survey system, such as the Delphi system (Schuster et al. 1985), may be useful as an in-house tool for resolving difficult authorship assignment problems, although it is recognized that assigning authorship is rarely a democratic process.

 

How are contributions assessed? One method that could be used is the actual time (hours, days, years) put into each of the five parts of the research investigation. A key problem here is the importance of experience. For example, how would you compare a two-hour contribution to a project's design from a person with 30 years of experience with a two-hour contribution from a person with little or no experience? Another method, admittedly subjective, is an assessment of the "importance" (relating to intellection) of contributions in each area. Again, a consensus type survey can be helpful in arriving at an acceptable and agreeable assessment. The development of some criteria for better assessment of contributions is needed. Time should be minimized, while intellectual contribution should be maximized, yet it is easy to visualize a project in which time is a real measure of effort.

 

Finally, there is a situation which involves teams of workers involved in one of the five parts. A realistic example would be having many workers assisting in data collection. Although the team's contribution may be large (perhaps 100% of the data collection), the relative contribution of each team member is small. The "points" given to this team may then be assigned to the team coordinator or leader. There is some question whether a "technician" should ever be a coauthor, especially if his or her sole responsibility is data collection or data collection and analysis, when the analysis is limited to performing routine operations rather than interpretation (Dickson et al. 1978).

 

It must be repeated that this system for determining authorship of scientific articles should not replace consultation among authors. However, it should be useful in delineating relative individual contributions when there are many, and it can help project coordinators or senior authors identify personnel who have contributed in a significant way to a study's conclusion. Authorship is a symbol that means taking responsibility for the contents of the paper (Jackson 1986). If the responsibility is there, inclusion as a co-author is appropriate. This worksheet should be helpful in defining this responsibility.

 

ACKNOWLEDGMENTS

 

For helpful comments on this essay, I thank C. Shugart, W. Howard, J. Aloi, R. Case, T. Tomasi, D. Anderson, R. Johnson, J. Tully, P. Moyle, and T. Salmon.

 

LITERATURE CITED

 

Croll, R. P. 1984. The noncontributing author: an issue of credit and responsibility. Perspectives in Biology and Medicine 27:401-407.

 

Day, R. A. 1983. How to write and publish a scientific paper. ISI, Philadelphia, Pennsylvania, USA.

 

Dickson, J. D., R. N. Conner, and K. T. Adair.1978. Guidelines for authorship of scientific articles. Wildlife Society Bulletin 6:260-261.

 

Jackson, C. I. 1986. Honor in science. Sigma Xi, New Haven, Connecticut, USA.

 

Jackson, C. I., and J. W. Prados. 1983. Honor in science. American Scientist 71 :462-464.

 

Kennedy, D. 1985. On academic authorship. American Council of Learned Societies, Office of Scholarly Communications and Technology, Scholarly Communication Reprint 4:1-5.

 

Schuster, E. G., S. S. Frissell, E. E. Baker, and R. S. Loveless, Jr. 1985. The Delphi method: application to elk habitat quality. United States Forest Service Intermountain Research Station Research Paper INT-353.

 

Robert H. Schmidt. University of California, Hopland Field Station, 4070 University Road, Hopland, CA 95449 (robert.h.schmidt@gmail.com).

(PDF) A Worksheet for Authorship of Scientific Articles. Available from: https://www.researchgate.net/publication/44018729_A_Worksheet_for_Authorship_of_Scientific_Articles [accessed Aug 25 2024].