Frerik van Beijnum Frerik van Beijnum 7 October 2012

Inviting co-authors

Tags: , , , , ,
Posted in Ethics, Getting published, PhD life

A problem I often encounter is deciding who to invite as co-authors. On one hand, you want to show appreciation to the people that helped you in the process of obtaining your results. On the other hand, generously adding authors will dilute the contribution of the people that made the largest contribution. In this post I would like to sketch a few hypothetical situations in which someone could be a co-author. The main goal here is to provoke some discussion on this subject, and learn about some good practices.

Undergraduate students

As a first example, I would like to take undergraduate students. From acquaintances in other disciplines than physics (psychology, medicine, pharmacy) I heard that it is rather uncommon to have undergraduate students as co-authors. An important reason is that many students are involved in the project, and including them on a paper would drastically increase the number of authors.

In my research however, I feel they often greatly contribute to the research. First, they do work that is otherwise not done, e.g. some wild new idea that can be explored. Second, they are often involved in many discussions, which may give new insights. Hence, if they were part of the project I would be inclined to ask them as co-author. On the other hand, some undergraduate students only do what they are told. If they are lucky, and some nice results come out, should they still be co-author? I would guess not.


Another interesting group of people is specialists. For example, someone that makes a crucial sample or someone that performs a specific and important measurement that you cannot do yourself (easily). Typically these people are not part of the entire of the research. Should these people be invited as co-authors, if so for how many papers? For example, a sample is made, on which different sets of measurements are performed, that are published separately. I would say that the person who made this sample made a crucial contribution to all papers, and hence should be co-author. However, it is the people who perform the extra measurements, that may put most work (i.e. time) in it. Ad Lagendijk’s survival guide for scientists suggests inviting the specialist as an author for one paper. I think this is rather meager, especially since they are never first authors but often second, third or fourth authors. Secondly, most sample production is so complicated that a lot of time is lost when an inexperienced PhD student does it, instead of the experienced specialist. Hence, the student can obtain more results because of the effort of the specialist.

Former colleagues

The last category is the “almost finished spectacular result”. What happens often is that someone invests his PhD in doing a spectacular and very hard measurement. The plans were too ambitious however, and the student leaves a paved road for his or her successor. The new PhD student has a huge benefit of all the work done, is able to get the job done and obtains the spectacular results that were hoped for. Sometimes, this may even be a series of very interesting results. I think in these cases credits are due to the student who started the project and got it running. Again, I feel this student deserves to be co-author on a number of papers that build on his work. This will, like the case of the specialist, not be first authorships, and therefore some generosity is fair. It is difficult, however, where to draw a line. For me, it should be leading that a student gets a fair amount of papers, compared to the quality of the work.

To finalize this post, I wish to note that all of this would not be a big deal if we would not try to quantify the quality of a scientist using the number of authored papers. This leads to the effect that PhD students already have to worry about the number of papers they write, and the impact factor of the journals they write for.

- - - - - -
If you like this post why don't you email subscribe to our new posts. Or subscribe to our RSS feed.
  1. Jacopo Bertolotti

    8 Oct 2012 11:26, Jacopo Bertolotti

    Personally I prefer to be a bit on the generous side when deciding who is co-author of a paper.
    Of course a nice chat in front of a coffee does not qualify anyone as a co-author, even if the result of the chat was a great idea that made the result possible. For such things there is always the acknowledgement section.
    But anyone who produced a significant part of the results (samples, experimental results or data analysis) should be an author. I am thinking in particular about the undergrad student who does not have yet a full understanding of the subject (not his/her fault, just need time) and thus performed the measurements and the instruction and guidance of someone more expert. Maybe the undergrad does not always deserve to be first author, but if he produced the results he deserve recognition for it.

  2. Frerik van Beijnum

    8 Oct 2012 12:52, Frerik van Beijnum

    Jacopo, thanks for your response.

    The coffee machine discussion is a nice one, I would indeed say that is part of how a group functions.

    With respect to the undergrad, I agree that you should forgive him/her not having a full understanding of the subject. However, some students need a lot of guidance, hardly ask critical question, and get a poor grade. If still something nice comes out, one should consider not putting the student on the paper.

    I think an undergrad should only be first author if he/she writes the manuscript.

  3. Unregistered

    9 Oct 2012 10:28, Jon

    This is why there is the phrase “material contribution” in most journal submission guidelines: an undergrad who did what they were told to do (not really material), a specialist who ensure that the experiments ran to time and budget and quality guidelines (material), a previous researcher whose work enabled you to accomplish your own work (material). Of course there are shades of grey for each of these, but the basic point will usually be “Could I have done this same work without the support and involvement of ‘x’?”

  4. Frerik van Beijnum

    9 Oct 2012 11:02, Frerik van Beijnum

    Jon, I like your question “Could I have done this same work without the support and involvement of ‘x’?”. I agree that this question works well in many cases. Sometimes, however, I would also often answer: “Yes, but it would have taken me x months longer”. Would the criterion then be whether x is large (>25% or so) compared to the length of the entire project?

  5. Mirjam

    10 Oct 2012 11:50, Mirjam

    I think it is impossible to formulate a closed set of rules, a lot will depend on subjective factors and intuition. Generally, I’d rather have 1 author too many than 1 too few. I also don’t view more authors as ‘diluting the contribution’ of the others. People need to get credit, even if it was just spending boring hours doing what you’ve been told or if in a split second you had a groundbreaking insight that made someone else’s project into a success that would otherwise not have been possible. Again, it takes intuition to judge what a crucial insight is, of course. Author order adds a degree of freedom that can be used to distinguish contributions. People are too worried about authorship, numbers of papers and impact factors… a seriously good scientist will be recognized irrespective of these statistics.

  6. Frerik van Beijnum

    12 Oct 2012 11:33, Frerik van Beijnum

    It is indeed hard to formulate a closed set of rules. However, objectifying the criteria for co-authorship a bit may help in making decisions and explaining them to your colleagues. Using subjective criteria and intuition may come across as randomness or personal preference. In some of these cases authorship may cause personal friction within a group.

    I am not sure whether people are too worried about authorship, although I agree they are poor indicators of people’s quality. However, if you try to get a tenure track, these statistics are used to select people (so I heard). Moreover, the success of this tenure track is measured using the same statistics. As long as these statistics are considered important, people will try to improve them and be worried about them.

  7. Mirjam

    16 Oct 2012 9:32, Mirjam

    These statistics are used to a much lesser extent than one is led to believe and in general they are viewed within the broader context of someone’s overall performance and professional and personal circumstances. Otherwise, there would be no need to ever do an interview or ask for letters of support when you hire or tenure people, you could just do the numbers…

XHTML: You can use these tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>


By submitting a comment here you grant this site a perpetual license to reproduce your words and name/web site in attribution.

Subscribe without commenting