Determining Authorship Order

My main advice in this task is that investigators discuss authorship and “credit” at the outset when projects are being planned and implemented (& as new team members come on board) so no unspoken assumptions are made that erupt at manuscript time. A short piece in Nature has some potential merit – but good heavens, I have no idea who would have the time to do this for each paper generated:

“Using a multi-criterion decision making (MCDM) approach, a group of potential co-authors decides on a set of items — such as figures, tables, text and ideas — that comprise a manuscript. They score each person’s contribution to each item as a percentage. Types of contribution vary across publications, but a group of co-authors is well placed to make judgements. As they may not always agree exactly, a range can be assigned that they can ‘agree to disagree’ on.

The group should then assess the relative importance of each item and put them in categories, in order of importance to the manuscript. For example, category A (the most important) might constitute a 15% weighting, category B 10% and category C 5%. This way, each item is given a weighting that represents its importance to the whole work. Finally, each author’s relative contribution to each item is calculated.”

Of course, this would at last explain something I see in many grant applications – the incorporation in research project timelines of 6 months of support at the end of the project period for publication …

8 Comments »

  1. PhysioProf said

    That Nature piece is nutso, at least from a biomedical science standpoint.

    I suspect plant scientists aren’t churning out a half-dozen papers a year on data with short shelf lives … but in cases when discussions aren’t held early on, breaking down the final product to objectively figure out who contributed what and who is responsible for what (as is increasingly required by journals in any case) is at least on the right track. It just shouldn’t take a 0.7 FTE statistician to do so. – writedit

  2. Yeah, that’s a little over the top. Drugmonkey and I discussed a similar issue regarding absolute authorship – whether an author should be included or not in the first place – and I see more utility in that arena than for authorship order. It seems order, outside of the first or last spot, matters less than I once thought.

    Often a footnote is included as to how first author position was selected or if the first two or three authors share equally as lead authors. See ref #30 in Science 15 June 2007;316(5831):1622-1625 as an example of how the order among equal authors is sometimes selected. – writedit

  3. whimple said

    That’s crazy. It is flat-out impossible for two people to have completely equal contributions to a paper. The coin flip and the footnoting of equal contributions are a cop-out on the part of the corresponding author. Go ahead and footnote that the first N authors all contributed approximately equivalently, but take responsibility for picking the order that is listed. Making these kinds of tough decisions is why the corresponding author is getting paid the big money.🙂

    Watson & Crick copped out in the same manner … – writedit

  4. PhysioProf said

    “footnoting of equal contributions are a cop-out”

    Most of the time it is not a “cop-out”, but rather a way to gain acquiescence from non-first authors that they are being treated fairly.

    Frankly, when I assess a CV, I completely ignore “co-first-authorship”. As far as I am concerned, the first listed author is the first author, as determined by the corresponding/senior author. The fact that this person was actually listed first indicates to me that the PI considered their contribution to be more significant, regardless of the designation of the second author as an “equal contributor”. Based on discussion with colleagues, this is not an uncommon practice.

  5. drugmonkey said

    agree with Physioprof that the “co-first” is a total buy-off of authorship. all that matters is the actual order on PubMed. I’ve also started seeing “equal contribution” down to third author position! Meaning equal from #3 and #4, not “co first” down to three that is. Starting to see “co-senior author” symbols creeping in too! absurd. nobody but nobody cares a whit whether you are third or fifth author and I can’t understand where people think this plays as currency. Far better to just adopt the extensive description of contributions from each author…

  6. […] to the way authorships for Nature letters/articles are actually determined! Writedit labors under a similar misconception that mere discussion of contributions and relative merit can solve the problem as well. Don’t […]

    No clue as to the basis for the statement above. My argument is that this should be laid on the table when the project is still a twinkle in someone’s eye and that everyone involved is clear on their relative contributions throughout the research process. When this doesn’t happen at the outset and continue thereafter, the choice at the end is either discussion or pistol duels. – writedit

  7. drugmonkey said

    I just think you come across a little, shall we say optimistic, on this issue. Agreed that discussion of relative authorship position throughout the process is a good idea. But the real knock-down fights that I see wouldn’t have been helped a bit by such a process. Mainly because the real form of the manuscript, the relative importance of each part, etc, frequently comes together at the end. Frame of reference being the high-impact journal article in biomedical science (read, almost exclusively molecular biology focused) which contains a wide variety of models, techniques, assays, etc. Yes, there are many fields and types of work that incorporate semi-monolithic projects in which the eventual paper design/structure changes little from the outset- notice these fields have less frequent and nasty authorship debates. So one needn’t focus attention and prescriptions a la the NatureJobs thing on the easy situation. That is kind of like the science ethics classes that discuss obvious straw-man scenarios and ignore the tough/complicated cases that cause the problem in the first place.

    I’ll give you an example: Imagine PostdocA has created a genetic animal model (KO mouse, say) which has multiple wide-ranging phenotypes. Let’s also surmise that the gene of interest is quite similar to a prior genetic model created by PostdocA resulting in a few publications. Thus, not only is the specific animal the domain of A but also the family of related genes. PostdocA focuses her attention on PhenotypeA over the course of several years worth of work. Meanwhile PostdocB in the lab works on PhenotypeB which is her primary expertise and “project” within the group. After some years of work, Postdoc B leaves the group with what should be the first-author paper on PhenotypeB “almost” completed. “Almost completed” being the first usual sticking point in that the postdoc always imagines that little work remains to generate a paper when in fact a LOT of work remains to wrap the story up, submit it , respond to critique and get it accepted. Then it transpires that B never so much as generates a draft and things drag on for a while. There is a story there so A takes over and wraps up the project. So who is listed first, is there a “equal-contributions” asterix, etc? B maintains that she did most of the work. A points out that as B left things, there was not, and wasn’t ever going to be, a paper. With me so far?

    Okay, now imagine GradStudent C who worked with B and contributed some of the data albeit of less than stellar quality. C also works on some related stuff as her major work over a period of years. A ProjectC which is originally conceived as going toward a different first-author paper. Now imagine that the PI gets wind that the original gene and/or proteins related to said gene are under review at Science/Nature/Cell, maybe even by two groups. The project under discussion has some additional features and can thus be argued it should be co-published or competitively published (“Hey Science, Nature and Cell are going to have papers on this and you don’t have one, will you take ours?” – yes this happens!). The point being the time is ripe RIGHT FRIGGIN NOW to publish this thing or maybe never have a shot at such a high-impact journal. Unfortunately for poor C, her work is the most closely related to the competing manuscripts under review. So the PI slips in C’s work, which isn’t close to ready to stand on its own, at the end of the manuscript. Now C is pissed because 1) her first author project has just disappeared because of scoopage and 2) she’s been helping A to clean up after B on the PhenotypeB paper. So now C is agitating for climbing over B in the authorship list and maybe even co-first with A. Now the fight is truly joined.

    And we haven’t even discussed the contributions of TechnicianD who makes beautiful histological analyses, CollaboratingLabPostDocE who contributed key reagents, ExPostDocNowIndependentPI-F who started the work on PhenotypeB with A and half-trained PostdocB, etc, etc, etc.

    You got your transgenic animal model to work?

  8. drugmonkey said

    Whether a transgene or ko “works” is immaterial in the hands of a good Cell/Nature type PI. Throw in enough array data, artificial expression systems, “rescues”, “recapitulations” and other crap….et voila!

    Nobody’s going to go through the same 10 person-years of work to end up with a “me-too” (unpublishable, insufficiently hot) or a “they’re completely wrong” (unpublishable, politics and the burden for proving someone’s data wrong is, ironically, quite high) result.

RSS feed for comments on this post · TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: