A throwaway line in one of my maniacal, whiny, ranting posts a few days ago about being told to add a coauthor to a manuscript triggered some debate that I think need to be pursued a little further.
The question is: what level of involvement constitutes being included as a coauthor on a scientific manuscript?
Should there have been some input to the manuscript? And if so, what part? Hypothesis, design, methods, analysis, writing? Is one of these sufficient or should the coauthor have contributed to more than one? Or all?
Exactly how much effort should one be expected to do to be included as a coauthor?
When I was in grad school, I put several weeks work to help out another student including two long field trips to collect data which formed the entirety of his thesis and subsequent papers. In that instance, I was included in the acknowledgement section of each publication. Was this politically motivated? I don’t know - we had the same advisors. Am I bitter about it? Sure, but I didn’t specify that I expected to be a coauthor before agreeing to help out, so that’s my own fault too.
If a student or trainee does some work for another PI, should their PI/mentor also be included as a coauthor? Even if they haven’t contributed anything (no reagents, consumables, equipment, nada) other than allow the student or trainee to perform the work?
Some PIs insist that everyone in the lab is included as a coauthor on every single paper generated, regardless of whether they worked on the study or not. Often it’s the same 6 or 7 names with the first name shuffled around depending on who’s project it was.
Does this practice dilute the fact that the first author (supposedly) had the most input or does nobody really care? At my grad school, publications are now assessed according to how many authors are listed and the faculty are given more credit for having smaller numbers of coauthors.
When you see two names on a paper, do you perceive that the authors did more work or contributed more to the paper than one that has five authors? What about a paper that has ten or even twenty authors?
What about in terms of looking at this on someone’s CV. If you see that this person has coauthored 20 papers before they finish their postdoc, would you view this as a person who was really productive or be more likely to assume that they are really good at one technique and that they run the same technique for everyone in the lab? Would you think that they actually had a substantial input into the paper/study?
If the person is 10th author out of 20 on every single paper, what does this say about their research capabilities? Does it dilute their perceived ability(s)?
Is it more important to include someone as a coauthor if they are a student or trainee than if they’re a tech? If you have a student that’s looking to get into med/grad school or a prestigious postdoc program, are you more likely to include them if they did one small piece of analysis? What about a career technician that has been in your lab for 10 or 20yrs for whom coauthorship isn’t important (I mean this in a career sense and yes, I know this doesn’t apply to every tech … and everyone likes to see their name on a paper don’t they).
If you had a high school student that came in every afternoon and washed the glassware, would you include them as a coauthor? I know some PIs that do because the student wants to get into a competitive undergrad program. Would you be more inclined to add the high school student rather than a research assistant that cleans the glassware as part of their 9-5 job?
Should you include mentors that really haven’t had any stake in your papers? Both of my postdoc mentors had some input into my papers so that’s not an issue for me at the moment, but when I was in grad school, one of my advisors was an advisor in name only and had absolutely nothing to do with my studies; he told me not to include him on my papers because he knew that he’s had no input into them.
At what level do you consider “input” to be worthy of coauthorship? Do you include people with whom a casual conversation may have sparked your idea for the study in the first place?
Do you include someone if they’ve given you something that you then used in the study? If someone has given you samples, should they be included? If it’s a cell line that you have been given that you have then cultured and treated, do you include the person who gave you the cells originally? Or should they be acknowledged as having given you a “generous gift”?
Should one seek to include famous coauthors? There’s a pioneer in my field whose name is on almost every paper that comes out (except those from our lab). I know where his lab is based and I know that there’s no way that he has physically been in the other labs so what sort of input is he having into these studies that has constituted being included as a coauthor? Is he having an intellectual input? Have they included him to get the manuscript over the line with reviewers? Has he actually provided samples or were samples analyzed in his lab? In which case why isn’t anyone else from his lab being included in the papers?
I’m not sure I have the answer to any or all of these questions, but as a beginning PI, these are important issues to consider. I’m sure that DrugMonkey and PhysioProf will have opinions on this and I’m really interested to hear their thoughts as well as any others.
Friday, August 29, 2008
Subscribe to:
Post Comments (Atom)
7 comments:
Number one observation: Since these questions are slippery over time and highly field dependent it is essential to understand the practices, rules of thumb and interpretive valence of your most-likely fields of comparison.
Two: Middle authorships in and of themselves are nearly meaningless.
Three: Symbols of equal contribution don't really mean squat when it counts.
Four: Middle authorships are meaningful in the context of an entire CV. I.e., raw numbers are helpful in some contexts. Also, they help to support the idea that the domain of the papers on which you are first author (or last author as PI) is really your domain within a lab (or collaboration).
interpretive valence
Dude, do you even understand your own fucking self when you say shit like that!?
HAHAHAH!
Seriously, I have strong opinions about all of this (surprise!), but unlike DoucheMonkey, I will let others chime in before giving mine.
Interesting, I have the same question...being new to the blog scene and at the bottom of the totem pole as a graduate student I probably can't offer anything substantial.
What I can say is....I doesn't really matter. Everyone is going to do something different and everyone is going to keep doing it their way.
My lowly grad student advice...get it in writing. If you are adamant about who gets authorship and you are not the primary PI, then you had better find out early on who gets put on the paper and get it in writing. From what I've seen ethics kinda go out the door (which is a whole 'nother issue in itself), its all about the politics.
Very gray area on many of these scenarios, and I just think it's important not to get your panties in a bunch when it comes to most of them.
In general, I think someone (and their advisor) should have their names on a paper if the person contributed something that resulted in a figure in the manuscript.
So, help that doesn't ultimately work out? Sorry, you're in the acknowledgments. Washing dishes? Acknowledgments (although I wouldn't even do this).
Samples or cell culture? This is trickier, I think. If someone gave me a cell line that I then modified, worked up on my own, etc, I would probably go with the acknowledgment. If someone gave me samples and data to back them up, I'd probably put them as an author.
There is such a fine line here, and it is probably best to consider all of the people involved before making decisions.
As to how to interpret author lists, I take them all with a grain of salt. 20 authors? Well, this had better be a pretty robust paper. 2 authors? You can be damned sure that the grad student did most (if not all of the work). You really need to look at the contents of these papers before judging them. That's the problem with a CV. Itemized lists of publications don't tell you everything you need to know.
1. Being simple myself, I try to keep things simple. If you contributed to one of the figures or tables, you're a co-author. Otherwise, you're acknowledged. Everyone in my lab knows this upfront.
I've never been in a situation where someone from another lab contributed in such a way, and outside the context of a formal collaboration. But I've always felt that two PI's should be able to come to a reasonable understanding, one that helps out the careers of everyone involved. (I know, that sounds pretty naive.)
2. As for multiple author papers, the comments here should resound greatly with those who write letters of recommendation. An applicant may or may not be able to "pump up" his role in a study that had 10 coauthors, of which she was sixth in the list. But a recommendation that takes the time to spell out why the placement in the author list does not do justice to the contribution (preferably, by explaining the vital contribution in explicit terms) is a very effective one.
Of course, such a letter gets very long if the applicant has many such studies. But even in these cases, some effort that explains, in positive terms, this collection of papers goes a long way.
I as a grad student find this very difficult and am glad for all advice I can get. In my personal opinion, I agree with art. But what is with the influence of one's supervising PIs opinion? If PI finds it important to include techs ("What about a career technician [...] for whom coauthorship isn’t important?") that are only in our lab for 4 years and give a ship about authorships, what if PI finds it important to include grad students working on the same subject just to make them feel better (PI did not say that, I find no other reason than this)?
What do I do in such a case? Opposing PI last time lead to lengthy discussions which don't help much and I find myself frequently in a situation where I have 4+ coauthors on my papers while I think only 2-3 or so really deserve it. I usually give in.
I always thought it does not do *me* as being the first author any harm, but when you say
"...publications are now assessed according to how many authors are listed...", I start panicking...
I've always thought that figure/data = co author. No data or figure = no co author. Of course this is not reality ;) at least not in my field... but the two last people are usually the money people, either the PI and/or the collaborating PI.
I look at middle authorships as a way of interpresting the lab culture. THe whole lab is on the paper and first is shuffled around, isn't an approach I like. It makes it dluted for all, imho.
And middle authorships, as in your question regarding 10 papers with only middle names, would point me into one of two directions. Either the person is a technique/interpreter e.g. micro array analysist, or the person hasn't gotten gis/her own project to work to the finish line.
A middle name as a first publication when changing labs points me into "I'm on the ball and soon I will have my first authoer paper". Then again, this is very subjective since this is what I have done ;)
and in regards to the cell culture/bacterial strains. I think it depends on how important the culture/strain was to the paper. Was it imperative and the paper wouldn't happen without it? Was there something done with the culture that you didn't do? Or was it just "you have a culture that you published before and know can distribute to others"?! I'd put my givers of bacterial strains in the acknowledgement IF they have published and described the strains lots before... otherwise I'd try to exchange some more data to put in the paper and make someone a coauthor. (for me, coauthor comes with reading and writing and commenting the manuscript... then again, I might be a tad bit naive here)
Post a Comment
Go ahead and say whatever you like. Comments will only be removed if they are spam or are of a vicious/derogatory nature.