Friday, April 1, 2011

Abuse of Power

If you've been following my twitter account (@profgears), undoubtedly you've seen a flurry of tweets about student expectations, terrible advisors, and politics of publishing. This stemmed from GMP's post the other day over at Academic Jungle and DrugMonkey's post on the subject clearly in GMP's camp. DrugMonkey's post heading The Care and Feeding of your PI: a tip for grad students and postdocs should tell you all you need to know about their position on the matter.

I clearly disagree.

As a grad student, you're not there to serve a professor at their beck and call. As a professor, you're not there to wipe the nose of your students. Both should have expectations and duties that are clearly laid out. I've already discussed this a few weeks ago. The professor (or PI, or whatever you want to call them) is supposed to be the student's professional mentor. The prof is supposed to be the student's guidem helping transition from a student to being a autonomous researcher.

This clearly breaks down when there is a lack of communication between both parties. For instance, the prof expects the student to publish as many papers as they can before graduation. But you'd be surprised how many students don't know that and how many profs don't tell their students that. So you have a prof that's expecting paper drafts and a student wondering what to do with their data. The student tries to schedule meetings but the prof has little time to discuss it. Months go by. Still no paper. There is increasing frustration on both sides.

Inevitably, a downward spiral happens and things turn to shit. The prof did a bad job of mentoring their student, yet from the prof's perspective, their student is a slacker. The prof thinks their student isn't working enough and so the student gets even less time than the minuscule amount the prof already gives them. The student gets frustrated because their mentor and career guide doesn't respond back to them and now the student starts worrying about graduation deadlines and whether they are even going to get a recommendation.

At this point, both are in the wrong because they didn't communicate with each other. But the professor is always in the power position and it should never be abused. And it sure sounds like some Profs/PIs do abuse that power by holding graduation and recommendations over their students' heads. Those are the same Profs/PIs that are supposed to be mentoring their students. That's a terrible situation to be in if you're the student because the one person you're supposed to go to for advice is giving you shitty advice and guidance. In the odd chance that student makes it to the next level, how likely is it for them to mimic their advisor's behavior? I say that's highly likely and I sure as hell don't want to be that kind of prof.


On the topic of publishing and who gets coauthorship, it's very hard to describe in 140 characters. My general feeling is this. Anyone who contributed to the writing, data gathering, interpretation, and oversight should be a coauthor. However, everyone one of those people should be capable of writing a significant chunk of the article. That doesn't mean only the person who wrote the article should be the sole author.

Here's how you test it. Ask each person if they could write the draft. If someone says they could probably get 60% of is done, that's fine (that's a significant chunk). If they said 10%, they don't belong as a coauthor. That's what the acknowledgements section is for.

In practice, most groups aren't that strict because everyone needs more papers to be considered successful. But if I adhered strictly to that, I'd be missing my prof on every paper. And for me, when I'm in that same situation, I want to be involved enough to always pass that test. Otherwise, I would and should expect my student to submit it without my name on it because I haven't contributed enough. But if it comes to that point, then there's obviously something else going on in our advisee-advisor relationship and there's a bigger problem at hand than coauthorship on a paper.


  1. The more I read blogs by the biomed folks, the more I realize what a different world it really is from my experience. A group of 10 graduate students is *large* in my field, not small.

    Consequently, the advisor-advisee expectations are rather different. My advisor may be neglectful, but to me, holding a thesis hostage until the papers are done borders on abusive. I agree it's all about communication and I think some professors don't realize that they aren't actually very good at it.

  2. Regarding coauthorship, I agree with you in principle, but the percentages you mentioned cannot be universal, because contributions are often hard to quantify. For instance, in condensed matter physics and materials science, there are people who specialize in the growth of complex materials; they grow a sample and then other groups build structures on them and measure different phenomena. I am not sure the growers could write 10% of the final paper, but their contribution to the work is absolutely critical and they are routinely included as coauthors. So I believe coauthorship criteria should be more about providing" salient contribution" to the project design, execution, data interpretation, or paper writing than about who could write how much text. For instance, Nature has you specify (at the end of the manuscript) exactly who among the authors did what.

    I am not going to comment on your derisive tweets and comments at my place. Nothing I wrote warranted their tone, and it's a shame you feel it did.

  3. True enough about certain specialties providing key contributions to the research. I have heard of instances you speak of in MatSci, ChemE, etc where the whole object [game] is to synthesize as many materials and samples as possible. I have heard of PhD students making complex samples and farming them out to every lab to shoot different lasers at it or eVs or whatever. Then they all write papers jointly and everyone is happy. I've heard of cases where PhD students are coauthors on 20 to even 46 papers (I think that's the highest I've heard). That seems a little absurd to me.

    And because my field in entirely different, I don't necessary understand why that's soooo great for publishing and science in general. It seems like a shotgun approach to build samples and get people to do sciency things with them and then get your name on papers.

    If you told me the majority of those people said "Hey, I'm going to synthesize this surface because it will be better for absorbing [whatever]. My theory predicts this." And then they go out and find a group that can measure if indeed their theory matches their measured properties, OK. But I am a little my cynical and much more skeptical and my guess is that samples are synthesized for the sake of synthesizing and they'll match the theory to the measured results after its been farmed out.

    And at a basic level, I use photodetectors, lasers, optical components that are vital to obtaining the research results. I don't go as the CEO of Thorlabs to be a coauthor though. Obviously that's a ridiculous case and those components aren't as "rare" but they are equally as vital. Even when I have custom optics manufactured even with crazy special coatings on them, I still don't think the people doing that warrant a coauthorship. That to me is something for the acknowledgements.

  4. You make a good point about putting CEO of Thorlabs as coauthor. My feeling is that if you pay for a tool or a sample fabricated by a company, that's compensation enough. Alternatively, if a sample is custom-made by your collaborator, coauthorship can be considered alternative compensation, epecially because those samples are likely fabricated using federal grants money; coauthorship on joint papers shows federal agencies where their money invested in materials growth went.

    I think no system for assignment of coauthorship is perfect, and, as in all human endeavors, there will be those who abuse any given system. I know both types of growers you mention: one of my closest collaborators is a grower, and the three of us design the whole project (the grower, a colleague who makes devices, and me on the theory end), specific materials/structures are grown for specific experiments because the theory says something novel can be expected; in this case all three of us write the paper together (with the involved students and postdocs contributing). One the other hand you have "growbots" who, as you say, mass-produce stuff and end up with ton of pubs that way, which does sound suspect (sounds like it would be better if they formed a small company so people could simply buy the materials if the growers provide no intellectual input). Anyway, no matter how you assign coauthorship, there are people who have integrity and don't accept coauthorship unless they did a significant part, and those who have no problem with more of a courtesy coauthroships...

    Btw, I don't know if you are aware that there are several hundred authors, in alphabetical order, on every particle physics paper -- everyone in these huge named collaborations gets on every paper. I was on a university committee trying to evaluate candidates from different disciplines, and here's this particle physics guy who had some 70-80 papers in a single year that way. How the hell is anyone supposed to evaluate any one person's contribution? Apparently, it can be a huge problem when such people are up for tenure/promotion, but there is no way their field is changing their norms...

    Anyway, to wrap up this superlong post: I agree with you in principle. Still, there is such a variety of different practices in different fields, that any formal criterion for authorship assignment is bound to fail in some cases. And there are always people who push the conventions to the boundaries of the ethical... At some point, you have to trust that people will take pride in their work and won't accept coauthorship on a paper where they haven't contributed substantially in some critical aspect.

  5. Very good point with:

    "...there is no way their field is changing their norms..."

    That sums it up nicely.

    Overall, it is a delicate decision about who gets coauthor, who doesn't, and what's the acceptable practice for the field. Regarding the growbots, that's something an outsider (like me) would probably have an issue with. But if you're looking for faculty candidates and there's a PhD student churning out 20+ papers during their PhD, I can definitely see why that person would get hired/interviewed.

    I have seen those types of papers with 70+ authors. I came across one that had 3 pages of authors and affiliations. Found it quite humorous. I don't get it but that's not my field and if the journal Ok's it, who am I to say what's right on that.