Thursday, May 22, 2014

Paper review

To stay in the academic loop, I'm a reviewer for two major journals in my field.  Been doing this since as soon as I graduated a few years back.  The past year I've reviewed about a dozen papers.

When I was in grad school, at first I thought everything was brand spanking new and going to change the world. That quickly changed as I realized that the research was cool and beneficial but 99.9% of it won't have any direct clinical impact. For the stuff that isn't basic science it's still so far cutting edge that the impact is unrealized. In industry, my work has changed the medical device world with tons of direct impact. Well, maybe not the world, but 10's of thousand's of peoples' lives. But it's not really brand spanking new besides a few things.  But I have to keep these things on the D.L. to keep other companies from utilizing it.  A lot of the ideas in the papers I've read have been thought of before. I've only read one article that I was amazed by, but it was written so poorly it nearly got completely rejected. I reviewed this paper this week and it reminded me about how awesome writing is. 

It's a great way to organize your thoughts. Throw some pictures and diagrams in a great document and you've just won yourself a multimillion dollar grant. Or look back at your old writing and discover how much better you've become. Or discover how you've changed as a person and scientist. But you don't necessarily need good writing for nonprofessional writing (ahem...this blog).

I love technical writing because it's the culmination of a lot of hard work. Through the planning, organizing, reporting, and discussing results, it's a lot of fun to pull everything together and let the world know what's going on. In order to do this, the language of most prevalence is English. While my English isn't the best I'm pretty decent at disseminating my work. This paper I'm reviewing is reminding me that if you want to do science you should either learn great English writing or work and publish in an area where you don't need good English. With this paper I'm reading it's just such a shame that the PI approved submitting this paper without reworking the writing and figures. I'm going to try and push this into major revs (so it's not killed) because it's great work. It's just very tough since the writing and figs are sub-par. As a big fan of writing, it just kills me to see work like this that's been approved by a PI. If this person tried to submit this crap to the FDA he would be heavily punished for it. The PI is doing the grad student that wrote this a disservice by not pushing the student to learn better writing skills. Some people will probably say, "students won't listen, are incapable, don't have time, etc".  To that, I was a mediocre writer when I entered grad school. My former advisor wouldn't let me submit anything for publication until he says it was acceptable. This improved my writing dramatically.  At first it took forever (months of constant rewrites) to submit pubs.  Holding off publication (and therefore graduation) is one of the few things PIs can use to control the students. And it's for good reason: learning to disseminate information is part of the scientific process, so this needs to be learned. 

Again, this is probably just my naïveté. But my advisor was just as busy as the rest and had time to teach me writing. I teach entry level scientists and engineers better writing. Why can't the corresponding author on this paper teach good writing? Or at least not let this garbage get submitted. 


  1. I don't know. I am getting totally disillusioned with teaching people how to write, because so many of them just really, from the bottom of their soul, don't give a shit.

    This is my unedited opinion. There are people who are worth my time, who work really hard on every draft, and who already on their 2nd or 3rd paper can produce a first draft that already looks like a real paper written by a professional scientist. I have no problem working with these students. But then there are those who either don't give a shit or are so incapable for whatever reason that they are hopeless and they are not a good investment of my time; when on paper 5 your first draft still looks like a freshman essay, you are officially hopeless. The thing is, when someone takes forever to learn to write science well (nobody is expecting Ernest Hemingway here, this is not that high of a bar, people who barely speak English can churn decent technical texts) then that means they really don't care or that they are not a good scientist, because you can apply the scientific method to totally deconstruct how papers are to be written (hint -- it's very formulaic; you can write solid, non-cringe inducing papers by upholding some basic tenets that can be deduced if your brain is on and you read enough papers).

    Which brings us to the fact that students don't read papers, and that's a huge problem. Again, those who seem to have their heart/head in the right place and are motivated do also read enough and also improve very quickly. The rest probably shouldn't be doing a PhD anyway.

    end rant

  2. I understand where you're coming from, but if nothing else do you have them make coherent figures with good scientific practice? This paper had missing axis labels with panels misaligned to one another. It had multiple samples run through testing but didn't even think of reporting a mean and standard deviation. It was just so sloppy, even away from the poor writing riddled with syntax errors. Unfortunately, I'll never know whether this student is a lost cause, or if the PI just didn't care.

    1. Oh, I am totally with you on this aspect -- I never let anything subpar go out because it reflects poorly on *me*. Often that means that, after several ridiculously pointless revisions, I have to say "screw it" and redo all the figures myself and ask for raw data so I can do some extra number crunching before I am finished with the text and the manuscript sees the world.

      But I do know there are professors who are considerably more hands off or don't care enough. That paper you saw was likely submitted without any PI input... Unfortunately, it happens. For instance, I remember my former postdoc when he interviewed for a position with me, I was appalled at the graphs in his presentation -- I thought who the f*** taught you how to make graphs, the lettering is microscopic, I can't see squat, the line colors were of mud and slime, and on top of that slides loaded with equations. I did hire him and he did well, but our first year was spent on me teaching him to write and make figures and it was very painful (he confessed that there was a figure in our first paper together that I had him redo exactly 27 times :-). It turns out his advisor was completely hands off and would edit only very minimally (typos). I saw one paper from my postdoc's PhD that he said he could not publish, it kept getting rejected and he was angry with the world; I offered to look at it and it was absolutely awful. Poorly written, horrible illegible graphs, just a disaster. I am not surprised at all it was not getting published (I am more surprised their other stuff did.) And the advisor signed off on it, apparently without putting in much thought.

      There. I feel better for getting that off my chest. :-)

      I think that the conclusion is -- all students start as clueless, that's a given. But, thereafter:

      1) Motivated student plus involved advisor = a student who quickly becomes a competent writer; papers that come out are professional

      2) Unmotivated student + involved advisor = advisor with hypertension, who eventually gives up on trying to teach student, does everything by him/herself; papers that come out are professional

      3) Student (motivated or not)+uninvolved advisor = disaster; papers that make reviewers cringe, pull out hair, and shake their fists at the world
      [Note: very good students can sort of teach themselves, but it takes long and is needlessly hard]

  3. Amen! PIs need to realize it will reflect poorly on them when releasing poor figures/writing.

  4. "It's a great way to organize your thoughts. Throw some pictures and diagrams in a great document and you've just won yourself a multimillion dollar grant."

    I wish this was true -- how wonderful it'd be if the grant success only depended on how it is written.

    1. True-though one thing I know for sure: a poorly written grant has never won me an award.

  5. I am an associate editor for a journal in my specialty. I just desk-rejected a paper that sounds like what came into your lap -- it was just god-awful. The quality was that of a first draft of a newbie student's first paper. Nuff said.

    I didn't even send it out for review, I rejected it with the explanation that it's nowhere near ready to be considered for publication and that I will not waste our referees' time, and that they can come back after they have drastically improved the presentation.

    1. I wish more editors would take your stance. I fear they're scared of being seen as an asshole, but people need to buck-up. It's far better to seem like a butt (though you're helping the authors by telling them they need to get in shape) than waste multiple reviewers' times on a POS article. Maybe of editor's names were more anonymous then fewer crappy papers would make their way to the reviewers?