Friday, March 18, 2011

Signaling Advice For Grad Students

I think a lot (though certainly not all) of the difficulty that the average grad student faces as they work their way towards the PhD can be avoided by thinking about interactions with faculty as a signaling game, and realizing that the game changes as you move from undergrad to grad, and even at different stages of grad school.  Specifically, pay mind to the change in the qualities upon which the faculty are conditioning their strategy (where that strategy may be assigning grades, writing letters, extending requests to co-author, etc, as the case may be).

When you're an undergrad, especially if you've majored in a non-technical field, faculty are mostly looking for signs that you're studious.  That's about it.  We'd love to require more, and some of us do, but odds are you can earn a very respectable undergrad GPA just by working hard, doing (some of) the reading, attending classes, following instructions, etc.  Because, sadly, most of your classmates aren't going to do much of any of that.

By the time you get to grad school, faculty assume they don't need to worry about whether you're studious.  Perhaps that's a bad assumption, but you as a grad student need to recognize that signaling studiousness is no longer a winning strategy.  Faculty are now looking to see that you can contribute, if only in some small way, to our collective understanding of whatever it is that you study.

That may sound trite.  But here's some concrete ways in which that makes a difference.

1.  Page counts: when undergrads are told to write 5 page papers and turn in papers that are 7 pages long, it impresses faculty.  Because it signals studiousness.

When grad students turn in a 40 page paper that was supposed to be 20, no one is impressed.  Most academics journals do not specify a minimum page count or word count, but they all specify a maximum.  That should tell you something.

2. Organization: when undergrads write or give presentations, they can basically engage in extended and unstructured infodumping and still be rewarded.  It's not ideal, but it's not penalized like it should be.  As long as it looks like a lot of effort was involved, that's fine.  The only people who will ever read an undergrad's work or listen to their presentations are people who have to do so.

As a grad student, you are supposed to be able to sell a wider audience on your work.  An audience that isn't going to be impressed that you put in a lot of effort.  You need to have a clear argument.  And you're not writing a mystery novel.  No one wants to wait until halfway through the paper/presentation to discover the answer to your question.  You should also provide the minimum amount of necessary detail and no more.  Learn to leave things unsaid so that you know what the first question in the Q&A will be.  Learn to use footnotes.  And appendices.  And that some things that you think are really cool and really bolster your claims are actually pretty uninteresting.

You also need to understand that research presentations have fundamentally different goals from scholarly articles and books.  The former serve as advertisements for the latter.  However much detail you think you should include, you almost certainly need to cut that in half.  Probably more.  Again, no one is going to be impressed at how much you have to say.  Get the idea across in the time provided.

3. Lit reviews: undergrads don't read, so there's nothing to say here.

But based on their belief that signaling studiousness is the path to success, grad students tend to write unstructured lit reviews that go on forever, citing everyone they can think of (which is entirely too correlated with syllabi from courses they took) without saying anything.

No one wants to read that.  Faculty want to see that you can make an independent contribution to knowledge.  At bare minimum, yes, that assumes you have done the reading.  But when you send a manuscript to a journal, they are not trying to judge whether you've done your homework.  The question is whether readers of the journal will learn something new, something of interest, by reading your article.  That's much more likely to be true if the author not only has read the literature, but understands what we've learned from it, what we haven't, which contributions were foundational and which inessential, and which rested upon faulty research designs.  So if you are asked to write a literature review (or comprehensive exam answer) in your first few years of grad school, don't treat it as an opportunity to signal that you've done your homework.  Use it as an opportunity to signal that you have the skills necessary to contribute to the field.

4. Criticisms of your work: most undergrads have learned that they can bargain their way to better grades, which usually has nothing to do with attempting to persuade the instructor that their work contains more merit than the instructor realized.  No, undergrads focus on persuading you that you have not recognized their worth as a person, which apparently is supposed to make you judge their work in a new light.  This doesn't work as often as many of them think, but it works far more often than it should.

As a grad student, you will inevitably face some very damning criticism of your work.  Your intuition will probably be to say that, yes, your work possesses this critical limitation, but there's really nothing you can do about it.  As if that somehow renders the point invalid.  Which, perhaps, it would -- if the point that was being made concerned your work ethic or some such.

But that's not the point.  So don't do this.

When you ask people to accept some conclusion about how the world works, if someone points out that they are not required to accept this conclusion because the argument you have made is insufficient to establish it, the question of whether you are personally responsible for that shortcoming is irrelevant.  Always remember that your goal is to contribute to knowledge, not to signal that you are really studious.  If your argument does not tell us what you think it tells us, then you've got a problem.  No one is going to pat you on the head and give you a cookie for doing your best.  Those days are gone. 

In fact, there was a subtle shift here as I went down the list.  Throughout much of grad school, there is something to be gained by signaling desirable qualities about yourself, as a person.  After a certain point, it's not about you anymore.  It's about your work, which is not an extension of your personal identity.  By the time you are ready to defend your dissertation, if not long before, you need to learn that even signaling that you have the personal characteristics one would expect of a scholar is only part of the story.

Eventually, your work needs to speak for itself.

I've written this post with grad students in mind, but looking back over the list, I can name at least one faculty member who is still struggling to learn some of these lessons.  :\


  1. Does the signaling in #3 change if you are a graduate student that clearly has something new to say, but you have a greater need to establish what you are doing in the broader context of mainstream IR?

    I am asking this question for... a ... friend.

  2. Hey, Matt.

    Yeah, I think when your work comes at things differently, it strikes me as perfectly appropriate to cite a wider swath of literature. But even so, the point is that your lit review should be focused and lead naturally to the argument you make in the rest of the work. If you find that existing work is lacking b/c it commonly assumes x, y and z, and you think we can learn a lot about the world by relaxing those assumptions, then you want to cite people who assumed x, y and z. But if they also commonly assumed v, and you're not so much dealing with v one way or the other, you can ignore that point, even if some of the works that assumed v are famous and were assigned in your classes.

    What I had in mind, in contrast, is lit reviews that bring up pieces that aren't directly related to the point the author wants to make one way or another simply because those pieces also had something to say that was tangentially related and the student read them in seminar and wants to prove that they see the connection (however loose it may be). Those lit reviews are painful to read.

    That said, some reviewers get pissed off if they think some piece they really like (perhaps because it was written by themselves or a friend of theirs) is tangentially related and you fail to cite it. And studies have found that articles with more citations themselves get cited more. So I guess the scattershot approach has its upsides... but I still stand by the claim that the lit review needs to feel like it's making a coherent argument.

  3. Terrific post. Fairly consistent with the advice above is Ashley Leeds' grad student guide for writing research papers (see esp. the bit about lit review):

  4. Thanks for the feedback, and for the link, Jon. I'll be sure to check that out!