Friday, October 4, 2013

Evaluating Ph.D. advisors

Wondering how to find a good Ph.D. advisor?  Stuck with an advisor you can’t stand?  A friend was complaining to me recently, and a funny thought occurred to me: never once, in the course of my three and a half years at McGill, did anybody in an official position ever ask me to evaluate my advisor, Andrew Hendry.  Nobody in the administration ever asked me about my experiences with him, whether I would recommend him to others, what his strengths and weaknesses are... nothing.


The nine types of PhD advisor (actually, I think there are a lot more than nine!).  Which type is your advisor – or which type are you?  (Click to see at larger size.)

Now, to get the suspense out of the way, I thought Andrew was a great advisor – for me.  Everybody has different needs, and every type of advisor has their pros and cons.  Andrew was a perfect fit for me, and as a result, I mostly enjoyed my PhD (and when I didn’t, it wasn’t his fault), and I was productive, and so forth.  The key thing in achieving this nirvana-like state is to evaluate yourself honestly, and look for an advisor who is right for you – even if that means saddling yourself with a slave driver or a control freak.  With the wrong advisor, you might get dragged through an awful experience that burns through years of your life, only to drop out without the degree.  You don’t want to be That Guy.


That Guy.

But that’s not really what I want to talk about.  I want to talk about evaluation of advisors.  It would be useful to know what type of person a prospective advisor is, and sometimes it can be hard to get a good read.  Furthermore, there are some types of advisors that really don’t have pros and cons; they only have cons.  They’re not shown in that comic up above, but you know what I’m talking about.  The egomaniac, who will take credit for your work and will try to keep you from being first author on your own papers.  The sexist pig, who will horrify and humiliate you and make your lab a living hell.  The flake, who will not live up to their commitments and will make it impossible for you to progress.  And so forth.  How can you dodge these types of advisors?

In the present system, the best way – really the only way, unless you are an exceptional judge of character – is to talk at length with your prospective fellow grad students when your prospective advisor is not in the room.  But that requires that you do an in-person visit (a lot of people wouldn’t complain about their advisor over email), and it requires some degree of trust and honesty all around (with people you probably just met that day), and it can be hard to get the advisor out of the room in a short visit – and of course the prospective advisor might be a great fit for your prospective grad students but a terrible fit for you, or vice versa.

The present system: how to know which advisor to avoid.  (Click to see at larger size.)

What occurred to me just now is: why on earth doesn’t the university administration take any interest in evaluating advisors?  During my undergrad, at the end of every course I took I was given an evaluation form to fill out, asking for feedback on all sorts of details about the quality of teaching I had experienced.  I think that form was four pages long, and I must have filled out 30 or 40 of them.  Those forms then went into the official record, and were used in tenure evaluations and departmental planning and all sorts of things.  Compare that to the feedback that was asked of me regarding Andrew as an advisor: None.  Nada.  Zip.

Wouldn’t it make sense for such evaluations to be requested at the end of one’s time as a PhD student?  Indeed, wouldn’t it make sense for annual committee meetings to be an opportunity to discuss the advisor’s performance, as well as the student’s, since both sides of the relationship need to be functional?  I can see huge issues with lack of anonymity, and revenge exacted in response to negative feedback, and so forth, and so it would have to be an optional process, and perhaps the advisor themselves could never be allowed to see the results directly; it might have to be filtered and time-delayed and anonymized to provide a safe context for the student to be honest.  Maybe the feedback couldn’t be given to the advisor until a few years after the student had left their lab.  Nevertheless, doesn’t this seem like something that ought to happen?  Such a process would let advisors who care about their performance receive suggestions and feedback from their students – feedback they are unlikely to get otherwise.  It would let departments evaluate the quality of advising of their tenure-track faculty, so they could choose not to offer tenure to advisors who made their students miserable.  And if these feedback forms were available online, it would provide an invaluable source of information for students trying to decide which lab to choose.

As the system works now (or fails to work, more accurately), my friend – who is not at McGill, I should state – is halfway through a PhD with an advisor with whom they are utterly incompatible.  They didn’t see it coming because they were unable to get a read on their advisor’s personality ahead of time.  Now they're stuck with a choice: another two years of misery, or an attempt to switch labs or even schools, or drop out and become That Guy.

Can’t we do better than this?

10 comments:


  1. Couple random thoughts come to mind.

    Another way to get an idea of the supervisor is to check their track record. Where do their grads end up? What's the pace at which the lab produces graduates? How many papers does a student publish, and how many/who are the co-authors. How many students are in the lab at any given time? Of course, this doesn't always work, much like your point about meeting students in the lab without the supervisor. It also doesn't work for new profs.

    Anyways, I also think this might be a slightly limited view since our experience is McGill and our colleagues at various institutions, and is also limited by most folks being primarily in biology. Granted, I know you know lots of other folks outside of bio. Grad school is very different in other departments/schools where the first few years are spent rotating through labs, so you DO get an idea of all the labs, get a chance to speak to all the students, the advisers, etc. For all we know, rotation-type programs or other grad programs might have evaluations and we are just un-aware. it also depends on if the degree is a research degree or a lecture based degree. Those degrees based on going to classes I would imagine have some type of evaluation.

    Also, evaluations are not always public, so even if you were asked to evaluate a supervisor, there is no guarantee you would get to see those evaluations as a method of determining if you would get along with the potential supervisor. I suppose there is always ratemyprofessor....

    Not being critical, just some points I'd toss at you if we were discussing this over a beer.

    ReplyDelete
  2. Yeah, great comments. My perspective is definitely McGill-centric. I have no idea how things work at other schools, really – except that I have several friends who are quite unhappy with their advisors, so however it works elsewhere, it does not seem to eliminate the problem.

    Checking the advisor's track record is an interesting idea, but I'm not sure how I'd go about it, really. How do you even find out the rate at which a lab has produced graduates (much less the number of people who have dropped out!), or where the grads have ended up? Unless the advisor posts such information on their website (which most don't), it seems like it would require an awful lot of detective work.

    Yeah, it kind of seems like there needs to be a "rate my advisor" website, like ratemyprofessor (which I am a big fan of, despite its obvious shortcomings). I googled "rate my advisor", but the most relevant website I found seemed to be deceased.

    ReplyDelete
  3. All I know is those bribes I keep paying to my current student seem to be effective in getting me new students!!!!

    ReplyDelete
  4. FWIW, I insist that prospective students visit my lab, and that their visit include spending a lot of time with my grad students without me around. And I tell my current grad students that they need to be totally open and honest with the prospective students. This is for my own benefit. Just as no student wants to end up with an incompatible supervisor, no (sensible) supervisor wants to end up with an incompatible student.

    ReplyDelete
  5. Jeremy, yes, that's the best approach, I agree; and students comparing various labs should probably be deeply suspicious of prospective advisors who don't have that attitude. I would note, however, that the same concerns also exist for postdocs, and the culture is rather different there. I recently applied for a raft of different postdoc positions, and several of them were interested in me, but none of them invited me to visit before accepting. In the end, I am about to move to France to work for a year and a half with someone I have never even met. Yikes! The evidence I do have suggests that it will be great; but it's really hard for me to know for sure.

    ReplyDelete
  6. And Andrew, I think you forgot to pay me my hush money. :->

    ReplyDelete
  7. Someday Ben. I was also thinking that advisors are not a one-size-fits-all thing. That is, the current students of a supervisor might love him/her because they have been self-selected to like the supervision style (Demigod, for example). This might be horrible for someone else. There isn't any magic bullet.

    ReplyDelete
  8. Definitely need something, as soon as you leave the area (e.g. for a job) advisors couldn't care less and can actually keep you from graduating. I've heard plenty of horror stories, and yes it leaves you bitter.
    You need a way to post your blog to facebook etc.

    ReplyDelete
  9. Wow if I only had read this at the start of my PhD. As it is I am in my 5th year and about to become that deranged drop out or apply to a new PhD program in another country.

    ReplyDelete
    Replies
    1. Ugh. Well, good luck. Maybe you can post a link to this article on FB or twitter – maybe it will save somebody else a miserable experience.

      Delete