"I'm stumped!"

Every qualifying-exam candidate should say “I don’t know”

Photo: “I’m stumped” (sorry, I couldn’t resist that one).  Geocaching Hanley Park, by Martyn Wright via flickr.com, CC BY 2.0

As I write this I’ve just come from a great talk: Kim Hughes’ SSE Presidential Address at Evolution 2016 (Variety is the spice of life: death, sex, and the maintenance of genetic variation).  Presidential addresses are very much a mixed bag: sometimes a ramble from a superannuated blowhard with a big captive audience; sometimes a charmingly offbeat mix of reflection and experimentation.  This was neither – it was just a terrific talk.

Part of what made the talk so terrific was that its details were so clearly anchored to one of the Biggest Questions in evolutionary ecology: why we see so much standing genetic variation in traits that are tightly linked to fitness (like lifespan or fecundity).  Such traits should be under strong natural and/or sexual selection, and the operation of that selection should quickly erode variation.  Why don’t populations quickly converge on the highest-fitness genetic variant?  As a field we’ve been chipping away at this question for at least 50 years.  We’ve learned a lot, but we’re not done, and the Hughes talk brilliantly took us through major contributions from the study of colour variation and mate choice in guppies.

What on earth (you’re wondering) does this have to do with qualifying exams and “I don’t know”?  Well, precisely because the talk was built on one of my field’s most fundamental questions, it was just loaded with perfect PhD-student qualifying-exam* questions.  Here’s one: name and explain two non-adaptive and two adaptive processes that help maintain adaptive genetic variation?  (Mutation, drift; genotype-by-environment interaction, frequency-dependent selection.)  Here’s another: what’s the relative strength of selection on mutations acting at young and old age? (Absent pleiotropy and tradeoffs, declining strength of selection with age.)

And since it’s never too late to have nightmares about your performance on exams – even those you took 25 years ago – I got to wondering how many of these questions I could answer well if I were asked them now.  The answer: some, but certainly not all.  So if I (re)took my own qualifying exam tomorrow, I’d find myself answering “I don’t know”.  Or at least – that’s what I should find myself answering.  (See – I knew I’d get to my point, and I hope you had faith too.)

Every qualifying-exam candidate should say “I don’t know” at least once during their exam.  Why?  Because I can guarantee you that every candidate, no matter how well prepared, will be asked a question they don’t know the answer to.  That’s how we run these exams: we keep asking tougher questions until we stump you; and if we have to escalate until we ask a question that nobody knows the answer to, that’s what we’ll do.  That doesn’t mean we’re horrible people who want to see grad students cry; it just means that we want to assess not just what you know, but also what you do when you don’t know something.

What you do when you don’t know something is important.  There are two common behaviours (not just in qualifying exams, but in life).  Some people say “I don’t know”; others try to bluff it out**. In a qualifying exam, the bluff can be a disaster, because believe me, we’ll give you a whole bunch of rope to run with until you’ve tangled yourself thoroughly. In a qualifying exam, that bluff hurts your odds of passing.  In the rest of science and life, it hurts your odds of getting to the best answer, to the optimum behaviour, to the deepest learning, etc.  And in any context, sooner or later, it will leave you looking like a prizewinning chump.  All of us have things we don’t know. We often worry that admitting the end of our knowledge will leave people unimpressed, but it’s often the other way around.  People who can recognize and admit the things they don’t know can garner a lot of respect (at least from me); and admitting you don’t know is the only way you can take a step toward knowing more.

So: if you’re in your qualifying exam and we ask you a question you can’t answer, what should you do?  You can try to fake your way through, but it’s almost sure to end badly.  Better, you can say “I don’t know”, and leave it at that.  Or best of all, you can say “I don’t know, but if I had to find out, I’d do X” – where X is “look it up in this particular resource” or “ask this kind of person” or “run this kind of experiment”. Over the years I’ve had a few candidates give me one of those “I’d do X” answers.  Without exception they’ve been candidates I’ve been extremely impressed with, and who have passed their exams with flying colours***.

The world is a complicated place, and our efforts to push back the frontiers of human knowledge mean we’re often operating in unknown territory. If we don’t frequently run into questions we don’t know the answers to, we’re not thinking very hard.  Say “I don’t know” on your qualifying exam; and say it frequently before and after, too.

© Stephen Heard (sheard@unb.ca) July 21, 2016

*^Nearly every PhD program has some variation of this ritual.  It may be called a qualifying exam, a preliminary exam, a comprehensive exam, a candidacy exam, or something else.  It may be closely tied to the student’s intended research or very general (sometimes there’s one exam of each kind).

**^I know whereof I speak; I’ve been really, really awful this way myself. It’s still an urge I have to resist, although I claim I’ve gotten somewhat better about it.

***^But now that I’ve tipped my hand, it’s no fair pretending you don’t know just to impress me.



5 thoughts on “Every qualifying-exam candidate should say “I don’t know”

  1. Sean McCann

    This is also a good point to keep in mind for students at defenses. The best and most probing questions at a defense are not “what did you do?” but “given what you have found, what are the implications…Speculate on X”
    These types of questions is where I see less-prepared students stumble regularly. An answer of “I do not know” does not suffice in this case, as it implies the student has not fully considered the implications of their research. A well-prepared student will have seen pretty well all exciting directions the research could go and will have a ready explanation of next steps and possibilities.


  2. pdodonov

    Interesting post…
    How about a fourth option: “I’m not quite sure, by I’d say it can be this, this or this” and try to develop a line of thought which might indicate some answers? Which means basically proposing different hypotheses which may, or may not, be correct.

    Liked by 1 person

  3. Jeremy Fox

    Yes to all of this. Though one of my proudest moments in ecology was at the end of my candidacy, when Mike Pace said to me “The purpose of these things is to push students to the point where they don’t know the answers, but we never got to that point with you.”

    Related: https://dynamicecology.wordpress.com/2013/03/28/surviving-your-comprehensive-exams/. Protip: if a committee member asks you a question, you answer, and then the follow-up question is “Are you sure?”, the committee member may just be trying to throw you off. In which case the correct response is “Yes.” 🙂

    And just for fun: https://dynamicecology.wordpress.com/2012/04/16/hardest-or-weirdest-question-you-were-asked-during-your-candidacy-exam/

    That tooting sound you hear was our own horn. Sorry. 🙂


  4. Pingback: Summer flashbacks | Scientist Sees Squirrel

Comment on this post:

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.