Giving a (Job) Talk: Notes from the Field

November 11, 2013

By Arthur Spirling

There can be few activities more anxiety-inducing than going on the job market. Once on the market, there can be few activities that are more anxiety-inducing than giving a job talk. This post is an attempt to share information and advice, and hopefully reduce the stress that many feel.

Given its importance as a decider of fates, it is perhaps surprising that so little attention is paid to teaching students about the job talk’s value, structure and content as they work their way through grad school. What follows is my advice, some of which I was given, and some which I learned on my own, or following discussion with colleagues: the usual caveat most definitely applies. Inevitably, different people will have different thoughts, and students preparing for the market should ask local experts for input. Quite who those “experts” actually are should be decided on a case-by-case basis, but here is a pro-tip on something to look for: the ‘expert’ has given a successful job talk (recently). Relaxing the conditions slightly, if you have never seen them give an interesting or intelligible talk about their own research such that you would wish to emulate their performance, they should probably not be your primary source on “what works” for closing the deal at the university where you seek employment.

In my current role as a faculty member, I’ve given several talks on talks to students and others, and have prepared a slideshow that discusses everything from the nature of the talk, its typical contents, how to display data, how to deal with questions, and so on.  To be clear, I come at this problem (mainly) from the perspective of a  political methodologist seeking a job at a research orientated institution, but I think the contents will helpful for most candidates navigating this difficult process.  You can find my slides here

For those looking for quicker tips on strategy, here are the five most important general points I wish to convey:

1. It is a job talk, not a paper, not a recitation, not a performance, not a lecture.

Though it is tempting to ‘script’ your talk, this is always bad news: candidates who have memorized (to the letter) what they “must” say on each slide come across as stilted and unintelligent. An even lower opinion results for those who have notes that they read for each slide. As an audience member, it is very hard to understand why speakers don’t know their own work well enough to discuss it in an unaided way for the 45 minutes or so that a typical talk lasts. Sometimes, candidates think they can get around this problem by packing their slides with text such that they can simply “read off” their presentation. This may be worse: see next point(s).

2. Visuals matter, so think about your slides from the perspective of an audience member.

If you have to use the phrase “I’m not sure if everyone can see that” because your table or font or diagram is too small, with probability 1 you have a “bad” slide. In methods talks, this seems especially common when candidates present regression results. If you must have a regression table, have it as a back-up, and go to it only if someone requests the whole thing. Very generally speaking, clean slides are better than busy ones, and slides with less technical detail are better than those loaded up with Greek letters. If you must use equations, be sure to introduce them intuitively and explain everything with words rather than symbols. If you are not sure how to write nice slides, go to other people’s talks and think about what worked and what didn’t in terms of the presentation. It’s fine to use tricks or displays that you’ve been impressed by.  If you struggle to know what should and should not be included in your presentation, ask yourself “what question does this slide answer?”  If, in fact, you can not think of the corresponding interrogative, you probably have a slide that should be edited— and maybe even removed.

Remember: it is often difficult to be succinct rather than verbose, but it will make the world of difference to an audience.

3. Take questions as they come up: don’t say you will only be taking “clarifying” questions, or that you will deal with “substantive questions” only “at the end”, or something similar.

When you give a talk, you are attempting to showcase your research to the audience, along with your ability to teach and to be a good colleague. Luckily, I have lots of good colleagues; tellingly, not one of them has ever asked me to wait 40 minutes while he or she completes a monologue before I am allowed to seek clarification on something they’ve said. My point here is that audience members have every right to question your work, and they have the right to expect an answer. If the next few slides (or one in reserve) will make something clearer, simply say that. If you want to give a short, helpful answer and return to the issue later, say that.  If you haven’t thought about the issue, try to say something intelligent that makes it clear you are intellectually flexible. It is a dull experience when a speaker is so uninterested in the audience’s comprehension of the material being presented that no one may seek to check their understanding along the way. Notice further, that you should be willing to take questions from everyone: grad students, junior faculty, senior faculty, visiting faculty etc. The committee that hires you wants to see that you can “speak” (literally) to the broad range of people in a department—certainly, most committees will solicit feedback from all their colleagues as to the merits of the candidate.

It is worth noting that candidates often, perhaps naturally, worry about getting caught off-guard: that a question will be so devastating or complicated that they will never get their talk back on track. This is unlikely to be the case if you practice your talk and Q-and-A (see next item). Moreover, never forget that you really are the expert on your own work: you have spent years on this project, and no one knows it as well as you.  Subsequently, the way that you craft your presentation, including what you seek to focus on and how, will go someway to funneling questions in terms of their topic.

Precisely because you will take questions, you should budget the time for the talk accordingly: it would not be an unreasonable goal for the talk to run around 35 minutes straight through, expanding to 45 minutes with interruptions. Quite how many slides that requires is up to you, though spending around 1.5–2 minutes a slide is a good (starting) benchmark.

4. Practice, practice, practice.

Your job talk should be prepared as early as possible: two weeks before the market opens (so, around the time of APSA) is not unreasonable for a first draft. Then, you should practice its delivery. A lot. For a grad student, practicing the talk everyday they are on the market is prudent: simply being in a room on your own and running through the slides out loud is a good start, but it’s even better if you can have people (grad students, friends, junior and senior faculty) there to give you feedback and just generally simulate an audience. I have found that practicing goes a long way to reducing anxiety for me, and for students. It’s fine, and even good, to be nervous about doing the best job you can: but once the talk begins, and you see the (by now very familiar) slides, you will hopefully find yourself relaxing and (even!) enjoying yourself.

5. Don’t obsess over how the talk went: get on with business

Under the stress of giving a talk (and the relief of having completed one), it is natural that your mind will obsess about how the audience responded. In truth, this is very difficult for the speaker to tell: there have been occasions (including times when I was on the market) when I thought the talk had gone well, though it hadn’t. Similarly, there were times that the audience seemed listless, quiet and unreceptive, though they had (or, perhaps, the key members had) actually enjoyed it very much. In the moment, it is much better to stick to the mantra that you did your best, and you now need to move on the next item on the agenda: typically one-on-ones with faculty, a lunch, or a dinner. It does no good to over-think these things after the event— except to learn for next time.