Writing sample as a job-talk?

In the comments section of my post, "Job Market Prep Questions?", Anonymous Grad asks:

Let's say I am fortunate enough to get a campus interview and give a job talk. Should I plan to prepare a job talk that is different from my writing sample?

As luck would have it, we actually discussed this topic almost a year ago! There, the consensus seemed to be:

1. If you can give a good talk on something other than your writing sample, you should do it. Giving a talk on your writing sample can make you look like a "one-trick pony."
2. However, if your writing sample is far and away your best piece of work, give it.

At least offhand (and I may be wrong!), I'm inclined to think that one should err on the side of (2). There's an old saying in baseball, "If you're going to get beat, get beat on your very best pitch." When you're giving a job talk, you want to be sure to absolutely kill it. You should give your talk on whatever piece you feel like you can hit a home run on, even if it's your writing sample. I personally know at least one person who scored an R1 job giving their writing sample as a job talk. If it were me, I'd rather lose the job knowing I'd presented my very best work rather than my second-best. But, perhaps this is just me. 

Am I wrong?

One final thing to note is that, as Rachel pointed out in the comments section of last year's post, it may be a great idea not to present a paper as your job talk at all, but rather "3-4 talks in one", giving a "big picture" presentation of your research program, along with how everything fits together, thus giving the people in attendance at your talk a really impressive presentation of the many things your research project promises to do in years to come. After all, you need to remember, (1) most of the people in attendance at your talk will not be on the search committee, but (2) their opinion on your job talk will count in the hiring process. While the search committee members may (or may not!) have read your research statement carefully, a lot of people in attendance--other faculty members, grad students, etc.--may know nothing about you.

Indeed, I was in a one-year position at a research university, and when our department had people in for job talks, their talk was the only thing most of us knew about the candidates. I can tell you--in no uncertain terms--that if someone came in and gave a kick-@#$, "big picture" presentation showing us how amazing their entire research program was, I would have been blown away. Too many of the job talks were on really specialized topics that I had no knowledge of, and which left me pretty disinterested (sorry, but it can be really hard to get interested in a problem that you have no background in!). In any case, since the job talk is likely to be your first (and maybe only!) impression on grad students and faculty members not on the search committee--and these people do play a role in who gets hired--I think Rachel's idea of giving a "big picture" talk is brilliant...if you can pull it off. :)

Comments

[Anonymous Grad]

Thanks, Marcus. SOO helpful!!

[Lee Walters]

I agree with Marcus about 1 and 2 above. But on where to strike the balance I disagree. But I suspect this will just vary from person to person. So my advice would be to make sure you have two good pieces.

A few comments to add to Marcus’s advice from one UK perspective:

For me, giving the same talk as your writing sample is very much a second best option, at least for a research job. It leaves the panel wondering whether this is all you have. I know of cases where this has definitely gone against people.

I for one would not be impressed by a big picture presentation and I know of cases where giving programmatic talks has gone against people. This might partly be a matter of the personality of the selection board but is also partly to do with expectations. In the UK research plans/direction is discussed at interview.

Follow the instructions you are given, or at least check if you plan to deviate. Part of the reason I wouldn’t advise giving a big picture talk in the UK is that this is not what you are asked for.

In the UK, usually the decision on appointments is made by a selection committee of around 4-5 people. The rest of the faculty may provide their opinion to the committee after having attended the job talks, read some of the written work, and looked at CVs, but the decision is not made by them. So the fact that the job talk is the only aspect of you that the majority of the faculty are familiar with is not so important.

The job talk is only one part of the portfolio that the committee look at. Having said that, I think there is some tendency to give disproportionate weight to it. Nonetheless, I know of plenty of cases of people who didn’t give the best talk getting the job.

[Marcus Arvan]

Anonymous Grad: :)

Lee: Thanks for weighing in. I do think it's a good idea to follow directions/check if you pan to deviate. You're also right that people differ. *I* would be impressed by a programmatic talk, you wouldn't. The question then is: what are the probabilities? Are people more likely to be impressed by a programmatic talk, or a specialized one? I suppose the answers here depend on the field. All I know is that I have seen some ultra-specialized talks first-hand that went way over everyone's heads, and in a way that did not seem to impress many people.

[Tom]

Having served on a R1 SC twice I can relay that it counts against candidates if they present their WS as a job talk, other things equal. This is just because many, myself including, expect that a candidate suitable for eventual tenure at an R1 would be able to come out of grad school with at least two papers that indicate a coherent, promising research agenda. Having only one presentable papers raises worries about productivity and narrowness.

[Lee Walters]

Hi Marcus,

Yes ultra-specialized talks are bad too. But there is a big sweet-spot between those kind of talks and overview of my research talks. I want to see a thesis and I want to see it defended, and yes I want it framed in such a way that my colleagues and I can understand.

[Chike Jeffers]

I agree with others in the comments that it should be considered unfortunate if you end up not having anything you feel you can present as a job talk that is different from your writing sample.

The job talk, by the way, should of course be accessible and not ultra-specialized... as compared with the writing sample, which can be more specialized because it shows just how deep you are in your field.

[yasmin]

Hello, Marcus.
This artcle is good and very helpful for us. Everywhere, we can get positive and negative comment but we have to follow only positive way. It gives me positive feedback to go achead. Thank you so much for this posts.