34 thoughts on “ask a scatterbrain: job talk”

  1. 1. be organized
    2. know your audience (know their central research questions, know their biases, predict and pre-answer 5 easy questions and 5 hard questions, don’t give a research talk for a teaching job, or a teaching talk for a research job)
    3. know your research and have it ”finished” so you don’t have to provide ‘excuses’ for the point your questioner knows something.
    4. Practice your talk (a job talk should never be given for the first time to new dept., give it at a conference in a different form, develop it)
    5. try not to worry, be overtly anxious… instead try to ‘own’ the room (come in, know how to set up any technology you need, figure out how everything work in a competent fashion, but in the end allow anyone that wants to help to help in a gracious fashion)


  2. I’m sorry to be pressed for time, but I can just cut-n-paste a list from a longer document that was drafted during my time at Columbia (with help from John Krinsky and Gina Neff) and has since been updated…
    (nothing too brillz here and much is redundant w. buridan’s comment)
    1. Make a PowerPoint presentation.
    2. No more than 17 slides for a 45-minute talk.
    3. Make them beautiful.
    4. Do at least 2 practice job talks in the department—the first with a sympathetic but honest audience, the second with the most general audience you can find.
    5. Memorize the talk but speak from your notes.
    6. Watch your body language. Avoid nervous ticks (“tells”), shuffling, wild hand gestures, and bizarre facial gestures.
    7. Eliminate “ums,” etc.
    8. Project and speak clearly and slowly.
    9. Expect that your “real” talk will be at least 5 minutes longer than your practice talks.
    10. Produce “red flags” in your talk—make your own questions.


  3. Wow! Good question – IMHO:
    – Substantive talk presenting real research with clear importance, including to sociologists not in the subfield
    – Concentrating on argument and evidence, *not* technical details
    – Offers context on relationship between this project and candidate’s planned trajectory
    – Able to discuss intelligently “on her feet” during Q&A


  4. Great responses above. One thing I’ll add is: be enthusiastic about your research. If you are not excited about/engaged with it, I don’t know I should I be. But strangely, I’ve seen this a number of times.


  5. Another: provide an 11×17 (folded) four-page brochure with your research summarized in a crisp, preferably heavily graphical leave-behind. Edward Tufte has written a bit on this. Often people will protest, “but my research is too complex to fit into four tiny pages!” I once saw someone do a detailed analysis of Chinese nuclear policy in this format, and it was brilliant. Big benefit: as soon as a slide leaves the screen, it’s gone. With a brief brochure, people can take your presentation with them to think about and form new questions and ideas.

    One more thing: these take longer than you might think to do well. Crisp and concise is hard, but well worth the effort.


  6. Make sure that the takehome message of your talk has two points: 1) your main finding and 2) the contribution it makes, and then make these points so clear that everyone in the room walks out the door with these in mind.


  7. One thing I’ve found is that ‘exciting and engaged’ is hard when you are heading down the end-stretch of the dissertation. By then you know your dissertation research, you know or should know all of its errors, problems, foibles, and are probably a bit embarrassed and wondering how it is going to pass, etc. Though it will pass, i’ve never heard of anyone failing that maintained good relationships with their advisors. I frequently encourage colleagues in this position to present a paper they have under review or an upgraded conference presentation. To present something they really would love to be working on, though they have to work on their dissertation. (There is likely a bit of granny ogg’s ‘headology’ in this approach.)


  8. Yay! Ask a Scatterbrain is back! Thank you all for your helpful advice on this and so many other topics, it has been incredibly helpful figuring out this strange world of academia.


  9. if you have the choice between presenting older (published) but polished work or new but still not finished work in your job talk, which one is better? Is it better to give an overview of your dissertation or to present a small piece of it?


  10. @10.socfreak, my 2cents: it’s better to present solidly finished work, but published is worse than not-yet-published. The ideal work is complete but not yet published. And I don’t think the overview approach is good – better to give the audience something specific to discuss and consider.


  11. @10: What AP said, and re: the overview vs. small piece. Present the small piece, introducing it with one slide of overview, so the audience can see how it fits in, what else you are doing. Optional: end with a bigger picture piece, showing how this finding will be developed further, will complement other pieces, begs another question you answer in another chapter, etc.


  12. My 2 cents. OK, inflation: 10 cents.
    1) Talk to the people where you will be giving the talk. The optimal talk is different at different places. Cross-examine the people you are talking to about who will be in the audience, exactly how long the talk window is and the expectations about the balance between your talk and the Q&A period, and whether they prefer a talk that goes into the nuts and bolts of a particular project or a talk that is geared to a more general audience. At my institution, we want the nuts and bolts paper: we really dig into method and argumentation. But some schools use the job talk to see how good a teacher you will be.

    2) Plan a talk, don’t read a paper. Thin the literature review to the bare essentials, the main points that are needed to understand what you are doing and tell people the full lit review is in the paper. Unless, of course, a major part of what you are doing is a critique of the literature. Most of the time, your methods and results are the things you should be focusing on. However, you DO need to be able to explain why this is an important and interesting problem.

    3) Talk about something that you know more about than you put into the talk. Have new things to say in response to questions. (It is particularly impressive to have slides prepared to answer questions that come up in Q&A.) Know what you are talking about in depth. If your best work is something that is already published or in press, I’d say it is ok to go with it rather than present something you are not fully comfortable with, but I’d try to have some new results or insights that are not in the publication. Perhaps begin with the published results and then some discussion of “what next,” especially if you can present some preliminary results not in the publication that tie to the what’s next.

    4) Be able to give a clear and accessible-to-nonspecialists and enthusiastic explanation of what the research problem is and why it is interesting. Beyond that, do not talk down to the audience. Be as clear as possible and be prepared to explain if, for example, people don’t know what regression coefficients are, but don’t alienate the audience by explaining simple stuff they know. (This is why interrogating people about audience is so important)
    5) Make sure you understand what technology will be available and have a backup plan. If you do slides, make them legible. For complex statistical results a good compromise is a visible summary of the main results on a slide accompanied by a paper handout with the details.

    5) If you have gotten bored about your dissertation, get un-bored. If it exists and you’ve moved on, or gotten to the stage where you just cannot stand to read it any more, take a short hiatus and then re-read it and remind yourself about why it was important. You will be asked about it repeatedly in the interview, and if you sound like you don’t know what it is about, it is extremely unlikely you will get the job. If the problem with your dissertation is that it does not exist yet (not even one conference paper or chapter), then talk about something else, but don’t be surprised if you don’t get the job.


  13. Would people say that the same suggestions apply to qualitative and quantitative research presentations? I ask as some of the advice clearly has quant methods in mind. Or any specific advice for qualitative presentations?


  14. Wow, great info to squirrel away, especially about what kind of thing to present.
    I’ve never had the honour of hearing or doing a job talk, but I would say for qualitative presentations – minimize the number of words on a slide. Less is more. Quotes from interviews just by themselves, or with a picture, work well (as long as you read them out and explain why they’re important).
    Worst, worst quali presentation error I’ve seen is when the slides and the talk don’t fit each other. You’re too busy trying to figure out why a given slide is up at this point in the talk that to listen to what the person’s saying. But surely no-one does that at a job talk?!


  15. I think the same principles apply qual/quant. Evidence is key, but the technical details aren’t (access issues, for example, or excessive self-reflexivity in ethnography).

    A slide show has become de rigeur, whether PowerPoint, LaTeX, or whatever, but heed this warning.


  16. I’ve never seen a pure quantitative presentation get a job, but I’m not sociology, i just read sociology journals and publish in its periphery. I’ve seen people who used methods appropriately get the job, but they were never only quantitative, always mixed method of some sort and almost always had a bit of history or theory. So, i’d say my recs cut across several disciplines and department types.


  17. I think it depends on what one means by theory, i can think of two ‘theories’ that are ‘methods’ and I can think of quite a few methods that are based on theories (actually i’m not sure that i can think of a method that does not rely on a theory). In any case, I agree that something that is theoretical is necessary, but I think my point was to say more that. I’ve never seen anyone that did not have research narrative that included their theory or at least the historical genesis of their research question, do very well. I have only seen two job talks that just leaped into hypothesis and then started mangling quantitative evidence.


  18. I’d like to second andrew & buridan re theory, especially since my comments could have been construed as downplaying its importance. I consider theory to be central to why your problem is important, and also is about what mechanisms or processes you think are important in the phenomenon you are studying.


  19. @21: The role of theory, or even what counts as theory, is something that I have a hard time communicating with graduate students about. Any tips on how to help grad students grapple with theory?


  20. Re theory, that’s a hard one, as people disagree about what they mean by theory. For me (and for most sociologists), it isn’t tying everything to some classical or modern social theorist, and it isn’t a propositional inventory. And I said to keep the “literature review” short. I personally rarely say to myself, “There was no theory,” although I’ve heard that said, and I have colleagues (especially demographers) who defend the value of “atheoretical” research, although they almost always actually have some theory in the way I think of it. The opposite of having theory is just talking about the relations between variables or just describing your qualitative results with no sense of context. What I want to hear is some deeper discussion of mechanisms and processes — what you think is going on to produce your results — that draws on and is linked to other thinking or research about related social processes. That probably does not help you explain it to your students.


  21. Re: theory…. I think that the audience, or at least those informed in your field should be able to say very clearly that your work is in tradition x and informed by said authors a, b, c, and that is just theoretical context and part of the research narrative. However, more deeply, there should be engagement with the thoughts of a, b, c, that demonstrates that you understand them and are misapplying them, or attempting to fit data to them that doesn’t fit their theories.

    One clear place theory is important is publishing. It i easier to publish in some journals when you have mastered a theoretical framework and are doing interesting work in that framework. This is also why, when someone gives a presentation using theory x which they indicate they frequently use, which traditionally has a home in journals, a, b, c, but they only have published in f and g, I get suspicious and pull their published papers to read. Granted, people do use frameworks and change perspectives, but… sometimes people disingenuously try to over-fit or match(but to strongly) their research to the audience.

    Theory then, for me, is the groundworks of your career, it is what I look to for the base and legitimizing narratives that ‘make sense’ of what you are doing. (granted other people will look to other parts of your narrative, such as skill in methods, appropriate application, use of data) So, if I don’t see the theory, i get this uneasy sense of that your research is meaningless, which sometimes I get for other reasons related to people’s misunderstandings of philosophy of social science too, but ehh, usually those are resolved with ‘does s/he get published’, if yes, then someone must find it worthwhile and i let it go.


  22. Theory, to my mind, is about the conceptualization process: what do the data *mean*, and why are you confident (or not) about their ascertainment? Data do not enter the world as data; they become data in a particular context (to paraphrase Marx rather poorly). Theory is the description and consideration of that context.


  23. I am enjoying this post and its responses quite a bit. A few items from my own personal experience, nothing terribly out of line with what others are saying.

    technical stuff:
    – I had transparencies, which would now be keynote slides (and hopefully beautiful). I also provided a handout packet with a few of my slides on them – not a 4 page brochure, but something to give for people to write on.

    – I had a qualitative thesis, and I used illustrative quotes from interviews and fieldnotes, and then spent a bit of time connecting them to the conceptual argument I was making. I bolded the critical part of the quote, and then treated it as data in the talk. I read the quotes with a kind of explanatory commentary, but I did not read my slides otherwise.

    – I talked about ‘theoretical linkages’ without talking about a literature review. My work connects to two sets of literatures, one an economic soc that treats markets as outcomes rather than taken-for-granted mechanisms, and one that sees markets as a set of organizational/cultural activities. I had all the literature in my pocket, but I moved directly to my own argument(s).

    – I wrote out the text of my talk, and practiced it enough that I could follow it without reading it (ie points made all over about being prepared)

    When I gave the talk at a couple places, I realized they were misinterpreting it. They wanted to talk about deviance, or ask me about how markets work. This was data that my talk wasn’t right, but also that the place wasn’t right for me.

    More generally:
    – My feeling is that schools want to know three things from junior candidates. 1) will they be finished; 2) will they be good/productive/central going forward; 3) will I want to be around & work with this person.

    – This means being knowledgeable & assertive without being a jerk. For Q&A, recall that questions are occasions to talk about your work, not just questions to be answered. And I would try answer these questions crisply and not ramble (unless there are not many), better still to connect multiple questions to themes you have already thought about.

    – Your talk will become one more piece of ammunition for your supporters to advocate for you, and for your detractors (or supporters of others) to advocate against you. You know that quants will complain about quals being not generalizable, that quals will complain that quants are just technicians, etc. Give your supporters as much ammunition as you can for them to convince their colleagues of your wonderfulness.

    – Finally, if you are doing a job talk, recall that this means they have invited you at time & $ expense for a job talk. No one wants a job candidate to fail at this point, they want you to be great. So go, learn about the departments you visit, ask questions about your potential colleagues’ work, and don’t try to be someone you aren’t willing to enact over the next 3-5 years.


  24. Three basic suggestions:

    1.) Make it interesting to all sociologists, not just people who work in your area.
    2.) Make an actual argument. Don’t just describe what you did.
    3.) Get to your data. GET TO YOUR DATA!

    I can’t tell you how many talks I’ve been at where there was 20 minutes of framing. So-and-so said this. So-and-so said this. So-and-so said this. So-and-so said this. So-and-so said this. So-and-so said this.

    Annoying? Most certainly. Get to your data. Tell YOUR story. What you did. Why it is interesting. I think it is a mistake to structure a talk like an ASR paper. People pay attention for about 5-7 minutes, in the beginning. So catch them then. Tell them something interesting early on, and tell them what your argument is going to be. And then get to your data!

    Did I mention: get to your data?


  25. A few small suggestions for delivering the job talk:
    1) When they’re making your schedule, request 15 minutes for yourself before the talk. If your talk is at lunchtime or in the late afternoon, bring a granola bar or other snack and eat this during the prep time so that you don’t run out of energy halfway through your talk.
    2) Ask about the room set-up. Will there be a podium? If not, have a plan as to what to do with your notes.
    3) Make sure you have a pad of paper and a pen to jot down key points of the question as they are being asked.


  26. I agree. Some other technical points, esp. for qualitative talks:

    * Let’s say your talk covers 3 main points that support your main argument. I recommend creating a Powerpoint slide with a very short version of the main argument and those 3 points. Bring this slide up near the beginning of the talk (“here’s what I’m going to talk about today”). Return to this same slide with each new point (“now, moving onto the second point that supports my argument…) , so the audience can keep track of where you are. You can show the slide again at the end quickly (“here’s what I told you today”). This is public speaking 101, but the slide helps a lot. In a conference presentation, you don’t need have to worry about this kind of thing. But a job talk is much longer and it’s easy for people to get lost in the details, esp. quotes and field notes.

    * As Jeremy Freese once advised me, a good qualitative talk–esp. an ethnographic talk–has at least one moment where the audience can imagine you in the trenches. Try to subtly work this into your talk somewhere, but obviously it needs to relate to your argument. (This is different than self-reflexivity.)

    * It is a little counterintuitive and different than most journal articles, but here’s one way to deal with the literature: at the beginning of the talk, use some citations to frame your central construct/argument. Also have one *very* brief slide that outlines a few major approaches to this topic and where your argument fits. Then, in the conclusion, talk more in-depth about your contributions to one or more literatures and how you’re moving those forward. Here’s why: In the beginning, the audience is antsy to get to the data/findings. At the end, the audience is ready to hear more about your theoretical and empirical interventions.

    No matter what–hopefully, this board is sounding repetitive!– keep the lit review short. For a 45 min. talk, you should be past methods and onto your findings by minute 12 or 15 at the very latest.

    * For any talk: the Q&A can make or break a candidate. Don’t seem defensive! Write up answers to common questions and practice giving those answers, just like you practice your job talk.


  27. Thank you!

    My job talk was the very last thing at the end of a two-day interview process, so I’m not sure how coherent I was by then. But here’s what I learnt:

    – Do follow @13olderwoman’s advice to find out as much as possible what kind of talk is expected. I was told something like “we want to know about what kind of scholar you are, what you do and why you’re excited about it”, which framed it nicely, I thought.
    – Do practice in front of real audiences. I planned two ‘dress rehearsals’ with slightly different audiences (including people in my field who know me to say hello to, but don’t really know what I do). I definitely needed both run-throughs: the first to discover that I was pitching/structuring my talk all wrong, and the second to get feedback on the major revisions!
    – For that reason, don’t work your talk to perfection for the first practice, because you might then change it a great deal.
    – Re. shakha’s and ellenhere’s points about getting to the data: I’m an ethnographer, so I started off with five minutes ‘in the field’ with lots of single slides of photographs, forecasting the theory and analysis, as it were.
    – Do practice a lot at home with the ‘real’ physical performance (e.g. standing up with appropriate level and pace of voice). Get your body to memorize the talk as well as your brain.
    – Do follow @1buridan’s excellent advice about predicting 5 hard and 5 easy questions.
    – Don’t use a lectern/podium if you didn’t practice with one, it will make you feel as silly as I did.
    – Don’t be put off by the scary expressions of people in the audience. They’re either concentrating or thinking about what to have for dinner. They might even be “trying not to look too enthusiastic” (as one listener put it!).
    – Don’t be surprised if some people get proprietorial about your subject. I realized (belatedly) that many people know the city where I do research, so they often have something to say about it that may or may not be relevant to what I want to say about it. I wasn’t quite prepared enough to spot the difference between questions that stemmed from their ‘ownership’ of the place, and questions that really related to my research, but I’ll know for next time.

    I also had a couple of strokes of luck. One was that a special journal issue on my subfield came out just before the interview, so I was able to quote the well-known guest editor saying why my kind of research was important, which my practice audiences thought was a nice touch. Presumably you could create a similar effect with older material. The other was that the person who chaired the question period did so in a very sympathetic way, with a light touch (it was a Friday afternoon, after all).

    Thank you scatterbrains!


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

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

Google photo

You are commenting using your Google 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.