Tuesday, February 12, 2013

Is the job talk portion of interviews a waste of time?

Via Daniel Drezner, an interesting attack by political science professor Dan Nexon on the job talk, which appears to be a central focus of major university hiring for political science assistant professor candidates:
The worst part is the job talk, which is an extremely artificial exercise bearing little resemblance to most of the professional activity of political scientists... 
...In fact, the job talk is most useful for… assessing the ability of a candidate to give a job talk. The reason we place so much weight on it * is that most academics (and I include myself in this category) are too damn lazy pressed for time to skim carefully read candidates’ portfolios. 
Professor Drezner demurs and says the following:
Dan's claim that it serves no purpose other than giving a job talk seems short-sighted to me.  In part, a job talk is an act of editing... A job talk lets me see whether this candidate will be able to talk to anyone outside of the five other people on the planet who know this specific topic cold.  
...If I've read the paper, I'm always curious to see how a candidate crafts his or her presentation.  And if the presenter can't hold my attention, that's a bad sign, because if they can't make their own work compelling, good luck keeping the attention of less interested students with work that's not their own.  
...Truthfully, however, the most important part of a job talk to me is not the talk, it's the question and answer session afterwards.  How well can a candidate respond to tough questions?  Stupid questions?  What are the reservoirs of expertise that lie below the surface?  In my professional experience, I can only think of a handful of candidates that blew their chances with the actual job talk.  I can think of a LOT of them, however, that deep-sixed their chances because they couldn't handle good questions.
Readers of this blog will recognize that Dan Nexon's negative comments about job talks could be made about both academic chemistry and industrial chemistry hiring interviews. The job talk is indeed artificial -- in the number of years since I left my postdoc, I've very rarely found myself in front of a group of strangers, wearing a suit and holding a laser pointer, while giving a artfully crafted 45 minute presentation while sounding relaxed, but energetic.*

I think that there are a number of things to be said for the job talk (or "interview presentation") for chemists. Like Drezner says, a job talk is an act of editing; it shows your ability to boil down the last 5 to N years of research into 45 minutes, your ability to ask interesting questions and to find out the answers. I think it's great, because, for the most part, you should have tremendous home field advantage over your audience -- after all, it was your own work.

I do think that the questions portion of industrial job interviews can be most illuminating of a candidates' ability to think on their feet and their instant recall of the fundamentals of their field (I can remember the moment, I think, where I sunk myself at the small company job I really wanted.) They can also be a time where the audience decides that they want to pick nits with your projects or play "you shoulda done this experiment!"

Are there interview tactics that are less artificial and more applicable to life in industrial chemistry?
  • Synthesis problems can be worthwhile, i.e. this is a molecule I would like to make -- can you talk about how you might make it? 
  • Scenario questions are fun, i.e. how might you scale up your 10-step total synthesis? Which steps might you focus your research on? 
  • I think there are probably legal strictures that disallow some sort of examination at the bench; it'd be interesting to see what people could come up with.
Readers, what do you think? Are job talks stupid? Can you come up with a better idea? 

*Maybe in sales?


  1. As a recent postdoct that got a "real" job in industry, I think the job talk is quite valuable for the evaluation of a candidate in both academia and industry. As a scientist, you will always have to explain your approach and defend your work, often times this will be done in front of people that are not intimately familiar with your research. Being able to put your work in context and do so gracefully may have significant bearing on whether or not you will get to continue it.

  2. Unstable Isotope tweets:

    In my experience, the seminar is the only opportunity some decision-makers will get to see the candidate. Very important.


  3. For an entry level position straight out of grad school, I think the job talk is very useful. But for experienced hires, it makes for an awkward experience trying to fill 45 minutes without divulging proprietary information and can be frustrating for the interviewee and the interviewers. One-on-ones and a solid resume review seem to be more suitable for experienced hires. If somebody previously managed to land a job at a reputable company, I usually give them the benefit of the doubt on their ability to perform.

  4. Lots of important things can be gleaned from a talk.

    -Obviously, pointed questions can reveal whether the person actually knows what they're talking about or if they're faking it
    -In addition to finding out if they can condense their research into 1 hour, can they organize them in a coherent fashion?
    -Did they stay within the time allotted? This is more important than you might think, it shows good planning, preparation, and in some cases an ability to think quickly and change the talk on the spot to fit into the right time.
    -Do they share credit?

  5. I think that in the interest of fairness every job talk should be preceded by a drug test. It'd be useful to know if the candidate needs propranolol, valium or a line of coke to get it going.

    1. Personally, I prefer giving the talk once in my underwear in front of the group, having them ask all the hard questions and comment on the fact that I haven't been to the gym lately, and then saving 1L of blood in a baggy. Then the morning before the interview, I give myself a transfusion in the hotel room.

      More seriously, it is the questions that always scared me the most. I always have friends ask me hard questions that an audience might ask, but the actual questions asked have always been different. Still, I'm really good at all this stuff, probably top 10% although I don't wear ties so it might count against me. So if I actually got an interview, I would kill on the job talk and questions part. So far no interviews though.

    2. "...probably top 10%..."
      Oh really now?
      I'd say coherent comments online are a good step towards killing your job talk.

  6. The job talk is most useful for assessing the ability of a candidate to get funding. If they can't convince a room of people and free food whose only investment is an hour of their time, how will they convince a grant board to give them $$$$$?

  7. The job talk is best way to break the ice and provide talking points for the subsequent one-on-one interviews. Everyone involved (except maybe the hiring manager) came into work that day and found 'Interview Candidate' in their outlook calender. If they read your CV five minutes before the interview, then you're already doing well. They want to ask technical questions to escape the inane HR 'tell me a story' questions but have no time to read up on whatever each of the three candidates does. Hence the job talk to provide some background and get the 'big picture' questions out of the way.

    And if they are hiring someone to be 'the expert' in X, then they are by definition not an expert in X and the candidate had better be able to talk about it to non-experts in a way they find useful and/or productive.

  8. CJ: "I've very rarely found myself in front of a group of strangers, wearing a suit and holding a laser pointer, while giving a artfully crafted 45 minute presentation while sounding relaxed, but energetic."

    Well, isn't this lecturing? And isn't the job for an assistant professorship? Seems valid to me.

    1. Fair point -- I think I was writing more towards industry.

      But I think that Nexon would say "That's not how I teach." Drezner would disagree, and say, "It's a good measurement of how you might teach."

  9. I think everyone should look their photos in the faculty profiles. Those pictures say a lot.

  10. At least they've done away with the snake fight portion for industry interviews, though academia still clings to tradition.