Thursday, October 30, 2014

How to design your academic job talk

These are the days when most of the faculty candidates who are fortunate to get invited are designing, and giving talks. I thought it might be a good idea to offer advice to those candidates since I am not sitting on the other side of the aisle. Mind that I used the same techniques to give talks.

1) Designing a talk for a broad audience is different from giving a talk in your department. The department that you are giving your talk in probably does not have any one who works in your sub-field (hence they are looking for a  candidate who could do what you are doing). Therefore, you must assume that the audience knows very little about your sub field, the jargon that you use, and how good you are in your sub field. Therefore assume that the audience knows nothing !

2) Following from argument 1; You do not, and should not cover everything that you have done till now. If you are invited for a faculty interview there are at least couple of large projects that you have worked on. Pick the one that is most awesome (in terms of results) and is most likely to be fundable in the future as well.

3) Following from argument 2; After picking up the awesome-st project give a high-level outline of the research. That means a lot of cartoons and diagrams. Yes, people are sleeping and they are lazy. It is your job to make them understand what you have done. The best way to make them understand is using diagrams (fancy diagrams) and animations !. The audience will keep engaged and will likely to understand something from your talk. They will feel smart and hence will like you.

4) Show them some results. Then tell them what the results mean in the larger context. Telling me that you imroved the time of the algorithm or the purity of your sample does not means anything to me. Tell me that your improvement of time results in blah blah blah. Otherwise the results are meaningless.

5) You have published your papers; that means that you have proposed and did something useful that other people have not. Tell me by comparing your results with that of the competitors and show why yours are better.

6) Have multiple slides on your future work. I want to know that you have a solid plan to take your research ahead and that you know where to target for your funding.

7) Do not have text after text in your slides and then to explain to me what is written. The audience can read the text; they are here to listen to you.

1 comment:

  1. Good points. I would like to point out though that #5 would be restricted to methods/techniques talks. It would be strange to tout one's results as "better" otherwise. Of course, the candidate could stress on what's "newer" rather than "better" in their findings to highlight their contributions toward moving the field forward.