You call it “wibble”; I call it “taking care of business”

Dr. Wibble’s astoundingly cramped view of what it means to give a good seminar makes it clear that there is need for further examination of this topic. Here I am lifting up out of the comments to this post my response to DrugMonkey’s correct conclusion that I was directing my advice there to trainees and junior faculty. This is an elaboration of my views on stage-specific purpose, and hence design, of a seminar talk.

Depending on the stage of your career, the purpose of a talk, and hence its design, will be very different.

(1) Grad student/junior post-doc: You are hewing closely to your data. No one knows who you are and has no reason to trust you or grant you the benefit of the doubt. Thus, you must stay close to the raw data, explain your controls and other experimental details, etc. You will get questions from your audience about things like the concentration of potassium in your internal electrode solution. You better be prepared to answer them. And any sorts of grandiose conceptual schemes you try to present will be ignored.

(2) Senior post-doc/job talk: Here, you are trying to convince people *both* that you have command of the experimental details of your research *and* that you have enough of a broader vision to lead a vibrant research program as a PI. This is tricky, and perhaps I will post a detailed analysis of job talks in the future. You will need to be able to field questions both about the nitty-gritty experimental details of what you present–mostly because faculty are trying to figure out what experimental techniques the applicant actually has command of and can implement in her brand new lab, and which were actually performed via collaboration–and how your research fits into a broader vision of your field. Detailed explication of experimental controls is still integral to your presentation.

(3) Junior PI: Here, you are mostly trying to convince people that you have some trainees in your lab and that experiments are occurring and producing results. You also need to convince people that you are successfully expanding your conceptual scope beyond nitty-gritty experimental details and focusing on the really big picture. At this stage, it can come across as a good thing if you field a question about how much sodium is in a buffer by stating that the questioner would have to ask the trainee who performed the experiment about that. Nevertheless, it is still good to describe experimental design in reasonable detail, including some discussion of controls.

(4) Late-stage pre-tenure PI: At this stage, you are really now trying to convince audiences of your breadth of vision and the growing scope and size of your research program. Your slides and presentation style are getting further from the nitty-gritty primary data, and more focused on the conceptual framework. Because you now have at least some reputation for being a solid PI, audiences will give you the benefit of the doubt about experimental design, such as controls and interpretation.

You can successfully field questions like, “How do you know that {boring artifact} and not {cool thing you care about} is what is really going on in your experiments?” as follows: “We have performed a complex series of control experiments that rule that possibility out. Since those experiments aren’t otherwise particularly interesting, why don’t we discuss it afterwards if you are interested?” Senior post-docs and junior PIs can’t usually get away with that.

If you have multiple research programs going on in your lab, it is important to present work from more than one of them. You have the time to do this, because your scientific stature at thus point in your career permits you to explain the science more conceptually, yet still be convincing. And this helps demonstrate to your audience–some of whom will be writing tenure letters for you–that you do, indeed, have a broad vision and expansive scope of science and will be a major contributor to your field.

If you have more research programs than you have time to talk about, it is useful to somehow allude to the existence of these other research programs even if you don’t described them. This can further impress the audience with the vast scope of your vision. Ways to do this include when mentioning the people in your lab, or in the very beginning of your talk you can state the science of your lab as broadly as possible and then say, “Today I’m going to tell you about a subset of what is going on in my lab”.

(5) Tenured PI: At this stage of your career, you can do things like openly embrace controversy with your peers–which you need to be more circumspect about pre-tenure, take a very conceptual and speculative approach–so long as it is interesting and provocative, and talk a little more sweepingly about the historical currents in your field and how your research has both responded to and guided those currents. But you really can’t just give a retrospective about your own life in science.

(6) Nobel laureate: At this point, your talks will be all you, all the time. How did you end up doing the things you did? How did your research come to so decisively dominate the course of an entire field? You can show pictures of your dog or cat. You can show pictures of each of your many trainees who have gone on to become leaders of their subfields and members of the National Academy of Sciences. You’re a Nobel laureate; you can do whatever you want.