I have the opportunity to give an unusual type of talk to my fellow postdocs and graduate students. The invitation is explicitly to discuss both our recently published scientific findings and the process by which we developed them. How did we decide to pursue this line of inquiry? What problems did I encounter, and how did I solve them? What was the biggest challenge of this project, and how did I get around it?
I think it's great that my postdoc association is encouraging talks in this format. Discussion of career issues, as I think you'd all agree, is critically important for scientists at all stages, and this talk provides a forum for "what best serves to advance your science" sorts of questions. So I was happy to agree to give the talk.
But because the talk is non-standard format (and because a reasonable portion of the audience is expected not to be in my particular subfield), I can't just recycle the powerpoint from previous talks.
What do you think the best set-up for such a talk is? My tentative plan is to spend the first half (~15 min) or so giving more or less my regular scientific talk, albeit with more explanation and less detail to accommodate the non-MyFieldists in the audience. Then I want to give the talk again, but in the order of how it happened, and what actually motivated our steps from one experiment to another, rather than the standard scientific Well Thought Out Post Hoc Justifications that I use in the first half.
That is, in contrast to the first half of the talk which will begin, "This system we study is of vital importance and we wanted to understand the mechanisms underlying one of its most salient characteristics," the second half should start something like, "I had finished up one project and was just messing around with ideas for another project, and really had no idea what I wanted to do...." etc.
For the second half, I'll put the figures in the order they were actually done, and I'll highlight the point at which the project turned the corner from Random Observations in Our Corner of Science to Hey We Have a Cool Paper Here (brilliant feedback from a member of my thesis committee, as it happens.)
What do you think? What would you most want to hear about in a "how it really went down" sort of talk?
15 years ago
12 comments:
Hey, that's a great idea for a series of talks! I like the idea of giving the polished story and then getting the "real story", kind of like a behind the scenes special.
I think it would be great to give trouble shooting examples that people could use for themselves if it applies (something like, I worked on this series of experiments using this standard protocol and got nothing, but then one day I decided to tweak this one variable because I heard something in a seminar and suddenly it started working...). Also I think it helps to really emphasize the connections between the initial observations and the hypothesis and why, of the many different experiments you could do to test it, you chose to show the one you did. Also, what are the caveats behind the experiments and how you can address those issues. Did the reviewers from the paper come up with anything you hadn't considered and how did you address that?
But that's a pretty long list for a half hour talk! Hope it goes well!
That kind of talk is totally fucking fun! It's what I do--albeit very informally--when I meet with groups of grad students and/or post-docs on seminar visits to other institutions. I'm not so sure it's really necessary or useful to give the regular "in order to test blaargh, we did bleeechcchch" seminar at the beginning. Why not just start at the beginning, with what really motivated you to embark on the project, and then present the actual historical progression?
This is a great idea! I was actually thinking whether it would be a good idea to prepare something like this for a job interview. I mean, all the while fully scientific, but highlighting the process of how I got there. I pondered that because I think that may put me in a better light, - being able to turn something that looked very bizarre in the first place in totally exciting science.
But I guess I will not dare to do such an unconventional thing...
This does sound cool. I am endlessly fascinated by how science work progresses in actual reality (same for all manner of art, or any creative process for that matter).
I have to agree with PP, just start with what actually you did, and then at the end discuss what you put where, and why. It seems that would be easier to follow as an audience member, especially if it's not in your subfield. Because for those people, the initial short talk isn't going to be enough to really stick in their heads, and thus the latter portion won't make much sense.
I think that's a good idea. I once had a conversation with a PhD student in which he complained about the fact that when you hear the standard scientific talk or read a journal article, and afterwards you talk to the investigators, you often get a very different story about the research. He was annoyed that it seemed like so much research was actually kind of disorganized, random, or accidental. (I think he was also annoyed that people were more non-committal about the awesomeness of their results, in person.)
Perhaps he was mad because we spend so much time talking about experimental design, and maybe he was feeling like that's all a bunch of crap since it never turns out predictably anyway. So I guess I'd encourage you, if you can, to see if you can weave in the notion that the random-walk approach that most research does end up taking should not keep you from having a roadmap at the outset (even if it kind of gets tossed).
This DOES sound like fun!! I think it would be funny to do the standard beginning only through maybe the first slide or two, and then morph into the "ACTUALLY, here's what REALLY happened" theme. Or just try to compress the "formal" version into something very short, like only the first 5 minutes, to still get to make the joke but leave the main bulk of the time for the real story.
I was actually thinking whether it would be a good idea to prepare something like this for a job interview.
Yes, but do *not* try to do this during your formal job talk. Your formal job talk should follow the standard "in order to test blaargh, we did bleeechcchch" format.
I think that sounds very fun!
I think the first run through, that you thought about doing in the "regular way" for 15 mins, could be maybe 3 mins? you knwo, go through the motions and say "normally you present it like this, A, B , C and D. In reality, it's really B,then D and then B and A finally... "
or something like that.
I wish you a good time and hope that you get good questions and a good dialouge. I'm going to prepare a talk now, a traditional one for my old department ;) scary...
What a cool format for a talk! That would be so fun to listen to, and probably a relief for a lot of grad students who might be sitting around with a mess of data wondering why everyone else's projects seem to follow such a logical path.
Thanks for the input! I like Arlenna and chall's suggestion to make the "formal" portion a lot shorter, just to give a flavor of why this story got into a good journal, and then quickly backtrack to how it actually started. (I didn't really have a plan AT ALL, was just sort of mucking around with stuff that had been insufficiently studied in the literature....)
And yeah, I think it's going to be fun--just wish I weren't more stressed out about the other deadline coming up that is far more important. But I guess that's what I have to get used to!
If the point of the talk is to discuss how you actually found your findings (heh), then I say tell it as it happened. In a formal talk, stick to "we tested blah, blah, crap."
That's extremely cool. I wish we did those.
Post a Comment