Tips to run a seminar series
I have been an organizer and participant in seminar series (eg: Boltzmann, Whizkey) and consider myself fairly well acquainted with some of the issues in running a series regularly and effectively. So, when the current crop of Boltzmann members were discussing how they could do it better, I wrote out a few tips. I post a slightly edited version here; perhaps somebody on the interwebs might stumble upon this and find it useful.
Schedule a time of the week that works for everyone and stick to it. Once you make this regular, and a part of habit, you're more inclined to meet up. Also, since you know in advance about the sessions, you will/should not let other plans clash with this.
Don't make the meetings very long. Preparing for a longer talk takes more effort. Sitting through a long talk if you lose the plot midway is painful. And the chance of schedule conflict increases. So try shorter sessions of about 1-1.5 hours, at least till you can make it regular. If someone has a lot to say on some topic, they can speak over multiple sessions, which increases the frequency of meetings. Avoid the rigidity of a single theme for too many sessions unless you can keep them reasonably independent. This is not a "course", and you should be allowed to doze off once in a while, provided you show up :-) <!-- For both reasons listed above, Saturday/Sunday evenings are danger zones :-) If you meet on a weekend afternoon, keep it short and/or early! Weekend mornings are an under-rated time slot. The tips on timings work differently in different cultures/environments (For eg: My experience with Boltzmann is quite different from the one with Whizkey) -->
Commit to talks at the start of a semester. Make a list and put it up on a webpage ;-) You'll then owe it to the rest of the group to give a talk, which will hopefully help the overall "conversion rate".
If 5 people meet for about 10 sessions every semester, that would mean 2 talks per person. You know the math. So whenever you come across an interesting topic you'd like to discuss, consider that for your next talk. I'm quite sure that you can come up with 2 interesting topics per semester, on each of which you could could talk for about an hour.
If #4 doesn't work out, here's an old trick: Have a list of attendees and go in alphabetical order. Assuming y'all really want regular Boltzmann sessions, this will be effective in getting you to commit to a talk.
Needless to say, these suggestions are not absolute. Your mileage may vary, and some of these suggestions deceptively look simple but are not easy to practise. I'm still trying to pin down a working combination for the Whizkey seminar.
blog comments powered by Disqus