Wednesday, June 2, 2010

CRV 2010, Teaching, and Good Presentations

Yesterday afternoon, I gave my first academic conference paper presentation. I have given talks at conferences before (including Grace Hopper), but they weren't on highly technical topics. My main goal for this talk was to stay true to what I believe is a good presentation style, even though it's not really the norm.

I've had the opportunity to rant about talks and poster presentations a few times lately. The main summary of these rants is that I hate that people put so much information in their slides or on their posters! The thing is, there's only so much detail I can take in in a short period of time. And guess what? If your second slide (or your poster) is filled with all kinds of facts and numbers and equations you think are useful...my brain turns off. You've lost me for the entire rest of your talk. I don't this on purpose - I just can't process all that stuff, no matter how clearly and loudly and well paced you speak.

I knew avoiding the temptation to put lots of detail in my talk would be hard when I presented my CRV 2010 paper on global context for SURF and MSER features. After all, this is a technical conference with an audience that actually knows this stuff. But somehow, I think I managed to do it. My slides, embedded below, certainly don't stand alone, though they do seem to give a decent general idea of what I talked about. I made sure I rehearsed, and found a good way to explain some of the details of the method and results when I spoke. I'm quite happy with the results (one very kind attendee introduced herself and then told me the presentation was "perfect"). Even still, I would probably prefer to remove some of the text. ;)


I like to take this idea to teaching as well. Many of those willing to listen to my rants argue that lectures are different. You need to include lots of text so students have notes to refer to later. I say if you want to give them extra information, use the Notes section of your slides software. Keep the main slide content to a minimum, and supplement with activities, videos, and discussions. This is much easier in the intro to computers class I'm giving to arts students, but should translate to programming and algorithms classes as well (to an extent, anyway).

Anyway, I don't mean to sound as pretentious as I think I may be sounding, but I think it's worth ranting on this topic every once in a while so that, one day, the norm can become a style that many more of us can learn from.

4 comments:

Unknown said...

Funny, I blogged yesterday about how to do good presentations too! I totally agree that less text is better. (The "less is more" comment.) The thing is, if you have a lot of text, people will read it instead of listen to you. And that's bad.

One excellent teacher of mine makes liberal use of "graying out". That is, he does have a decent amount of text (equations, algorithms) in his slides for students to refer to later, but as he presents, he only "ungray"s one sentence at a time. He also uses pictures whenever possible to describe algorithms.

Keita Wangari said...

Great post. I read somewhere that if people ask for a copy of your presentation then you had too much info on your slides! May not be 100% true (I always like to have the presenter's Additional Resources slide) but probably a good guideline.

iosparkletree said...

I enjoyed this post a lot Gail, thank you! I wish more scientists talked about sharing their work in quality ways like you.

Gail Carmichael said...

Thanks all!

Post a Comment

Comments are moderated - please be patient while I approve yours.

Note: Only a member of this blog may post a comment.