Avian’s Blog

Today was the last day of seminars on Information and Communication Technologies for this year at the Jožef Stefan International Postgraduate School. I believe I managed to attend 77.7% of the 18 talks given by other PhD students and gave one short presentation on spectrum sensing. That should bring me above the mandatory attendance with a sufficient margin of error.

With a few exceptions, most of these talks were wrist-cuttingly boring. Considering that the final conclusion was that this year's presentations were better than usual I dread to think what previous years had to endure.

It was suggested to me that I might not understand all the finer nuances of the pedagogic process. However I still believe that if students are giving these talks to learn how to a present their research topic, more realistic feedback on their performance might be helpful. So here is mine.

I think two mistakes were prevalent:

First was gross mismatch between audience and the basic knowledge required to follow presentations. ICT field of study at this school is so broad that you will have someone parsing tweets about celebrities on one side of the classroom, someone working on protein databases in the center and someone doing motor bearing fault prediction on the other side. As an electronics engineer I had the misfortune of acquainting myself with sentiment analysis before. Still it doesn't make sense to talk to me about Segmented Discourse Representation Theory if you can't introduce it in a way that someone who has not been studying it for the past 6 months can understand.

Personally, I like best the format where presentation is roughly split into three parts: the first one can be followed by anyone with an elementary education, second one by someone familiar with the field and third one by your supervisor and colleagues from your research group. I find that I can follow such presentations with interest even when topic is completely outside of my specialties.

I realize that with the suggested 15 - 30 minute length of these ICT seminars this might not be possible. But I think it's better to drop the last third than the first. Your supervisor already knows what you are doing and will skip the talk anyway. Others can read your paper.

"Because Magic" by Egophiliac

And that brings me to the other problem I noticed. I know that it is the de facto standard that academic slides are walls of text and are expected to stand alone relatively well without the accompanying speech (as much as I like LaTeX, Beamer was a terrible invention). I also know I enjoy every talk that breaks this mold. Don't recite a table of contents of your talk at the beginning. Make slides colorful - you're not paying Springer's color print fees here. Explain the topic using examples, plots, photographs. Don't fill up slides with 10 pt text that I can't even read from the back row. Use slide notes if the slide deck must be usable as a stand-alone publication (by the way, it doesn't in these seminars - the school doesn't even publish slides on the web).

Presentations in an academic setting certainly must be on a much higher level than your usual marketing pitch with stock photos, cute kittens and a stevejobsian act, but that doesn't mean they must make me sorry I didn't rather spend the afternoon debugging some PHP code.

In conclusion, I think everyone would be better off if these seminars would be given a bit more exposure. Obviously there is a lot of interesting research being done in the offices at our Institute. But as long as presenting it is viewed strictly as a formality and there is no push towards teaching people how to talk about their work in an understandable way, it will stay behind those closed lab doors. Presenting scientific discoveries to the public can be done by postgraduate students at informal seminars as well as by big-name professors with 3-hour extravaganzas in the large auditorium.

Maybe if that would be the case, the school wouldn't need mandatory attendance and enforce one (1) mandatory insightful question per student after the talk to give an impression of an academic debate. We might even get some of that much-sought-after cooperation between departments.