How To Referee
Resources:
- Presentations and the thesis oral exam
- powerpoint (needs to be updated)
- handout of the talk, to be given to the class
- notes (needs to be updated)
- A Guide for New Referees in Theoretical Computer Science
Ian Parberry, Dept Computer Science, University North Texas, April 1990
Overview
While there are some 'naturally born' presenters, most of us have to work at becoming good speakers. This presentation describes some basic tips for giving effective presentations to audiences.
The accompanying paper by Ian Parberry is another good resource. While it is somewhat oriented toward theoretical computer science, the tips in it are applicable to any speaker.
To do on your own
Attend and critique presentations. There are many opportunities to view presentations, e.g., by professors, by fellow students, in the graduate and faculty seminars, at conferences. People will vary greatly in their presentation skills and their styles. You can learn from all of them, be they good or bad. While your listening, take notes about the person's presentation style. What worked well? What worked poorly? What was left out? How could they have fixed it? What methods did they use that you thought you could try? Knowing how to judge other people's presentations is the first step in knowing how to prepare your own, and how to discover and fix your own flaws.
As part of the above, see how well the presentation followed the guidelines mentioned in class. Was there a clear message? Was media used well? Did the talk have a good structure where it flowed from point to point?
Practice and have others critique your presentations. Always give a dry run of your talk to a 'safe' audience. By safe, we mean people (e.g., fellow students, your supervisor, your lab mates) who are willing to say where your talk fails and how to improve it. Ask them to take notes, and have the audience publicly give you any specific problems and recommendations (public as discussions will often come out of it). It is far far better at having a safe audience critique your talk so you can improve it, instead of giving it (warts and all) to your intended audience.
