Sunday, June 17, 2007

Anti-PC Travel

Over at the new Computational Geometry Discussion Forum, they are discussing program committees, and Ileana Streinu argues that every PC should have a face to face meeting for final decisions. I couldn't disagree more.

Academic life involves a significant amount of necessary travel -- conferences, talks, NSF meetings, research collaborations. Requiring travel for a PC should be avoided, especially these days when any air travel is an unpleasant hassle, and electronic PC tools (like easychair) are becoming reasonably good. My other main objections:
  1. Travel wastes PC members' time. For people with families, the cost is particularly severe, as overnight travel specifically takes away from family time.
  2. PC travel costs. It either increases the cost of the conference, or PC members have to pay with their own grant money.
  3. Extra air travel is ostensibly bad for the environment. Be aware of your carbon footprint!
What are the supposed benefits of face-to-face PC meetings, and how to respond to them?
  1. It leads to better decisions. This seems to be a matter of faith with some people, but I have seen no real evidence for it. In my experience, online discussions can be as good as face-to-face discussions. In both cases, there are various types of noise, and decisions will not be perfect, but hopefully nearly so.
  2. It saves time, by gathering people all in the same room to make decisions. Given the time required for actual travel, I doubt this is true. Even granting this, it isn't worth the loss in time flexibility. Getting on a plane means missing class, family dinners, etc.
  3. It leads to better feedback. I do not see how this can be possible. Electronic PC meetings allow recording of comments and discussions, which should make it easier to provide better feedback, in the form of appropriate cuts and pastes, in the reviews to the authors.
  4. It provides an opportunity for colleagues to get together. This I can't deny. Face-to-face PC meetings are usually more fun and informative than electronic ones. Does that really make up for the time/cost/annoyance of travel?
So in general, if a face-to-face PC meeting is required, I'll likely be saying no, unless I have some ulterior motive that makes the travel worthwhile.

3 comments:

Luca Aceto said...

Michael,

I think that your assessment of the relative merits of electronic and physical PC meetings is well balanced. I basically agree with it wholeheartedly.

As far as feedback to the authors is concerned, I believe that electronic PC meetings provide a much better record of the discussions related to a paper, and make it easier for the PC chairs and PC members to update their reviews to take into account the issues/suggestions that were raised. IMHO, it is not rare to see very valuable comments emerge from the PC discussion, and the use of any framework that makes these comments available to the authors should be encouraged.

Sure, an electronic PC meeting is longer than a physical one. However, it is certainly less disruptive than a physical PC meeting taking place in a far-away location.

Luca

P.S. Welcome to the ICALP 2008 PC for track A.

Anonymous said...

Is it not possible to combine the benefits of face-to-face communication and electronic discussion by a video conference? The technical media needed for this should be mostly available, aren't they? I don't know, but I think it should be not to hard to record video conference meetings for later purposes.

Regards.

edwardahirsch said...

After attending a physical PC meeting (of ESA), I can say that physical meeting stimulate more discussions and urge the PC members to produce comments where they would keep silence in the other case. It is possible to ask whether anyone verified Lemma X.Y and get the answer. It is also possible to have in some cases an opinion of those PC members who do not have the paper in their assignment.

On the other hand, indeed some noise is introduced by the time constraints for decision.