Conferences vs. Journals in Computer Science: An Alternative Perspective from Education

At the CRA Conference at Snowbird this week, there was a session on "Publication Models in Computing Research: Is a Change Needed? Are We Ready for a Change?" During the discussion, the question was asked, “What other models are out there that don’t put conferences against journals?”

This question reminded me of an experience I had years ago. My research interests are in Human-Centered Computing (HCC). As such, I work across multiple disciplines.

In one of my collaborative projects with an education researcher, we decided to publish our work, and my collaborator suggested we submit a proposal for our work to the American Educational Research Association (AERA) conference (http://www.aera.net). He informed me that this was the primary conference in education research and everyone attended this event. The 2012 conference had 13,000 attendees to give you an idea of the size of the conference. We submitted our proposal, which was accepted. As you can imagine, I thought we had to prepare a paper for the proceedings. To my surprise, he informed me that they don’t have proceedings for the conference.

He went on to say that we would give a presentation on our work at the event, but no paper submission was required. So, I went to the AERA conference that year and was surprised to learn that there were several book and journal editors in attendance scoping out presentations. After the conference, my collaborator received an email inviting us to submit a full paper for consideration in an education research journal. Hmm? That was strange.  After some additional investigation, I learned that book and journal editors attend sessions and browse the abstracts to identify potential papers for their respective venues.

I thought, what an interesting concept. Could computer science do the same? What if our conferences required proposals for presentations and then selected some for consideration in journals and book chapters? Would this be a compromise to reduce the tension between conferences and journals in computing? Would this provide an incentive for journals to quickly review submissions? Would this format allow computing researchers to do more extensive research knowing that the results are not limited to a 4-6 page conference paper?

I don’t think we can have a single computer science conference like AERA, but I do think some of our venues could benefit from this format. At a minimum, I think it’s worth discussing.  
 

Juan E. Gilbert is IDEaS professor & chair of the Human-Centered Computing Division in the School of Computing at Clemson University.