Personal tools
You are here: Home 2003 Vol.35 No.4, September/October 2003 TOCHI Retrospective
Document Actions

TOCHI Retrospective

In this letter to the editor, Former Editor-in-Chief of TOCHI, Jonathan Grudin gives a retrospective look on his experiences.

In the course of six years as editor, I have thought about the role of journals in our fast-moving world, and the role of Transactions on Computer-Human Interaction (TOCHI) in HCI.

Scientific journals were invented to communicate and archive results. Today, results are communicated more quickly at conferences and on the web. Within HCI, conference proceedings are now archived and widely accessible.

Watching papers evolve – almost every published TOCHI submission has been substantially revised at least twice – I concluded that the journal serves an important function even if no one reads it. Most authors (and reviewers) are forced to think carefully and comprehensively about the topic in the review and revision process. They return to their research, teaching, or practice with a deeper and broader understanding. This varies with the kind of work being reported, but it is generally true. Some papers are improved so much that the reviewers almost deserve co-authorship credit.

A recent review of the ACM Digital Library download statistics were a happy surprise. Assuming that no bots are at work on the ACM site, papers are downloaded much more frequently than I imagined. This is accelerating as the digital library membership spreads and more people become aware of it. CHI conference papers are downloaded more than those of other ACM conferences, and TOCHI papers are downloaded more often than CHI conference papers on average.

I encourage readers to make the effort to submit work to HCI journals. You and the community will benefit. Your work will be read. Digital archives are likely to remain available for a long time.

Partly in light of these observations, I want to respond to a proposal in the first SIGCHI E-Bulletin issue: that we follow a SIGGRAPH model and institute a revision and re-review cycle for our conference submissions, then publish them as TOCHI journal articles. Perhaps it is appropriate for SIGGRAPH, but this model has little merit for human-computer interaction. The purposes it might serve can be realized in other ways.

Introducing a conference revision cycle is a great idea, as was done by DUX2003, for example. But if the elevation of conference papers to journal status were undertaken, each year 12 TOCHI papers would be real journal submissions and about 100 would be conference papers. The real journal articles average three times the length of conference papers and are more substantial. They require more work. Would authors feel any incentive to write one if others get the same credit faster and more easily with a conference paper? Readers searching online could have trouble distinguishing solid journal articles from weaker conference papers. This model could discredit our field with tenure committees, NSF review panels and anyone who discovers that we are masquerading conference papers as journal articles. Rather than elevating the status of our conference papers it could diminish the status of our journal articles. It seems to manifest insecurity, the impression some people have reported from the CHI Letters experiment. We should be more self-confident. HCI is an accepted discipline. We are getting hired. I have written many tenure and promotion letters for HCI researchers and every candidate was successful. Some people may never accept HCI, but calling a lightly revised conference paper a journal article won’t change their opinion.

There is a better way to accomplish the apparent goal: “Best Paper” awards. These stand out on a résumé or curriculum vitae more than a conference paper elevated to journal status. Best Paper awards are noticed. I voted with the majority against incorporating Best Paper awards in the CHI conference in the mid-80s. Today I agree with Jack Nicholson, who said that he loved awards because they help some people and don’t hurt anyone. I think that’s true – getting one is great, but there is no sense that not getting one is bad. I think the CHI conference should designate some papers as particularly strong and then accept a lot more, enabling more people to attend. If we decide CHI conference papers are journal articles, it is likely to drive the conference to be even more exclusionary, the wrong direction for an umbrella conference in a mature field.

Powered by Plone CMS, the Open Source Content Management System

This site conforms to the following standards: