Kevin Cheng  

The Why of CHI 2005

April 15th, 2005 by Kevin Cheng :: see related comic

Last week, Tom and I attended CHI in Portland. Aside from the paper presentations I mentioned last week, CHI also features Interactive Exhibits, Panels, a debate and some special interest group meetings. This format is pretty much consistent with the previous CHIs. Worth noting, and somewhat tangentially related to our comic this week, is that CHI2006 will actually be divided into communities, one of which is Education. The others include Design, Reseearch, Usability, Engineering and Management. This new format should give more focus in each area to perhaps facilitate more interesting discussions and submissions. We shall see.

Tomorrow, I’ll be writing about one of the highlights of CHI, Bill Buxton’s Lincoln Laboratories panel. In the meantime, go check out There was a lot of discussion on some mailing lists about the vlaue of CHI to practitioners and why people did and did not go to the conference. I’d love to hear one, some or all of:

- why you did or did not go to CHI2005
- whether you think CHI2006’s format changes will help improve the conference for you personally
- if you attended CHI2005, what the hi and lo-lights of the conference were for you

Stay tuned for more coverage from us.

8 Responses to “The Why of CHI 2005”
Domenick J. Dellino wrote:

I attended CHI 2005 and was extremely dissappointed in how few of the academic papers had any grounding in the reality that we as practitioners face every day.

I’m fearful that UE professionals are not being prepared for what awaits them in the real world (if, indeed, that’s where they’re headed).

CHI 2005 High–Jared and Eric debate (although even that was more of a semantic discussion than I had hoped for).

CHI 2005 Lows–Just about any paper presentation from anyone at a university.

Susan F wrote:

High: Randy Pausch’s (CMU) Opening plenary on VR creative/engineering teams was very entertaining and inspiring, Bubble Cursor seems like an important basic interaction breakthrough, student design competition showed promise, the many posters were high-quality and very diverse, and there were free Google t-shirts and WiFi.

Mildly interesting: interactive exhibits, and the finding that small round mouse cursors are faster to use than arrows are. (Comparing Cursor Orientations for Mouse, Pointer, and Pen Interaction, Po, Fisher, Booth: UBC Canada)

Low: Most of the paper presentations I went to, although interesting at times, were not very compelling or useful to me. Food was both scarce and unappealing at all the venues, lack of chairs next to electrical outlets caused the half with laptops to sit on the floors.

Things I saw that are probably useful for someone:
(single usability score calculation)
(a new visual interface for DJs)

Overall impression: Like reading a stack of Scientific Americans. Ecclectic, often interesting, very technical, very narrow, not much grounded in practice (lots of basic research of interest to industry, a few prototypes). A few practical ideas for social good. A few scary ideas for social control. Not as interesting as CHI used to be.

Interesting posters:

TXTmob: Text Messaging for Protest Swarms

A Gesture-Based American Sign Language Game for Deaf Children

Ruug Long-Distance Communication (synchronized pairs of heat-sensitive rugs that communicate over the internet and show glowing impressions of the other person on each).

A power cord that glows according to consumption.

Designing the “World as your Palette” (painting app for kids that samples the environment for pattern stamping/brushing) MIT Media Lab

The closest thing to a useful item for me personally was the “Tool for Accurately Predicting Website Navigation Problems, Non-Problems, Problem Severity, and Effectiveness of Repairs” by Blackmon (U of CO), Kitajima (AIST Japan), and Polson (U of CO).

It uses LSA corpi to analyze navigation labels and flag problematic ones, based on the books various user groups theoretically have read (and thus what their working vocabulary theoretically should be). Problem is matching corpi to actual humans and groups is hard and very theoretical.

Dave wrote:

I went to CHI for 2 reasons:
1. the Development Consortium. There is going to be a UXnet hosted site with a lot of information about the outcome. You can get a taste by looking at the program page fo the consortium. To put it short (or is that too late), a bunch of UX related orgs got together to discuss how we can better cooperate and collaborate in the future.

2. I got the oportunity to present IxDG at a booth. Highlights manning the booth were my very short talk w/ Bill Buxton and my very long talk with Alan Marcus.

The other reason I go to any conference is to network, talk, engage. I think that is the main flaw with most of the CHI conference is that it is about publishing and presenting and not a heck of a lot to do with engaging. Most engaging went on at fun events like the IxDG hosted Dinner/Beer night and the CHI Reception. I met some great people and had a lot of fun, though I was pretty scared about the floor caving in while dancing.

As for the new format, I believe it is short-sighted in a certain way. I don’t believe that you need to make CHI be the UX conference of the year. That really isn’t fair to the other conferences and I hate this notion that the conferences are in competition with each other. Instead, to the point of last week’s article, academic/researcher presentations need to be made more accessible. Don’t present your research, present how your research is relevant to ME! a practitioner. Tell a good story based on what you learned from your research. I can read the published paper if you intrigue me afterwards.

More acurately put, academics need to make their presentations accessible and interesting to practitioners. Using the same publishing based format but bringing in “designers” (like a designer is going to create something like this.) is not going to create any meaningful x-pollenation. The designers will stay in the designer track and the academics will stay in the academic tracks.

This is one way that DUX works. It is a publishing/presentation format where usually the presentation had only loose relationship to the published case study, but there was only 1 track. This meant everyone was engaged w/ the same sets of experiences throughout the 2-3 days. Can’t wait for DUX this year!

The one thing I do wish DUX would do differently this year is find some time for moments of workshop and education than there was last time.

Victor wrote:

Depending on your background and interests you will find this one or that conference interesting or disappointing. However, as Dave said, the main point of going to a conference is the networking aspects of it and you can always meet interesting and great people in any conference. If you just go to the presentations in any conference I guess that it can be very boring.

My background is in interactive systems development/HCI stuff and I find more interesting UIST ( than CHI. It is smaller and more focused and that works better for me. But that is me!

It is not always the case, but the papers in CHI often deal with specific aspects of very concrete interactive systems (either design, development, evaluation, etc and that has to be new in terms of the previous literature).

The level of detail and the novelty requirement makes it difficult to see “what-is-bigger-picture” and how can that immediately be relevant to practitioners. However all the basic usability methods that we use today were probably discussed in CHI or a CHI related conference many moons ago.

David Heller wrote:

For those interested DUX just announced their CFP and more information about dates and location: go to the DUX site for more info -

Matt McKeon wrote:

I thought the Jared and Eric debate was trite and pointless — Crossfire for the usability set. Each took a position obviously contradicted by reality and got called on it by the audience. Trying to claim that you can train anyone to be a usability expert in 5 months is ridiculous, as is the claim that you can do effective usability without some kind of formal process. Quote of the session from an audience member: “Jared, you can’t expect to just make numbers up and not get called on it.” I like Jared quite a bit, but I think the setting and the artificiality of the topic forced the two of them into untenable positions.

The tutorials I attended were OK. The cross-cultural UI design tutorial taught me that the model chosen by the instructor was a poor predictor for design. However, it proved to be pretty useful for guiding a formative study and analyzing its results. The prototyping tutorial was basically a bundle of tips and tricks, including a surprising usage of Excel to rapidly generate semi-functional mockups. In general, tutorials tend to be a lot more focused on practitioners. I think the future CHI model of smaller, cheaper, quicker classes is a great idea — hopefully more practitioners will be able to afford them.

Some of the papers were interesting, if you knew what to look for. I particularly liked the papers on using images in communication and Nokia’s context-enhanced messaging study.

Frankly, academics can’t write papers on web design and get tenure. They have to push on less mature areas of usability research - ubicomp, security, augmented reality, etc. I think the academic / professional divide in HCI is definitely a problem at the moment, however I don’t think it will be solved by making every HCI program a trade school. Perhaps an HCI program should require a professional internship as part of its graduation requirements.

And, as usual, the networking at CHI was great. I met scads of smart folks, had my world rocked by one or two conversations and have a walletful of business cards from industry and academia. Although (was it Tom or Kevin?) didn’t want to join in our sketch battle during the Google party. :p

Keith Instone wrote:

I missed this - “Bill Buxton’s Lincoln Laboratories panel” - did it get written? Sorry, cannot find it.

Kevin Cheng wrote:

I only just posted it this week. Thanks for bringing that up!

Leave a Reply

OK/Cancel is a comic strip collaboration co-written and co-illustrated by Kevin Cheng and Tom Chi. Our subject matter focuses on interfaces, good and bad and the people behind the industry of building interfaces - usability specialists, interaction designers, human-computer interaction (HCI) experts, industrial designers, etc. (Who Links Here) ?