Journeying Through CHI 2006

By Elizabeth Bacon

Published: June 5, 2006

“As our field continues to have to build its standing among entrenched disciplines such as software engineering and product marketing, recognizing and ideally controlling the impact of politics is frequently crucial to the success of a UX design endeavor.”

When I signed up to attend CHI 2006, for the very first time in my seven-year career, I didn’t expect that I’d spend most of the event helping to staff our company’s exhibit space and drive hiring for the St. Jude Medical Human Factors Engineering team. In 2001, a paper I’d co-authored with Robert Reimann was accepted for CHI, but I was unable to attend due to conflicting project duties. Over the years, events always seemed to conspire against my attending CHI, although I’ve had the pleasure of attending other conferences such as DIS and DUX. At CHI 2006, I hoped to educate myself about leading research and fresh trends in the field of computer/human interaction, as well as network with folks I’ve worked and communicated with, especially through the Interaction Design Association (IxDA). In the end, though, working the St. Jude Medical booth consumed the bulk of my time. More on that later, however.

I was glad to attend two courses: a short course called “An Exercise in the Politics of Usability” and a full-day course called “Designing for User Efficiency.” I was generally disappointed by the course on the politics of usability, which was structured around the single, idiosyncratic example of “Mary the Usability Engineer” and her particular organizational challenges. Plus, the discussion that resulted from examining the political stratagems the instructor presented wasn’t as free-flowing and creative as I’d hoped it would be. Perhaps it was wrong for me to expect much edification from a ninety-minute course?

Nevertheless, it’s important to acknowledge that understanding and negotiating politics is a fact of professional life, especially within the corporate environment. As our field continues to have to build its standing among entrenched disciplines such as software engineering and product marketing, recognizing and ideally controlling the impact of politics is frequently crucial to the success of a UX design endeavor. One valuable take-away from the class included the suggestion to apply user-centered techniques to the design of any educational or process-definition efforts. Such methods would include performing customer research with colleagues, in order to know your users’ needs and concerns, and insisting on measuring the success of your UX design activities.

In the course “Designing for User Efficiency,” I learned about a concrete basis for improving the effectiveness of complex systems. Although the instructor, Deborah Mayhew, relied heavily on a single example from her work consulting for a call-center—a less complex domain in comparison to the sophisticated clinical systems I design at St. Jude Medical, most of which use touch screens—she managed to provide a solid framework for analyzing software ease of use that should benefit my design work.

“By separating ease of learning—so often referred to as the all-important intuitiveness of interactive systems—from ease of use, one can separately study the productivity of experienced users using a system on a regular basis.”

Firstly, by separating ease of learning—so often referred to as the all-important intuitiveness of interactive systems—from ease of use, one can separately study the productivity of experienced users using a system on a regular basis. With the crucial assumption that the users of a given system are experts, ease of learning becomes less important than ongoing ease of use. The framework Mayhew presented for efficiency analysis has its basis in the Keystroke-Level Model (KLM), which is a variant of the GOMS (Goals, Operators, Methods, and Selection Rules) model. Using KLM, one quantifies each specific step of an interaction with a system. For example, a task sequence requiring a user to find and click a button, select an insertion point in a text box using a mouse, then move her hand to the keyboard and type five characters would break down into four discrete operators. A code summarizes these operators as follows: P (point with mouse), BB (mouse click), P (point with mouse), BB (mouse click), H (hand movement), and T(5) (five typed characters). Each coded operator is associated with a measured time; thus, sequence analyses such as this one can measure the actual efficiency of system tasks. Now, regarding how KLM techniques translate into the analysis of touch-screen systems, I have some homework to do!

My major criticism of the class, despite its highlights, was that the design process Mayhew presented relied too heavily on task replication as opposed to goal-oriented improvements. Also, the specific “more efficient” design examples she covered for almost half of the class felt dated and unsophisticated, giving little hint of the dynamic expanse of rich, visual, modeless interactions available to designers today. My professional area of expertise is interaction design (IxD) with a broadening emphasis on the larger issues of user experience (UX) design. I also perform customer research that affects strategic product definition and regularly address information architecture problems as well as deliver user interface solutions. My three years of consulting with Cooper provided me with not only a very solid methodological foundation for design, but also a set of powerful IxD principles and patterns that help me to conceive innovative solutions. Participants in this class would have benefited from a deeper and richer palette of interaction design concepts that deliver efficient experiences.

“IxD was definitely a hot topic of conversation at CHI 2006. The organizers explicitly attempted to respond to past criticisms that the conference didn’t serve design practitioners well.”

IxD was definitely a hot topic of conversation at CHI 2006. The organizers explicitly attempted to respond to past criticisms that the conference didn’t serve design practitioners well. However, CHI does need to balance its focus on theory versus design, because no one disputes that it is the foremost conference for HCI research in the world. With its high-tech noodle shops and primary-hued walls of glass casting a creative influence on the crowd, I generally found the conference venue in Montréal, the Palais des Congrès, conducive to thinking about the expanding influence of design on CHI.

The field of IxD is coming into its own today, especially with the recent formalization of IxDA. The IxDA/UXnet dinner on Tuesday night was memorable. The more than 40 people who attended the dinner enjoyed delicious Italian food and wine while connecting with like-minded souls who had gathered from around the world. The growing participation of industry leaders at CHI is another sign that practitioners merit increased focus at the conference. Google seems ready to hire every interaction designer in the world, competing with Yahoo!, eBay, SAP, and other major companies for world-class talent. St. Jude Medical joined the hiring fray this year, representing the medical technology industry, which offers some of the most rewarding challenges to practitioners in our field.

Though the obligations of manning our booth dominated my time at CHI 2006, this activity gave me a unique perspective on the conference. CHI 2006 was a well-organized event, and the exhibit organizers provided good support in our securing and setting up our exhibit area. In many ways, I enjoyed spending time in the exhibit hall, because it was the central gathering point of the conference. During morning and afternoon breaks, attendees enjoyed snacks and drinks. There was also surprisingly good food at the opening reception and this year’s new Job Fair. Our booth drew a steady stream of intelligent and interesting individuals at all stages of their careers who were curious about what St. Jude Medical was doing at CHI. We quickly learned to begin with the line: “We’re not the children’s hospital!” (St. Jude is the patron saint of lost causes and is often invoked in the medical arena.) In fact, over the course of the event, we conducted interviews with over two dozen potential candidates for our Human Factors Engineering team, which was a tremendous outcome for us.

“CHI Madness, which followed each morning’s plenary session, was highly entertaining to the crowd, as HCI professionals reached deep into their souls to manufacture buzz around their work in less than 60 seconds.”

Between stints extolling the career opportunities at St. Jude Medical, I attended a variety of events, including the plenary addresses, CHI Madness sessions, a SIG on “Rhetoric and Argumentation,” an alt.chi session on design and research for hand-held devices, and various paper sessions. The SIG on rhetoric was thought provoking, because effective argumentation is a powerful tool in our political quest for professional legitimacy and advancement. Although another participant mentioned later that the session wasn’t as collaborative or discussion oriented as the SIG format was meant to be. Perhaps the SIG presenters had their own persuasive agenda?

CHI Madness, which followed each morning’s plenary session, was highly entertaining to the crowd, as HCI professionals reached deep into their souls to manufacture buzz around their work in less than 60 seconds. The format produced such lines as:

  • “If you care about salvation or sausage, you might learn something about social computing.”
  • “If you’re interested in coupling…on tabletops…then come see the papers about gestures and visualization.”
“Nothing in the world is necessarily manifest destiny, …one or two humans have the capacity to shape culture and evolve technology.”

Although I only dipped my toe into the waters, CHI 2006 provided multiple, engaging streams of events, ranging across the entire spectrum of HCI research and design practice. The conference closed with a diverting presentation from Scott McCloud, the iconoclastic cartoonist who broke through communication barriers with his seminal work, Understanding Comics. As he discussed the ongoing and only semi-satisfying transition of comics from paper to digital media, I found myself wondering about the exact thrust of his message for UX professionals. He contended that nothing in the world is necessarily manifest destiny, that one or two humans have the capacity to shape culture and evolve technology. So, are we meant to pursue our own idiosyncratic paths in order to build a universe of interactive systems that please and delight people? Or should we instead continue to work toward a future in which research and practice, academia and industry collaboratively define satisfying systems that advance civilization? I tend to believe CHI should continue evolving into a forum that fulfills the promise of greater unification among the disciplines of our field, to the betterment of all.

5 Comments

Thanks for a good review. I haven’t been to CHI for more than 10 years. Obviously, it’s changed and I’ve changed, and I’m not sure if I’d want to return, but reading this gave me a good perspective on what it was like for you to attend, and that’s enjoyable as well as helpful.

This is definitely a great review. As part of the Committee that worked for CHI2006, I am delighted with your last sentence, since it was one of the main objectives for 2006, and would still be for the future!

“CHI should continue evolving into a forum that fulfills the promise of greater unification among disciplines of our field, to the betterment of all.”

Marta Rey-Babarro University of Michigan CHI2007 Communications’ Chair

As one of the co-chairs for the Design Community facet of CHI ‘07, I can report that we are working hard to build on the momentum you sensed at CHI ‘06. You noted that the SIGCHI organizers “explicitly attempted to respond to past criticisms that the conference didn’t serve design practitioners well.”

The CHI 2007 “Call for Particpation” includes a section that encorages submissions in the area of Contemporary Trends and Experiences. Those submissions use the shorter, Extended Abstracts format that has been so successful at venues like DUX.

We are also working hard to make sure that the review process for that area of submissions is tuned properly. The questions we ask in the peer-review process for that area will not be the same as the questions for the full papers—not reviewed to academic publication standards. The questions will be more suited to matters of practice—as befits case studies and the like.

We have certainly run into many folks like Steve. People that are not inclined to give CHI another try. That is certainly understandable. In the meantime, we are thrilled that folks like you are seeing evidence of incremental change.

We sincerely hope that many readers of UXmatters will enter work, help review submissions, and attend the conference in San Jose next year. The more IxD and UX folks contribute, the more they will find relevant content.

See http://www.chi2007.org/community/design.php

As one who was pleased with the progress CHI 2006 made toward becoming a more practitioner-friendly conference and hoped for more at CHI 2007 in San Jose, I’m sorry to hear that you’re modeling your Contemporary Trends and Experiences track after DUX. That extended abstracts format leads to superficial coverage of topics and, thus, tends to be boring rather than informative to anyone in practice. It’s also a frustrating format for presenters and very difficult to pull off successfully. Your adoption of that format would seem more likely to discourage practitioners from attending CHI rather than encourage them to do so.

Other aspects of DUX were successful—particularly their tutorials and plenaries—but not their extended abstracts format.

Something that makes a good design conference is the inclusion of invited speakers who do not go through a review process. The best design conferences I attend, like IDSA, do not actually review submissions the way CHI or even DUX does at all. All speakers are invited, and it makes for a wickedly good design conference.

I think the idea of trying to make a design conference using a CHI/ACM-like format doesn’t make sense. Designers are not the same types of practitioners as researchers, and their model of education and practice differs significantly from other UX professionals. (I think there can be a huge mistake if we think that all practitioners of UX are designers.)

If we do want to encourage submission, the IA Summit style is my preferred format. It’s a very open format, with a peer-review process done by committee, but nothing publishable is submitted for review. You submit, at best, your abstract, and that is what is accepted. This allows flexibility for presenters and panelists, so their talks tend to be a lot more timely and thus more practical than submissions made six months in advance.

Join the Discussion

Asterisks (*) indicate required information.