I attended CHI 2007 in San Jose. CHI is a large (2,500+ attendees) conference that focuses on academic HCI research, but also includes many panels and sessions on trends in user experience research and design. Here are some of my observations from CHI. —Abe
Trends in design
Bill Moggridge’s keynote talk emphasized the need for “intuitive design” to help navigate complexity and create designs that people enjoy. He quoted Eames: “the design is an expression of the purpose… it may if it is good enough be judged as art.”
He then laid out a framework of “design skills”:
- Frame (or reframe) a problem or objective.
- Create and envision alternatives.
- Select from a range of alternatives (solution space).
- Visualize and prototype.
- Synthesize a solution within constraints, and understand the impact of design changes on results.
He gave a usefully simplified example of how cultural context can influence design. In Japan, most professionals commute on trains, and the social norm requires maintaining a respectful quiet. As a result, Japanese commuters work quietly with mobile devices, using many small buttons and a complex UI. Americans, on the other hand, commute primarily by car. Fiddling with small buttons and screens and difficult and dangerous while driving, but we can “shout to our hearts’ content,” so voice input is a promising alternative.
BM also showed an entertaining video of a Japanese woman trying to purchase a soft drink with her i-Mode cell phone. This was a usability disaster, and illustrates how the design of many mainstream products is still basically broken. Many videos are included with the DVD accompanying his recent book, Designing Interactions.
A later panel, “Who Killed Design?” followed up on BM’s ideas of intuitive design. Bill Buxton described the crit [that’s a critique, as in MFA and similar design programs] as “a fundamental part of the design process,” but one that’s rarely written about or discussed. He argued the design community needs to reflect more carefully on the role of structured critique and feedback in the design process. The trend in design education, presented by educators Terry Winograd and Meg Armstrong, is for heterogeneous teams (e.g., business, engineering, and design) and incorporating reflection on design activities. A challenge for both students and practitioners is to distinguish “problem setting” from “problem solving” and learn when to apply each approach.
Buxton argued passionately that “there is a calculus [of design], it’s just not the same calculus we use in the sciences, and it must be respected.” The CFO’s opinion on design should not be considered equivalent to an experienced designer’s opinion, just as it wouldn’t be considered equivalent to an engineer’s opinion.
Prototyping
A panel on prototyping took up the challenge of “what’s wrong with prototyping in HCI research?” In a nutshell, current prototyping practices are weak—prototypes aren’t real (representative of the ultimate system or design idea being investigated), and evaluation using them isn’t real (participants, tasks, questions, and time periods are limited or unrealistic). Unfortunately, “getting more real isn’t realistic,” because creating more detailed prototypes and extensive evaluations is prohibitively expensive in many cases.
The panel focused on addressing this dilemma in HCI research, but design and UX practitioners likely confront similar problems. Two approaches seemed particularly promising. Jonathan Grudin emphasized the bias inherent in having the prototype creator (e.g. an interaction designer or information architect) evaluate the prototype. Even when another person (e.g., usability engineer) evaluates the prototype, she is likely to be a friend and collaborator. Grudin suggested an exchange approach: “you evaluate my prototype, I’ll evaluate yours.” Of course, this may be difficult in cases where confidentiality is an issue. But an IA working on an intranet might be able to evaluate a prototype for someone in another industry working on an ecommerce site, and vice versa. Ron Baecker suggested evaluating multiple prototypes at one time, to get a broader range of opinion and more useful suggestions from users. But Jared Spool countered that, in practice, “the same design team can’t create meaningfully different prototypes” and it would be too expensive to engage multiple teams in prototyping for the same project.
Spool also emphasized the importance of failure, saying “it’s very important that we talk about failure.” He suggested morbidity and mortality conferences in medicine as a potential model for UX/HCI. The idea is to focus “not on success or failure, but what you’ve learned.” Grudin suggested the “overoptimistic” history of videoconferencing as a cautionary tale. Initial research on videoconferencing was very encouraging, with successful small-scale evaluations, but there was a lack of attention to serious problems that emerged with widespread use. He concluded that “HCI as a field needs more reflection and consolidation of what’s been learned.”
Web 2.0 and enterprise software
“Web 2.0” ideas have taken public, “consumer-facing” sites by storm—can they do the same for internal, enterprise software? Jonathan Grubb of RubyRed Labs asked us to “please never say the word ‘enterprise’ again.” Instead, he encouraged designers to “pretend you’re making consumer software for people who work in big companies.” He contrasted the power and rapid adoption of social tools, such as Facebook, versus administrative tools such as a university’s directory. The idea is to replace the traditional “top-down push” and centralized software with individual-led adoption. The very structure of traditional enterprise software inhibits rapid iteration, which is critical to effective design.
Grubb suggested designers try to “Give people something that will inspire envy in their coworkers.” He gave the example of “showing off” Basecamp in an organization that was using MS Project—coworkers literally envied Basecamp’s simplicity and visual appeal. He further argued that “niche consumer sites have been successful, so why not niche business applications?” A continually improving ecosystem of public tools, such as Yahoo Pipes, enable designers and web developers to quickly prototype new concepts.
New approaches in ethnography and design research
Informances
Ron Wakkary presented “informances” as a way to bridge ethnography and design. His team worked with families as “everyday designers.” In the informances, students who conducted the ethnography went on to act out the role of a participant. He showed a video of an informance that highlighted how one participant in the study struggled to use cell phone voice recognition. The benefits of this approach include understanding embodied action (such as the physical movements involved in using a cell phone), and developing a shared understanding of the insights from fieldwork. Wakkary cautioned that, like participant observation, informance is a practiced skill, and not everyone is comfortable performing.
Dialogues in design research
Johanna Brewer presented her experiences with workshops that combined fieldwork, group discussion, and design. The workshops were intended to explore the theme of “in-betweenness,” as expressed in activities such as public waiting (lines, etc.), and in transitional times/spaces (such as commuting). The workshops included “research speed-dating—2 minute introductions” to match participants, scavenger hunts and observation tasks in the city, and the presentation of interesting examples from the field. Participants then engaged in a design activity based on ideas from the field data, and finally reflected on the results and their conceptual significance.
The workshops led to interesting concepts to explore futher with research (e.g., the theme of ‘legitimacy’—the right to be waiting/lingering in a certain place at a certain time). In addition, provocative design concepts also emerged, such as “myst-air,” which would create clouds of water vapor, to mark one’s ‘territory’ in a public space. Brewer characterized the workshops as a cyclic process between practical engagement and conceptual discussion, creating a dialogue between theory and practice with a complex topic. This is an intriguing alternative to standard UCD and participatory design processes.
Triangle User Experience 