Back to curiousLee
DUX 2003 Notes by Amy Lee
alee AT aarp.org
Converted to HTML by Mike Lee 6.10.2003
Added remainder of Amy's Day 2 notes 6.11.2003Conference site: http://www.dux2003.org
Aggregated links here http://iaslash.org/node.php?id=7379
:: SFMOMA Tour
This group is the education department of the museum.
They have an ASP-based publishing tool (Pachederm) that publishes simultaneously to Web, DVD, CD and kiosk. It uses Flash. This limits the interaction design because the designers can only design to the lowest common denominator (Web). The kiosks they use rely on a mouse because of this (rather than touch screen). They chose to limit themselves because it was more important that they are able to efficiently publish to multiple platforms.
They use paper time-based flow charts to coordinate the development of multiple, concurrent content assets, rather than Gantt charts.
Staff consists of 3 full time staff (who do the project management) and 6-9 interns who do most of the develoment. Because of the constant changeover of interns, the authoring tool was made simple and uses lots of thumbnails so interns can be sure of what option they are selecting.
There is a lot of up-front work done on paper-based storyboards. The content, the flow and the design is all worked out before anyone gets on the authoring tool.
Art museums usually keep kiosks outside of galleries. SFMOMA is a little different in that they do actually put the kiosks discretely in the galleries. In order to get people to notice the kiosks (flat tables with hand-held earpieces), the kiosks have speakers and designers put the volume on low to "draw people to the buzz" without interfering with the a traditional museum experience.
They test effectiveness of kiosks by Time and Tracking (they track where people physically go and how long they spend at each spot), as well as interview people coming out of the gallery.
They are struggling with a "dot bomb" economy, which makes it harder for them to go back and fix usability issues.
They include games/interactivity that everyone agrees adds little value to the experience but 20-30% of visitors who use the kiosk will play with the game.
They experimented with PDAs, putting video on them so people could carry them around while they look at the paintings. They found that people preferred to sit down with video. Having learned this, they are looking for ways to incorporate tablet computers into furniture and putting that furniture into the galleries.
:: Day 1, Session 1: DUX in Practice 1
This session turned out to be largely about devices.
:: Jesper Kjeldskov: Maritime Communicator (I-Paq)
Problem: Used for container ships. The bridges of container ships actually have complex multimedia computer systems that people are accustomed to using so overcoming technophobia wasn't an issue. The ships are as long as the Eiffel Tower is tall and people are typically at each end and the center (bridge). Communications is typically done over walkie-talkie and with the noise of the environment, it is difficult to make sure your information gets across. In addition, the captain has to keep complex sets of instructions (orders) running concurrently for different parts of the ship when docking or leaving dock. The device was meant to make the communication less reliant on the audio and to allow the captain to know what orders had been given and which orders have been completed.
They found that, during interviews, interviewees didn't verbalize everything they needed (like the thing about keeping track of the orders). It wasn't until these guys got out on the ships that they really saw what needed to be done.
These devices are not really in use yet.
:: Ellen Ayoob: Drowsy Driver Device (for truckers)
Problem: Sleepy truck drivers kill themselves and others, as well as wreck expensive trucks and cargo.
This device would be built into the dashboard (it is still in testing) and works on "pupil-ometry" to detect drowsiness.
The design team worked hard to create extensive human interaction flow charts (rather than using flow charts for system interaction/integration). They found these charts invaluable as a means for communicating between the truckers, the HCI people and the technical people. "Pictures worked."
Originally, the devices (which are meant to be an assistant to the drivers and not some sort of "big brother" device) only had an on/off switch. A key change was to give the drivers a lot more info about how tired they were so they could make an educated decision about whether to stop or not ("No one wants to crash"). They added gauges and sensitivity settings akin to a CB radio's.
:: Eric Paulos: Connexus device (the ultimate "reach out and touch someone" device)
The device is a watch that has sensors built in to vibrate, communicate heat, and "mood ring"-like colors (even a heartbeat). The idea is to send a non-verbal cue that "I'm thinking of you". This is an attempt to take a physical touch behavior and transmit it over distance.
This grew out of observing teens with SMS. They like the "stealth" aspect of messaging over phones and frequently were sending out "touch" messaging like "Hi, I'm bored." He also noted that teens were creative in coming up with "languages" uniquely suited to the device.
He isn't sure how it will ultimately be used but hopes to release it to some teens to "see what language they come up with" in using the device.
:: Tim Brook: Agri-tech (finding new ways to use technology in vineyards
Did "deep hanging out" in vineyards ("yeah, it's good work if you can get it"). His actual job in Intel's R&D division is to look for new ways to use ubiquitous computing.
He found that what he was told in interviews was only a small part of the actual story/opportunity. He found people were much more forthcoming with specific information when they were reacting to a designed device/application.
The system is actually a network of wireless devices that measure, among other things, micro climates in a vineyard, mold, insects, readiness for harvest.
He found that is was also important to know what people in the fields were doing, and where they had been doing it (like spraying for insects). There are state requirements for documentation on this. He also found that there was a need for positional notes. The field workers would often use colored ribbons to mark a vine where something was happening (mold, bugs, not enough water). The ribbons frequently fell off or workers forgot what the issue was. Tim's system was able to help with this.
The system put into place could gather vast amounts of data that Tim and his team presented to the vineyard managers. The managers were unimpressed with the quantity of data and only wanted a subset of the data that could be gathered. "If you give me data about something I can't do anything about, I don't want the data."
:: Tom Brinck: Redesigning a Business School Web Site.
Problem: They just wanted to redesign the site and "make it better than all of our competitors". (*groan* Where have I heard that before?)
The school's previous design firm sort of shot themselves in the foot. The school asked "How do we know our new design in better?" The designers said "Because we are good designers and we say so." This made the school a good client for a usability firm.
The existing site had typical problems: lack of consistent identity or navigation, integrating or combining like tools, integrating an extranet. They had a diverse group of audiences they had to serve (faculty, students, alumni, potential students, the rest of the world). They found that while students had different tastes than the alumni or faculty, they wanted the site to show a classy face to the world since they wanted good job prospects after graduation. The designers found a common goal among the main audiences and then designed to that.
The proved the quality of the new design with task completion rates.
Though the entire project, they tested 180 people (many more than anyone else did).
:: Day 1, Session 2: Informing DUX
:: Daniel Makoski: Purchase modeling and loyalty programs
He used card sorting as a brainstorming tool.
Problem: Project for Wells Fargo to find out what makes people more loyal to their bank/credit cards
They originally sort the cards by task or type, but found that when they got people in, that they sorted differently. They sorted by "These make me happy" and "These I do every day". They also tended to sort more quickly by the frequency with which a task was done than by grouping like tasks. This was called "contextual" card sorting (visualized the context of the user's daily life and tasks). The team was able to use this to define/redefine the problem space.
:: Anu Kankainen: User Centered Conceptual Product Design
Her team was testing the motivational level needs ("Why") and action-level needs ("How")
She used dolls to simulate social situations described in the narratives. (She was working with kids on this project.) This helped her get deeper info about the situations.
She is a psychologist and seemed to approach this from a much more scientific (and deeper anthropological) approach than most of us can manage.
She specifically noted the importance of documenting user responses in a consistent way across different tests and phases of the project.
:: John Pruitt: Personas
Problems he experienced using personas:
Characters not based on data. They are, basically, designed by committee and feel fictitious to members of the development team later.
Personals not communicated well. "Resumes with a picture" are not enough.
No clear understanding of how to use a persona (no use scenarios).
No support from above (executives) to mandate the ongoing reference of and use of personas.
He created a Persona Foundation document with all of the personas in it. It has the "resume and a picture", but also scenarios and the specific data that back up every assertion.
:: Brian Herzfeldt -- Alignment Wall
Problem: Trying to keep track of all of the considerations, tasks, content, etc. of a complex project (a medical device for neurosurgeons).
He used Post-Its. It always seems to boil down to Post-It notes.
He took over a wall, taping printouts of the user's mental model of the task at the top of the wall. From that beginning, the Post-It notes started to grow. These notes reflected all of the different assets that were a part of that task -- but everyone had to keep referring back to the mental model.
He found the "alignment wall" a good tool for the team to communicate across disciplines. He got into the habit of pulling the entire team together once a week to talk over the wall.
:: Jesper Kjeldskov: TramMate -- acting out in context
This is another device -- a device to use on an Australian tram that brings together tram schedules and Outlook.
He used trained actors, as well as small groups of non-actors. These people were give props (pieces of wood) and asked to use them as they would a TramMate, then report the ways in which they would like to use the device. Researchers followed the users through their days and made observations.
He said it was an interesting experience, but took rather a long time and was a bit expensive.
:: Day 1, Session 3: Focusing on User-to-Product Relationships
:: David Danielson: Web Credibility
Nearly half of the comments in the study had to do with visual design. He found that, while they couldn't really articulate it, most people were visually very sophisticated. They see through the smoke and mirrors when the visual design is propping up a poor application or lack of content. "Development teams can't expect to just slap on a pretty face at the end of the development process." Every design decision must be carefully thought through.
If you don't "pass the test" with the first visual impression, web sites are rarely given the "second chance" with good content.
He found that people aren't very good a predicting their own behavior, but are very good at making attitudinal observations.
:: Jeanna Frost: Imagery in Diabetes Education
Found that logs didn't work for diabetics. No one really kept them up and it was typically too much uninteresting data for the doctors to be interested in ever reading. This team decided to give the participant diabetics digital cameras and had them just take pictures of their meals. The theory was that this was easier to do and would provide more concrete representations of what was actually eaten.
The team then synchronized the photos with data graphs of blood sugar ups and downs. This allowed participants to see what behaviors (eating certain foods or exercising) affected their health in a very tangible way.
The unexpected result was that the pictures often uncovered other issues in the environment (like an empty potato chip bag laying under the table with the salad on it, or the refrigerator with milk, orange juice and a case of beer). People started extending this from taking pictures of their food to taking pictures that represent behavior (such as a picture of the staircase to demonstrate how she could get some extra exercise).
:: Jane Wood: Mobile applications for breast cancer patients
Handheld devices were loaded up with personalized healthcare information. She had short video vignettes of people talking. They were largely older women.
Users liked it because:
It was an alternative to chasing down a busy doctor.
It was a mechanism for remembering. If they could go back to the device, they didn't feel like they had to remember every detail.
She is hoping to later incorporate text messaging for person-to-person support and to build in a mechanism for storytelling.
:: Day 1, Panel: How Far Can Design Really Go?
The whole panel was blogged live via Hydra and posted to http://www.spy.co.uk/DUX2003
:: Day 2, Session 1: DUX in Practice 2
:: Francine Gemperle: The Hug
This is a device that is designed to facilitate familial communications across distances. It is big, plush, vaguely animal-looking thing that acts as a 2-way communicator (like phone) and transmits a physical vibrate when the connected Hug is hugged.
An alternate interface for one-on-one communication.
Meant for grandparents (over 70 years old) and grandchildren.
:: Sarah Ljungblad and Maria Hakansson: Context Aware Photography
Studied "lomographers". Lomographers say "don't think, just shoot." They use crappy Russian cameras and part of the pleasure is the unexpected result of their shooting.
They used this as a starting point and built a digital camera that has filters built in that react to heat, pollution, noise, etc that exists when the photo is taken.
The device hasn't been tested yet.
:: Michael Kronthal: Portal research and design
He started out just doing requirements writing and forced himself into the design and prototyping stages of the process.
Because the development schedule got bumped, he had a week to do his usability testing on paper prototypes. He pulled it off and, in the process, uncovered new features. This sold the organization on the value of usability and he was granted more time and resources the next time around.
He noted that once users get used to a new product/interface, they start demanding more features. Doing testing before the release helps to identify those potential demands early.
:: Day 2, Panel 2: Dealing with Constraints
:: Bill Stubblefield: Micro Part manufacturing
The manufacturing process is a very linear process. The tool he originally created to track the process was originally linear. One thing had to be done before the other thing could be done. He found that his was causing roadblocks and people weren't using the tool.
He redid the tool so that the areas could be filled in non-linearly (with checks to make sure everything was filled in). This helped.
He also had a problem with a particular engineer who wouldn't use the tool. The rest of the manufacturing team sort of ganged up on him to make him use the tool. Basically, the user community liked the tool and made sure that everyone used it properly, rather than the designer, UX person taking the responsibility.
:: John Armitage: Too much work, too little time, and it has to be in German
Objective: Survival on an undeliverable consulting project. (Huge project contracted at flat fee before it had been scoped out.)
Big question: What do we not do to get this done?
Consulting firm on hard times. Lots of people in the office without a lot to do, but none of them were trained to do this work. There was not going to be any additional resources brought in. He decided to "give rifles to the cooks" and teach them all to be I.A.s.
He created teams of 3 people each who:
- Spoke German
- Could draw
- Could do research
He had people focus on creating storyboards for the development rather than wireframes. It seemed to help.
Really funny stories about how this project came off.
:: Steve Portigal and Lynn Shade: Research in Japan
"Cute" gives people a way to have a friendly relationship with a product or company. It is also a more palatable way to give issue unpleasant instructions or apologize. This why the Japanese seem to have an obsession with cutesy cartoons. They make extreme use of this.
These sorts of observations can lead to interesting conclusions in the right hands. For example, there are devices in ladies rooms that create a rushing water sound since Japanese ladies are often embarrassed of their own sounds and would flush to cover it up. To Lynn, this indicated a unique cultural sensitivity to noise that also affect the development (or lack of development) of voice actuated equipment for the office.
:: Day 2, Panel 3: Organizational/Business Issues
:: Bill Bachman: Common UI guidelines in Adobe
Symptom of the problem: In their products, Adobe has 7 count 'em, 7 different spell checkers.
Interfaces and tools were all over the place. Adobe needed more common UI elements.
Solution 1: They wrote a 700 page manual covering all aspects of UI. They had to do this because the techies would ding them if they missed some little thing. Obviously, no one but the UI department ever cracked this sucker open.
Solution 2: They wrote a 3-page "cheat sheet" of the most key UI elements. This got cracked open sometimes, but not enough.
Solution 3: They would send out weekly emails of "UI Trivia" putting the guidelines into neat little nuggets. The staff of Adobe didn't particularly appreciate the spam and it was stopped.
Solution 4: Common UI Report cards (1-page feature adoption matrix based on the guidelines in the 700-page document a huge effort and they are still figuring out how much detail every report card should get into). They got the organization as a whole to agree to really back up the common UI need and allow them to "grade" products according to these report cards. The development teams weren't terribly happy about being graded but didn't have a choice. The adoption matrix is turning out to be a valuable tool for documenting and comparing products as they are released.
:: Peter Morville: Adaptive Path and PBS
Peter is standing in for the guys who were supposed to be speaking because they couldn't make it.
Problem: The PBS affiliates all had different sites of wildly different design quality and usability. They wouldn't accept orders from headquarters, but needed some sort of direction
Adaptive Path's approach (what set them apart) was their emphasis on evangelizing directly to the affiliates. The design of the templates that the affiliates would be using as a basis took 3 months. Evangelizing and explaining those templates (as well as good design and usability) took an additional 3-4 months.
They did a heuristics report card for each affiliate. Since they were a 3rd party and not PBS headquarters, the affiliates were more willing to accept the results of these report cards. Then they worked with the affiliates to use the templates and redesign the sites.
The templates allowed enough flexibility for individual look and feels. They decided that constancy wasn't as important because people would go immediately to their local affiliate. They wouldn't be comparing a site for an Arizona affiliate with one in Indiana.
:: Plenary Session
:: Stephanie Yost Cameron and Sara Little Turnbull
Cameron is an executive at NeoPets. This is a global gaming community that has survived the dot-bomb era and proven to be very successful. 80% of the site's users are under 20 years old and most of them are girls.
Different language speakers see precisely the same site at the same time only the text is different. They aren't seeing "sister" sites where they are only interacting with others from their own culture. Japanese kids can message directly to German kids in real time and they both see the messages in their own languages (with some obvious stilted language, sometimes). This is very unusual… and was very difficult to do.
They've had to custom develop language, filtering, and messaging applications that would handle the load and the unique needs of this young audience.
Turnbull was talking mostly of her history in design and anthropology. Fascinating stories.
Back to CuriousLee