5th Human Factors & the Web: The Future of Web Applications

CONFERENCE PROCEEDINGS:

5th Human Factors & the Web: The Future of Web Applications
June 3, 1999
National Institute for Standards in Technology
Gaithersburg, MD

The drive down was a treat. Despite the torrential downpour in the Pennsylvania mountain country, Cam’s Jetta only hydroplaned once. Every once in a while Derek would loosen his grip and let the blood flow back into his knuckles.

Following a fitful sleep at the Holiday Inn, we headed on over to NIST to grab out nametags and fill up on bad coffee and day-old pastries. After receiving a truly cheesy bag o’ goodies, we settled in for the conference introduction.

COGNITIVE STRATEGIES IN WEB SEARCHING
Raquel Navarro-Prieto

http://zing.ncsl.nist.gov/hfweb/proceedings/navarro-prieto/index.html

The first presentation was entitled Cognitive Strategies in Web Searching, and was presented by Raquel Navarro-Prieto of the School of Cognitive and Computer Sciences at the University of Sussex, England. Of Mexican descent, Navarro-Prieto rushed through her 20 minute presentation speaking far too fast and nervously for most of us to understand half of what she said. What we did comprehend came from the default white text-on-blue Powerpoint presentation. Navarro-Prieto focused on the cognitive and real-world problems of searching the web and user-testing. In a nutshell, they attempted to put down on paper the cognitive strategies followed by users as they attempted to search the web. Despite the fancy flowcharts and big words, we both came away from this presentation with the impression that the study was deeply flawed and should not be trusted.

WHAT DOES GETTING WET (Web Event-Logging Tool) MEAN FOR USABILITY?
Michael Etgen

http://zing.ncsl.nist.gov/hfweb/proceedings/etgen-cantor/index.html

After the crash-and-burn of the first presenter, the WET presentation was a welcome relief. Given by Michael Etgen of the User Experience Engineering Division at AT&T Labs, we were treated to a showcase of a truly useful tool for usability and user-testing professionals. The Web Event-Logging Tool is a Javascript-based (client-side) tool that allows for complete and absolute user-tracking, especially when enabled within a testing suite. The sample tool they showed used a Cold Fusion database to store the data captured by the WET tool, including number of seconds and nanoseconds between clicks, the position of the mouse during a click, the IP number of the user, the name of the file being requested during the HTTP event, and the state of the file’s retrieval (success, failure, server error, etc.) The demonstration they showed also used a DHTML layer that was positioned above the page that required the end user to start/stop the event logging. This is entirely optional, depending on the test suite setup and the desired results of user-testing. Overall, the WET Javascript tool looks to be a very valuable and beneficial addition to every usability and/or testing lab that deals with end user interactions and web-based cognitive processes.

GETTING BACK TO BACK: ALTERNATE BEHAVIORS FOR A WEB BROWSER’S BACK BUTTON
Saul Greenberg

http://zing.ncsl.nist.gov/hfweb/proceedings/greenberg/index.html

This particular presentation did not turn out to be as interesting as the title allowed us to think. I was looking for truly alternate behaviors for a browser’s back button like the much-talked-about “smart back” behavior, and what we received was a proposal for changing the default behavior of the back button from a stack-based model to hub-and-spoke based model. Potentially, this could hinder an end user’s understanding of where they have been in the past as it reorganizes the history of a browser’s locations based on a parent/child hierarchy. While this was a truly interesting concept, I am skeptical about changing a browsers’ behavior so dramatically, especially a behavior that accounts for 30-37% of all navigational acts. More interesting were the facts and numbers they were able to dig up, including the fact that 6 out 10 pages a user visits are pages they have visited before.

THE EFFECTS OF CULTURAL MARKERS ON WEB SITE USE
Charles Sheppard

http://zing.ncsl.nist.gov/hfweb/proceedings/sheppard/index.html

Easily the worst presentation of the eight we attended, we found it to be deeply flawed, based on incomplete or skewed data, and presented by someone who had little or no experience with how to handle such a rowdy and intelligent audience. The concept of the presentation was clear, focusing on how a web site should be designed with cultural markers in mind. For instance, given the fact that green is a popular and widespread color in the Middle East, the designer would design a site for a Middle Eastern audience to use green more prominently. The biggest flaw was the decision to use ornamental textual header graphics (green) in place of the typical textual header. The designer’s decision was based on the idea that a Middle Eastern audience would react more favorably to big green text because the color green was a cultural marker. The manipulation of the site to use cultural markers was poorly executed, gave little thought to usability, performance, or any of the basic fundamentals of site design. This study cannot be trusted, as it was based on incomplete data, poorly thought-out research, and completely ignored common site design standards, and even breaking them to accomplish the study’s desired results.

THE MAX MODEL: A STANDARD WEB SITE USER MODEL
Gene Lynch

http://zing.ncsl.nist.gov/hfweb/proceedings/lynch/index.html

Based on the understanding that the web is maturing, the Max Model is an attempt to define a typical end user and/or standard web site user. The Max Model is broken down into a Personal Profile which includes cultural and psychological characteristics, a System Profile which includes system characteristics (speed, RAM, HD browser, etc.), Internet connectivity (transfer speed, loading factors, etc.) and a Cognition Profile (semantic and inference, set rules and behaviors). Overall, this presentation was done very well, had good research to back up the ideas and hypothesis, and the presenter was convincing.

GLOBALIZATION OF USER-INTERFACE DESIGN FOR THE WEB
Aaron Marcus

http://zing.ncsl.nist.gov/hfweb/proceedings/marcus/index.html

With the rapid worldwide expansion of the Internet and the web, the recognition that interfaces for the Internet do not hold up as well in some cultures, societies and countries. For instance, where the desktop metaphor of files and folders works well in any industrialized and modern society/culture, it does not work as well in less-developed countries. A book and chapters metaphor would work better? It is important to take into consideration the various intercultural metaphors, mental models, navigation concepts, and interactions of both industrialized societies and not. Give thought to internationalization issues and intercultural issues such as religion, culture, language, and aesthetics. The globalization process requires planning, research, analysis, design, implementation, evaluation, and documentation, all the while keeping international and intercultural issues in mind, so as to prevent any unwanted cultural taboos or mistakes. Understand the issues of design and typography across cultures and societies, the problems a designer faces upon page orientation (left-to-right or right-to-left), and potential color choices and combinations that may be offensive. Overall, Marcus and Associates have done an excellent job of researching these issues and has presented us with an amazing insight into the future of web design, worldwide web site usage, and implications of such. All of this will be published in forthcoming book from Lawrence Erlbaum & Associates.

TOWARDS WEB MACROS: A MODEL AND A PROTOTYPE SYSTEM FOR AUTOMATING COMMON TASKS ON THE WEB
Alex Safonov

http://zing.ncsl.nist.gov/hfweb/proceedings/safonov/index.html

Essentially, stating that the web is moving towards a truly interactive medium, that uses web macros to deliver pages based on a user’s pre-determined profile and history. Safonov talks about the differneces between emerging technologies that manage these actions: Infobeans, Internet Scrapbook, AgentSoft’s LiveAgent, and the basic Web Macro. Safonov used the Bookmark (or Favorites) as the case for change. Better bookmark management systems are emerging, which will eventually lead user-driven processes for automating repetitive browsing and data entry tasks.

PANEL: PERVASIVE ACCESSIBILITY — ACHIEVABLE OR PIPE DREAM?
Mike Paciello, Jakob Nielsen, Shawn Lawton Henry

http://zing.ncsl.nist.gov/hfweb/panel/panel.html

Easily the best part of the conference, the panel about widespread accessibility within the web was truly excellent. The typical diatribe about ALT tags, screen readers, and overuse of graphics was prominent. However, the speakers focused on the idea of educating the people who build the web about accessibility issues, instead of trying to pass laws to require accessibility guidelines are met. Most interesting was the proposal for an emerging framework for web site architecture that allows for pervasive accessibility. Instead of building 10 static sites for 10 different audiences, you build one dynamic site that delivers the correct data, information, and interface depending on which audience your end user is part of. Truly excellent.

WEB NAVIGATION: HOW TO MAKE YOUR SITE FAST AND USABLE

http://zing.ncsl.nist.gov/hfweb/proceedings/tiller-green/index.html

This sums it up: “Design for your audience.” I didn’t learn much from this presentation.

SUMMARY:

It was a great conference, despite some bad presenters, flawed data, and skewed presentations. Next year’s conference will be held in Austin, TX in early July.

Posted by Cameron Barrett at June 3, 1999 11:59 PM

Leave a Reply

Your email address will not be published. Required fields are marked *