SIG HCI Research Symposium (ASIST 2006)
Notes and comments on the SIG HCI Research Symposium at ASIST 2006.
HCI and Meta-search Engines, Amanda Spink and Carolyn Watters
Marchionini (1996) showed that there is very little overlap between meta-search engines. They used a dataset from Dogpile.com that sounds a lot like the AOL dataset that was released — if the AOL dataset had been released through an academic project, would people have reacted the same? Users in the dataset tended to be on the high end of search behavior, with ~4 terms per query and ~4 queries per search. The duration of interaction was only a few seconds, but higher than average for search engines. In general not much difference in behavior between meta-search engines and normal search engines.
Good question — what are the limitations of web log analysis? It’s generally only one facet of the picture. When interesting patterns are found in the logs, they would do separate studies based on those patterns. Someone else mentioned that Google itself has realized that the amount of information available from logs is finite, and that they are really pushing into ethnographic research to evaluate search behavior.
User Behavior During Web Information Tasks, Melanie Kellar
Big question: What are the high level tasks on the web? How do users interact with their web browsers in standard high level tasks?
Focused a lot on web information/activity monitoring. Found that most users considered RSS too overwhelming. Users are interested in changes to specific information, not necessarily an entire page (though this behavior is very much task dependent).
For designers, this implies that we need to provide better tools for users to monitor frequently changing information, at least information that people would be interested in monitoring. While RSS is one way to do this, it is still necessary to offer alternative tools.
See session later this week.
Design Of Digital Libraries, Tony Moore (Drexel)
All of the digital libraries that he has mentioned are DSpace based. While most have modified the template, they still all look the same. Talked about the lack of visualization in DSpace. The default DSpace UI is incredibly dull. There seems to be a lot of low hanging fruit in this area when it comes to visual design and interaction design. He’s working on a mockup for a more visual interface for DSpace, could be interesting to follow.
Question: Did you think about the IA of the page? The order of information on the page?
Research Tool For Investigating User Interactions With Retail Websites, Allan Harold and Brian Detlor (McMaster)
How is consumer interaction with retail websites influenced by information seeking behavior and information presentation. Mentioned a big issue, the researchers looking at web behavior need to build viable simulations of real world websites that can also gather research information. They used Amazon Web Services to create an experimental front end. I wonder if they did any evaluation of user credibility for the experimental website versus real retail websites.
Desktop Virtual Reality Environments, Jamshid Behesti and Andrew Large
Browsing is mainly a visual activity. Using familiarity with libraries to create virtual environments for browsing digital libraries.
Games And Motivating Children To Engage In Health Applications, Carolyn Watters et al (Dalhousie)
Games increase engagement in treatment, increase motivation, provide a distraction, and have social benefits. Targeted children who are managing chronic diseases. Kids wanted to share and communicate with their friends.
Measuring Usability of International User Interfaces, Dania Bilal
Bourges-Waldegg & Scrivener, 1998 — Regardless of the language of an international interface, the design of the interface should be understandable to all audiences.
Sadly, I have to duck out now, but Jenny Preece (who is sitting two seats away from me) is about to talk.