The tangled Web we wove: a taskonomy of WWW use

Paper: 
Michael D. Byrne, Bonnie E. John, Neil S. Wehrle, and David C. Crow. 1999. The tangled Web we wove: a taskonomy of WWW use. In Proceedings of the SIGCHI conference on Human factors in computing systems: the CHI is the limit (CHI '99). ACM, New York, NY, USA, 544-551.

Review:
Authors claim that little attention was given to analyize the user tasks while browsing the www (internet). Their goal was bring a more usable and effective user interface for internet users (who brows the www) from the tool side and from the usability of the web side.

Authors work was new in identifying and understanding user tasks while browsing the internet. Unlike the previous studies which focused on the user navigation patterns such as: click-studies. This provide an advantage of better understanding the user tasks and behavior when browsing the internet.

Experiment was conducted on 8 participants. They were asked to browse the internet as what they usually do in their daily routine. Each was asked to provide a verbal protocol that describe his/her browsing activity. Additionally, a video camera was recording protocols and users screens. The amount of analyzed data was between 15-74 minutes for every participants.

For the analysis purpose, they have constructed a taxonomy of user tasks when browsing the internet. This taxonomy was constructed using their previous work, and by observation of the first three users in the experiment. This taxonomy is hierarchal taxonomy of user tasks while browsing the internet. WWW tasks are as the following:

• Use Information
• Locate on Page
• Go To Page
• Provide Information
• Configure Browser
• React to Environment

Every task encapsulate sub task. They used those tasks to construct user tasks through the experiment. For example, a user wanted to download a college paper, his top-level goal was to use information (download) and the rest of sub-goals go as the following:
UseInfo(download)
    GoTo( bookmark)
    ProvideInfo(search criterion)
    Locate (related)
    GoTo(hyperLink)
    Locate(related)

In their results, they found that the six main tasks that they used to capture user tasks were useful and worked well in that matter. Also, they found that some tasks can be formed from other sub-goals. This lead them to the conclusion that their hierarchy was not strictly hierarchy (nearly flat).
In their discussion, they had two type of design implications: on browser design, and on page design. One interesting thing about webpage design was that its not necessary a bad design to have a webpage full of text (from their experiment). Another intuitive thing that they discussed was the need to optimize the page load time as most users spend good amount of time waiting for pages to load (I guess this is not a big issue now as it used to be at 1999).

In my opinion, they did a great job by constructing the main task hierarchy that form user goals when browsing the internet (figure 1). Another issue that I found that they use the time and the number of events as the main two factors to evaluate every user task. Time factor can be biased sometimes. Sometimes users needs some time to learn their way through website on their first visit. I think this issue should be addressed in the limitation of the study.