|
According to WWW''s simple, easy, and versatile user interfaces, related techniques'' advances, and government''s great support, the numbers of WWW''s users and Web sites are increased greatly and quickly. However, we can easily find many usability problems in most web sites. Web sites with bad user interface will not only waste much users’ time on the already jammed Internet but also damage the users’ image about the owners of the web sites and thus lose the customers. So, building web sites with good usability is an important issue currently. One of the most important concepts in user interface design is user-centered design. Usability evaluation is one of the most important tasks in user-centered design. However, the understanding on web site usability evaluation is not yet ready. Providing the properties of suitable usability evaluation methods for web sites is thus an important research issue currently.
In this thesis, some usability evaluation methods suitable for web sites adopted from traditional software are introduced, these methods’ most important properties that is necessary for web site designers are presented by experiment method. The methods being compared include Heuristic Evaluation, Cognitive Walkthrough, Heuristic Walkthrough, Usability Walkthrough, and User Testing. The properties used for distinguishing these methods include number of problems found, thoroughness, validity, reliability, cost effectiveness, average of severity, and weighted thoroughness. The experiment shows that Usability Walkthrough has better thoroughness, reliability, and weighted thoroughness. Cognitive Walkthrough has best validity. Group working on Cognitive Walkthrough has poorest cost effectiveness. Heuristic Evaluation can find more sever problems.
|