The purpose of this thesis is to test the hypothesis stating; that since users with domain expertise browse, perceive and use domain specific systems differently than domain novices, involving domain novices into the evaluation process would help to locate and improve more disadvantages that would be overlooked, but also yield more results about the advantages of the system that are less focused on content, but the overall usability aspect instead. The hypothesis has been compiled from the literature on the subject of usability evaluations. We have chosen this hypothesis, as we want to contest the prevalent importance of using participants with domain expertise (intended users) in usability evaluations, as we in our studies in Information Architecture have experienced that novice test participants can contribute to usability evaluations too.

The first part of the thesis consists of a presentation of the general theory on the subject of usability evaluation and Information Architecture, which also covers the literature stating the importance of using intended users as participants. This is followed by recent research studies on the subject of domain experts compared to novices in usability testing, as there is uncertainty on when to use experts or novices as usability test participants.

The test methods used to test our hypothesis are Think-Aloud and Card Sorting, which are being utilised on twenty participants, who consist of ten domain experts and ten domain novices. The experts were chosen from their expertise on mountain biking and the novices were chosen at random, so to study the differences between a carefully selected group of participants and a group consisting of participants chosen at random. The mountain biking domain is chosen, as it fits the context domain of the websites used for the usability tests.
The thesis is concluded with our recommendations for the choosing of participants in usability testing. The recommendations are based on our findings in the analysis of the data provided by the two participant groups in the usability tests, and reflect the results of when we found the use of domain novices to be as useful, or more useful, than using domain expert participants.

The conclusion shows that the participants with no domain knowledge can in some aspects of the usability test be useful, and it will include our recommendations for what future usability and Information Architecture evaluators could consider, when completing their research design.