With the belief that adaptive interfaces have been largely unsuccessful due to their unpredictable nature and users' loss of control, the authors hypothesize that providing user support will improve the learnability and ease of use of the system.
This paper presents the results of a user study in which two sets of users (9 participants in each group) used an adaptive interface to view video clips. The first group received no support, while the second group received support via the interface.
Participants were given a mental model test before using the system and after completing a task. The authors hypothesized that the support group would score higher on the post-test than on the pre-test and that their overall scores would be higher than the unsupported group. In the test, the unsupported group actually scored higher than the supported group, refuting the authors' theory.
The authors also sought to measure ease of use by evaluating user performance and satisfaction. During the test, the supported group understood their device better (according to the performance metrics) and trusted the device more than the unsupported group (according to the satisfaction questionnaire).
The authors conclude that the negative aspects of adaptive user interfaces can be overcome without actually improving users' mental models of a system. They then raise the question whether user understanding of the system is actually required for a sufficient ease of use.
|