Welcome to the third issue of the Journal of Usability Studies.
In this issue:
The invited essay and peer-reviewed articles in this issue all address aspects that are not necessarily in the main stream, everyday usability studies.
In his invited essay, Charlie Kreitzberg asks “Can collaboration help redefine usability?” whereby collaboration refers to having a common, dynamically evolving domain knowledge space. Charlie’s thesis is that a collaborative space like this will not only provide access to an integrated knowledge base of research and practice but also help re-define usability in a more dynamic, collaborative fashion.
Agnieszka (Aga) Bojko introduces us to the benefits of using eye movement metrics to compare between user interfaces and assess usability. In her paper “Using Eye Tracking to Compare Web Page Designs: A Case Study” she is suggesting that using eye movements metrics can reflect the “cognitive process” in addition to the “overt performance”, and that it is particularly relevant for application where efficient and effective visual search are required.
Ever faced the challenge of assessing the usability of a large-scale system where aspects of safety are critical and the availability of specialized users of the system is extremely constrained? Elliot Hey in his article: “Case Study: Conducting large-scale multi-user user tests on the United Kingdom Air Defence Command and Control system” shares his experience in conducting such a study. Some of the interesting features described in this case study are the use of training materials in conjunction with user training and the concurrent multi-user observation.
Finally, a topic that we tend to avoid, particularly in usability studies: statistics. Jim Lewis and Jeff Sauro suggest that some interesting statistics can be computed even for the typical small samples in usability studies. In their article “When 100% Really Isn’t 100%: Improving the Accuracy of Small-Sample Estimates of Completion Rates” they suggest that the computation of confidence intervals can be used even for small samples with complete/incomplete type metrics.