The usability canary in the security coal mine: A cognitive framework for evaluation and design of usable authentication solutions
This addresses usability issues in security design for system designers, offering a novel approach to integrate authentication with user tasks, though it is incremental in applying cognitive psychology to a known bottleneck.
The paper tackles the problem of security mechanisms being too unusable, leading users to circumvent them or abandon services, by developing a cognitive framework to help designers select the 'best fit' authentication methods at the design stage, and demonstrates in a validation study that the model can predict user performance and satisfaction.
Over the past 15 years, researchers have identified an increasing number of security mechanisms that are so unusable that the intended users either circumvent them or give up on a service rather than suffer the security. With hindsight, the reasons can be identified easily enough: either the security task itself is too cumbersome and/or time-consuming, or it creates high friction with the users` primary task. The aim of the research presented here is to equip designers who select and implement security mechanisms with a method for identifying the ``best fit`` security mechanism at the design stage. Since many usability problems have been identified with authentication, we focus on ``best fit`` authentication, and present a framework that allows security designers not only to model the workload associated with a particular authentication method, but more importantly to model it in the context of the user`s primary task. We draw on results from cognitive psychology to create a method that allows a designer to understand the impact of a particular authentication method on user productivity and satisfaction. In a validation study using a physical mockup of an airline check-in kiosk, we demonstrate that the model can predict user performance and satisfaction. Furthermore, design experts suggested personalized order recommendations which were similar to our model`s predictions. Our model is the first that supports identification of a holistic fit between the task of user authentication and the context in which it is performed. When applied to new systems, we believe it will help designers understand the usability impact of their security choices and thus develop solutions that maximize both.