Thesis Committee

Eugene Santos Jr., Ph.D. (Chair)

Mark Borsuk, Ph.D.

Vikrant Vaze, Ph.D.

Abstract

Task structures serve a crucial role in computer interface design and usability analysis where they assist in modeling users in order to design an interface that minimizes error and maximizes performance. However, a gap exists within computer interface usability and task analysis whereby usability methods and task models do not fully account for the cognitive attributes of interfaces, tasks, and their users. More specifically, these structures do not consider the overall problem-solving efforts involved in both use of the interface or solving the problem, resulting in a large variance in human performance. We propose a simple high-resolution task framework that supports cognitive attributes of both users and tasks. Using our framework, we designed and implemented a user study to determine the impact of cognitive load and interactive elements on user performance while validating our framework assumptions. The results of the study correspond to our theoretical framework analyses and build upon the understanding of user performance in the presence of increased element interactivity. We show that not only is performance linked to increased element interactivity, but also the way in which that interaction takes place. We show how the presence of a simple strategy can significantly increase performance even in the presence of high element interactivity and expected cognitive load.