Eye Movements and Interface Component Grouping: An Evaluation Method
Evaluation of the computer interface is traditionally conducted via checklists, task scenarios, walkthroughs, heuristics and other subjective techniques. Analysis of user's eye movements during user interaction can add more objectivity to these evaluations. The objective of this study was to determine if eye movement behavior is influenced by interface design quality in a manner that can be characterized and used to identify design weaknesses. Eye movement scanpaths were collected from 12 subjects interacting with 4 interfaces utilizing different component grouping strategies. Interface quality ranged from “excellent” to “unacceptable”, as assessed by 50 typical users and 30 interface designers. A series of quantifiable measures were developed to characterize the resulting scanpaths. These measures examined raw scanpath data, fixations and saccades, and their interaction on both a temporal and spatial basis. The measures proved to be good indicators of interface quality. The “good” grouping strategy resulted in few fixations (minimal processing) and relatively efficient scanpath behavior (minimal search). The “poor” grouping strategies produced scattered scanpaths covering a larger portion of the interface area (extensive search) and requiring a higher number of fixations (more processing). The applicability of eye movement analysis as a quantitative interface evaluation method was further discussed.