Evaluating informational characteristics of oculomotor system and human–computer interfaces

Abstract

Interest in adapting eye movements to everyday interaction is apparent in recent consumer technology: more and more interfaces with gaze as input will be developed in years to come. Usually throughput and other Fitts’ index of difficulty based methods are used to evaluate human’s ability to perform a coordinated movement. However, it is questionable if such methods apply to gaze based interaction. This paper explores measures of information transfer rate and channel capacity—a better alternative to throughput. Based on difference of initial and final entropy of the extent of the movement, these measures can be used to evaluate and compare any pointing devices (mouse, joystick, eye tracker, etc.) because they evaluate only information processing capacity of oculomotor channel, without including informational characteristics of the methods to perform object selection.