Recent evidence about positive impact of expanding targets on user performance during manual pointing raises a provocativequestion whether a similar effect can be expected for eye gaze interaction. We present an experiment designed to examine the benefitsof target expansion during an eye-controlled selection task. The experiment involved twelve participants and a head-mounted eyetracking system EyeLink used as the input device. Our results suggest that both movement time to an expanding target and selectionerror rate are reciprocal functions of expansion factor and monotonic functions of distance to the target. Furthermore, movement time isa quasi-linear function of dwell time on the target, whereas error rate is invariant with respect to dwell time. Finally, movementdirection has an effect on error rate, but not on movement time. Overall, the results confirm the boosting effect of target expansion bothin terms of pointing speed and accuracy. These findings indicate that target expansion has the potential of making eye gaze input moresuitable for practical applications. Ill. 7, bibl. 6 (in English; summaries in Lithuanian, English, Russian).
Read full abstract