Analysis of Individual Typing Performance and Influence of Eye Gaze in Flick Text Entry Method

Abstract

Flick text entry on smartphones is so popular that a majority of users prefer it to the more conventional QWERTY keyboard. Yet, the typing speed and accuracy of such text entry varies greatly depending on the individual. Focusing on eye-gaze behavior, we analyzed typing data of 33 individuals recorded with a touch logger and eye tracking data of 31 individuals recorded with a glass-type eye tracker to clarify what causes the individual variability. Correlation analysis revealed the following four findings. 1) The typing speed and the reciprocal of time interval from a release to the next landing showed a high correlation over 0.90. 2) The higher the gazing ratio at the keyboard was, the slower the typing speed and the higher the accuracy tended to be. 3) Some expert users achieved a fast typing speed and a high accuracy. 4) The majority of errors were landing errors to adjacent keys. On the basis of these findings, we propose a keyboard layout that considers eye gaze.