Smartphones and tablets are becoming indispensable to many different types of people’s daily lives. These touchscreen devices, however, generally require visual attention. Some work is emerging in the literature on the subject of eyes-free interaction with mobile devices. In addition to voice input, some researchers are exploring the possibility of eyes-free touch-based interaction.

This paper focuses on analyzing the performance accuracy and the important factors of eyes-free single-handed thumb interactions. The experiment and analysis in this work include some relatively obvious conclusions. For example, there is a Gaussian distribution of touch point locations; the variance of touch points is higher with eyes-free interaction than with visual attention; and accuracy drops when the touch target size is smaller. But the work also includes some interesting observations. The authors note that touch point location variance is significantly smaller on the x-axis (horizontal direction) than on the y-axis (vertical direction); the centroid of touch points per target offset to the target center’s left along horizontal direction and to the center of the screen along vertical direction; and touch targets on the top-left side of the screen have the worst accuracy.

The work first describes in detail the different literature on human interaction accuracy on touchscreens. It also describes the motivation and research questions behind the eyes-free interaction study. The experiments are done with 1D targets and 2D targets on the screen, and the study is performed with a significant number of users and collected datasets. The collected datasets are presented and analyzed in detail, and studied for performance factors such as distribution of touch point positions per target, touch accuracy, and the offset and variance of touch points. Overall, this paper is useful. The key observations on touch accuracy can serve as a foundation for future studies and various applications of eyes-free touch controls.