Bimanual Input and Patterns of User Behavior

G. Evreinov

Department of Computer Sciences University of Tampere, Tampere, Finland

The assumption that two-handed manipulation save time is not always the right solution for interface design. As it was shown in Guiard, 1987, the preferred hand articulates its motion relative to the dynamic frame-of-reference determined by the non-preferred hand. This has an immediate effect upon both the behavioral strategy and the way the subjects could perform the task efficiently in two-handed interaction.

This paper describes the results of the study (data gathered from 20 subjects) of using two analog buttons in a target acquisition task illustrating behavioral principles for two-handed interfaces. The analog buttons allow the user to manipulate the cursor position across the entire area of the screen with a resolution of 256x256 pixels by having only 4 mm displacement.

The results showed that despite the apparent contradiction of the motion artifact to the one-to-one mapping principle, Y-axis changed with the left hand whereas the X-axis was controlled by the right hand, the subjects normally did not notice the artifact as such. This can be explained by the nature of the input mapping designed to accommodate the genuine structure of bimanual manipulation.

Our observations also revealed that the integral behavioral pattern averaged on 49 trials could be divided into three phases. The first is the motor programming phase. It is the time span between the onset of the movement and the time, when either of the coordinates cross the aiming-on-target zone (36 pixels in the current study). The final point of the motor programming phase coincides with the end of the sudden movement of the cursor along the dimension controlled by the dominant hand. The second phase is the closed-loop phase during which the subjects manipulate the buttons to get the cursor over the target. When either of the coordinates gets as close to the target as 8 pixels away from its center, the third phase begins, the final target acquisition phase. During this phase, the other coordinate should be caught up, so that the cursor finally gets over the target.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv