Optimizing the control of high-ID movements: rethinking the power of the visual display

A recent experiment by Boyle et al. (Exp Brain Res 223:377-387, 2012a) demonstrated that providing a sine-wave template for participants to follow enhances performance and transfer on difficult (ID = 6) Fitts tasks. Another experiment (Fernandez and Bootsma in Acta Psychol 129:217-227, 2008) demonstrated the effectiveness of applying a nonlinear transformation of the visual feedback provided to participants executing difficult aiming movements. The present experiment was designed to determine whether these two enhancements when used together would provide further enhancements to difficult aiming movements. Participants were randomly assigned to one of the three acquisition conditions. Participants in the Fitts and Fittslog condition were asked to flex/extend their arm in the horizontal plane at the elbow joint (wrist stabilized) in an attempt to move back and forth between two targets as quickly and accurately as possible. In the Sinelog condition, participants were asked to track a sine-wave pattern. The timing for Sinelog conditions was set to result in total times (movement time + dwell time) similar to that anticipated for the Fittslog condition. The feedback displays for Fittslog and Sinelog groups were subjected to a nonlinear transformation, but not for the Fitts group. Following 54 acquisition trials (17.5 s each) under their assigned condition, Test 1 was conducted under the same conditions as the participant experienced during the acquisition trials and Test 2 was conducted under Fitts conditions with the nonlinear transformation of the display data. Test 3 was conducted under typical Fitts conditions with no transformation of the display data. The results for Tests 1 and 2 indicated that total time and movement time for the Fittslog and Sinelog groups were reduced relative to the Fitts condition. In addition, dwell time was significantly lower, a larger proportion of movement time was spent in the acceleration portion of the movement, and normalized peak velocity was significantly lower for the participants in the Sinelog condition than for participants in the Fittslog condition. On Test 3, an untransformed Fitts condition was imposed; the Sinelog group outperformed the Fittslog condition on all variables except hits, endpoint variability, and peak velocity where the Fittslog and Sinelog groups performed similarly.