Abstract
Our previous work has shown that the use of vision during rapid upper limb reaches is optimal at high limb velocities. When providing vision only above .8 m/s, reaching endpoints are as precise as in normal vision and vision only provided below .8 m/s does not yield better endpoint control than without vision. In the current study, vision could be provided during 3 limb velocity windows above .8 m/s (Early [between .8 m/s & 1.4 m/s], Middle [between 1.4 m/s & 1.4 m/s] & Late [between 1.4 m/s & 0.8 m/s]). All possible combinations were used in a factorial design, yielding 7 vision conditions presented in a randomized order. Each vision condition was presented 20 times and a no vision condition was presented 140 times. Full vision pre- and post-tests were also performed. Our main dependent variables were tied to the variability (i.e., precision) and bias (i.e., accuracy) of movement endpoint distributions. In the primary movement axis, movement endpoint control was more precise when vision was provided in both the early and middle vision conditions than in the no vision condition. Providing vision in the late vision condition resulted in worse endpoint precision than the full vision pre- and post-tests. Our results indicate that visual information may be used most efficiently for endpoint precision when the limb is moving quickly early in a movement and in the portion of the trajectory that includes peak limb velocity. In contrast, vision above .8 m/s but below 1.4 m/s late in a movement does not appear to contribute to endpoint precision control. Thus, the use of visual information may be tied to the kinematics of a movement and be most effective early in a movement.