Variability and Location of Movement Endpoint Distributions: The Influence of Instructions for Movement Speed and Accuracy

An influential theory of motor control predicts that targeted hand movements should be aimed at the target center and that the variability of movement endpoint distributions should fill the target region (Meyer et al., 1988). Because increases in the amount of movement endpoint variability correlates with increases in movement speed (Schmidt et al., 1979), centering the distribution on the target center and expanding variability to the limits of the target boundaries should allow for maximization of movement speed, without the production of movement errors (i.e., target misses). Slifkin and Eder (2016) recently found that those predictions only held over a range of small target widths; however, as target width increased the endpoint distribution variability increasingly underestimated the variability permitted by the target boundaries and the location of the distribution center increasingly underestimated the target center. There was a strong relationship observed between the unutilized target region and aim points shifting away from the target center. Those results suggest that the downward shift in endpoint distribution location was based on “knowledge” of the amount of endpoint variability relative to the unused space in the target, and such downward shifts may reflect a reduction of travel costs (e.g., movement distance). Thus, there is a possibility that there is a link between unused space and how much distance minimization occurs. Here, we extend the results of Slifkin and Eder (2016) by explicitly manipulating endpoint distribution variability through a manipulation of task instructions, thereby allowing a more direct investigation of the link between unused space and distance minimization. The instructions emphasized either 1) movement accuracy, 2) both movement accuracy and speed, or 3) movement speed. Participants generated movements under different target width and amplitude requirement conditions. Variability increased as the emphasis on movement speed increased. In turn, as variability increased within a given target width condition, the amount of unused space within the target region decreased. The results provide support for the notion that the relation between aiming and knowledge of variability was maintained, but the nature of the relationship was influenced by the instruction conditions. The implications of these results on models of optimal motor control will be discussed