Amending Movements

Abstract

This paper examines the relationship between the degree of a mechanical disturbance, outcome accuracy, and amendment times to produce response corrections. Movement time error and amendment times were generated by systematically increasing the duration of acceleration and deceleration perturbations. Subjects produced discrete timing responses (700 msec-70°) during which perturbations were interjected into the ongoing movement on random trials. Amendment times were generated from acceleration curves along with a number of related kinematic parameters (e.g., Movement Time, Peak Acceleration). The results showed that as the degree of the mechanical disturbance increased, timing error and amendment times to the perturbations also increased. At low force levels, the percentage of accurate responses to a decelerating perturbation was approximately equal to the percentage of accurate responses for control trials. As force level increased, however, timing error increased and the percentage of accurate responses decreased. In addition, as the magnitude of the disturbance increased, changes occurred in the kinematic properties of the perturbed movements which contributed toward the degree of outcome accuracy of the response. Collectively, the results are discussed in relation to an error correction system that operates in an interactive fashion based on the characteristics of the error and the constraints of the task.