Impulse characteristics in rapid movement: implications for impulse-variability models.

Three untested assumptions of the impulse-variability model were examined in two experiments utilizing rapid, uni-planar limb movements. Experiment 1 varied movement distance (A) and movement time (MT) in a rapid-timing paradigm where the subject moved a lever through a certain distance in a certain time. Experiment 2 varied A in a reversal response where the S made a rapid elbow flexion and extension in a given MT. Displacement recordings were made on every trial. KR (knowledge of results) about MT was given after every trial. The results can be summarized as follows: (a) As predicted by the model, variations in impulse size and velocity were directly related to the impulses size; (b) There was no correlation between the accelerative and decelerative impulse durations recorded during the reversal response supporting the notion that the impulses might be independent; (c) Negative correlations (-.20 to -.50) were demonstrated between peak acceleration and impulse duration for both experiments, counter to the predictions of the model; and (d) Counter to the predictions of the model, timing error (VEt) increased as A decreased for rapid-timing responses. When the correlational results are taken into account, the model has the capacity to account for curvilinear relationships between relative timing error and movement speed. Overall, the results suggest that the impulse variability model requires some restructuring before it can be considered a viable model for the control of rapid limb movements.