| Popis: |
Errors in task performance, such as the failure to strike a ball with a bat, are common in everyday actions. How do humans respond to such errors and the corresponding absence of rewarding outcomes? We test the hypothesis that action performance failures drive implicit refinements in motor plans. We employ a learning paradigm wherein we manipulate the location and size of a reach target while “clamping” visual feedback of the motion in a fixed direction, resulting in either performance failures or success. Our results fail to support the view that performance errors trigger implicit recalibration, and rather suggest that they set in motion a distinct, verbally-responsive, volitional strategy to select actions which rapidly reduce future errors. Our findings also indicate that for implicit learning to occur, exposure to sensory prediction errors is necessary. When prediction-error-driven implicit learning and performance-error-driven deliberative strategies combine, improvement in behavior is greater. However, when actions are always successful and strategies are not employed, performance gains are smaller. Flexibly employing cognitive strategies could thus be a way of controlling how much and how quickly we learn from errors. |
