Influence of augmented feedback on coordination strategies

The authors examined the influence of knowledge of results (KR) and concurrent feedback (ConFB) on coordination strategies in learning a 2-finger discrete force-production task. In Experiment 1, 4 groups learned the task under the following feedback regimes: ConFB, ConFB with knowledge of no-KR test (ConFB test information), KR after every trial (100% KR), and KR after every alternate trial (50% KR). Results show that the ConFB group had lower errors during acquisition but the highest errors in the no-KR transfer test. An uncontrolled manifold analysis showed that participants in the ConFB group adopted a strategy that tended to use multiple solutions to achieve the goal during acquisition, but they could not retain this strategy in the no-KR test. Both KR groups retained the same coordination strategy from acquisition into the transfer test, even though this strategy was not conducive to producing a constant force. In Experiment 2, the authors observed these differences even when the time-to-peak force between the groups was constrained to a criterion bandwidth. These results show that ConFB facilitates using different solutions from trial to trial to achieve the same goal, and this may be a reason for poor performance when feedback is removed.