Department of Mechanical and Aerospace Engineering, University of California, Irvine, 4200 Engineering Gateway, Irvine, CA 92697, USA.
Exp Brain Res. 2010 Mar;201(2):119-31. doi: 10.1007/s00221-009-2014-z. Epub 2009 Sep 29.
Performance errors drive motor learning for many tasks. Some researchers have suggested that reducing performance errors with haptic guidance can benefit learning by demonstrating correct movements, while others have suggested that artificially increasing errors will force faster and more complete learning. This study compared the effect of these two techniques--haptic guidance and error amplification--as healthy subjects learned to play a computerized pinball-like game. The game required learning to press a button using wrist movement at the correct time to make a flipper hit a falling ball to a randomly positioned target. Errors were decreased or increased using a robotic device that retarded or accelerated wrist movement, based on sensed movement initiation timing errors. After training with either error amplification or haptic guidance, subjects significantly reduced their timing errors and generalized learning to untrained targets. However, for a subset of more skilled subjects, training with amplified errors produced significantly greater learning than training with the reduced errors associated with haptic guidance, while for a subset of less skilled subjects, training with haptic guidance seemed to benefit learning more. These results suggest that both techniques help enhanced performance of a timing task, but learning is optimized if training subjects with the appropriate technique based on their baseline skill level.
性能错误会影响许多任务的运动学习。一些研究人员认为，通过触觉引导减少运动错误可以通过演示正确的运动来帮助学习，而另一些研究人员则认为人为增加错误会迫使运动更快、更完整地学习。本研究比较了这两种技术——触觉引导和错误放大——在健康受试者学习玩电脑弹球游戏时的效果。该游戏要求学习在正确的时间使用手腕运动按下按钮，以使击球器击打下落的球到随机定位的目标。根据感知到的运动启动时间错误，使用一种机器人设备来减小或增大手腕运动的速度，从而减小或增大错误。经过错误放大或触觉引导训练后，受试者显著降低了他们的时间错误，并将学习推广到未训练的目标。然而，对于一组更熟练的受试者来说，与与触觉引导相关的减小错误相比，用放大错误训练会产生显著更大的学习效果，而对于一组不太熟练的受试者来说，用触觉引导训练似乎更有利于学习。这些结果表明，这两种技术都有助于提高定时任务的性能，但如果根据受试者的基线技能水平，用适当的技术训练他们，学习效果会得到优化。
