Montagne Gilles, Buekers Martinus, Camachon Cyril, De Rugy Aymar, Laurent Michel
CNRS and Université de la Méditerranée, Faculté des Sciences du Sport, UMR Mouvement et Perception, Marseille, France.
Q J Exp Psychol A. 2003 Apr;56(3):551-67. doi: 10.1080/02724980244000620.
This study was designed to better understand the process underlying the learning of goal-directed locomotion. Subjects walked on a treadmill in a virtual reality setting and were asked to cross pairs of oscillating doors. The subjects behaviour was examined at the beginning of the learning process (pretest), after 350 trials (intermediate test), and after 700 trials (posttest). The data were analysed at three different levels, each representing a specific aspect of the global response: performance outcome, displacement kinematics, and current arrival condition. While some aspects of performance outcome suggested the presence of a ceiling effect in the intermediate test, both displacement kinematics and current arrival condition clearly highlighted continuous transformations of the control mechanism involved. The learning process is best described as (1) the establishing of a relationship between specific information and a movement parameter and (2) the optimization of this relationship. The optimization process is characterized by the further exploration of the available behavioural repertoire and by the refinement of the dialogue between information and movement.
本研究旨在更好地理解目标导向性运动学习背后的过程。受试者在虚拟现实环境中的跑步机上行走，并被要求穿过成对的摆动门。在学习过程开始时（预测试）、350次试验后（中期测试）和700次试验后（后测试）对受试者的行为进行检查。数据在三个不同层面进行分析，每个层面代表整体反应的一个特定方面：表现结果、位移运动学和当前到达条件。虽然表现结果的某些方面表明在中期测试中存在天花板效应，但位移运动学和当前到达条件都清楚地突出了所涉及控制机制的持续转变。学习过程最好描述为：（1）在特定信息与运动参数之间建立关系；（2）优化这种关系。优化过程的特点是进一步探索可用的行为库，以及细化信息与运动之间的对话。
