Sævland Werner, Norman Elisabeth
Department of Psychosocial Science, Faculty of Psychology, University of Bergen Bergen, Norway.
Front Psychol. 2016 Jun 7;7:808. doi: 10.3389/fpsyg.2016.00808. eCollection 2016.
Implicit learning is usually studied through individual performance on a single task, with the most common tasks being the Serial Reaction Time (SRT) task, the Dynamic System Control (DSC) task, and Artificial Grammar Learning (AGL). Few attempts have been made to compare performance across different implicit learning tasks within the same study. The current study was designed to explore the relationship between performance on the DSC Sugar factory task and the Alternating Serial Reaction Time (ASRT) task. We also addressed another limitation of traditional implicit learning experiments, namely that implicit learning is usually studied in laboratory settings over a restricted time span lasting for less than an hour. In everyday situations, implicit learning is assumed to involve a gradual accumulation of knowledge across several learning episodes over a longer time span. One way to increase the ecological validity of implicit learning experiments could be to present the learning material repeatedly across shorter test sessions. This can most easily be done by using a web-based setup in which participants can access the material from home. We therefore created an online web-based system for measuring implicit learning that could be administered in either single or multiple sessions. Participants (n = 66) were assigned to either a single session or a multiple session condition. Learning occurred on both tasks, and awareness measures suggested that acquired knowledge was not fully conscious on either of the tasks. Learning and the degree of conscious awareness of the learned regularities were compared across conditions and tasks. On the DSC task, performance was not affected by whether learning had taken place in one or over multiple sessions. On the ASRT task, RT improvement across blocks was larger in the multiple-session condition. Learning in the two tasks was not related.
内隐学习通常是通过个体在单一任务上的表现来研究的，最常见的任务是序列反应时（SRT）任务、动态系统控制（DSC）任务和人工语法学习（AGL）。在同一研究中，很少有人尝试比较不同内隐学习任务的表现。本研究旨在探讨DSC制糖厂任务和交替序列反应时（ASRT）任务的表现之间的关系。我们还解决了传统内隐学习实验的另一个局限性，即内隐学习通常是在实验室环境中，在持续不到一小时的有限时间跨度内进行研究的。在日常情况下，内隐学习被认为涉及在较长时间跨度内跨多个学习阶段逐渐积累知识。提高内隐学习实验生态效度的一种方法可能是在较短的测试阶段反复呈现学习材料。这可以通过基于网络的设置最轻松地完成，参与者可以在家中访问材料。因此，我们创建了一个基于网络的在线系统来测量内隐学习，该系统可以在单个或多个阶段进行管理。参与者（n = 66）被分配到单个阶段或多个阶段的条件下。在两个任务上都发生了学习，意识测量表明，在任何一个任务上获得的知识都不是完全有意识的。比较了不同条件和任务下的学习情况以及对所学规律的有意识程度。在DSC任务上，学习是在一个阶段还是多个阶段进行，对表现没有影响。在ASRT任务上，多个阶段条件下各块的反应时改善更大。两个任务中的学习没有关联。
