网络化学习中学习者认知负荷的评估方法

Method for Assessment of Cognitive Load in the E-learning Environment

本研究通过设计模拟的网络化学习实验，探讨网络化学习过程中认知负荷的评估方法。结果显示，心理努力、任务主观难度、注视时间、注视次数、主任务反应时、主任务正确率等评估指标对认知负荷变化敏感；采用多维综合评估模型对认知负荷进行测量总体上比采用单一评估指标的测量更为有效。研究表明，BP网络和自组织神经网络两种神经网络模型对认知负荷的测量结果优于传统的因素分析方法。

Through a simulated E-learning experiment designed to find the best methods to assess cognitive load in the E-learning environment， we found that six indices prove to show sensitivity to changes in cognitive load: mental efforts， perceived task difficulty， duration of fixation， the number of fixation， the response time and correct rate in the primary task. In addition， it is more efficient to assess the load based on the multidimensional synthetic assessment， rather than single assessment index. Two ANN models， namely， BP neural network and self-organizing feature map， show a better result in measuring cognitive load than the traditional technique of factor analysis.