学习分析视阈下在线学习行为模式研究

Research on Online Learning Behavior Pattern from the Perspective of Learning Analysis

传统的在线学习行为分析方法通常将学习过程割裂成独立的行为操作单元,不利于从整体分析把握在线学习活动的过程与规律。学习分析视阈下的行为模式分析则从学习行为之间的时间序列与相互关系入手,旨在揭示个体和学习群体的学习行为特征,识别不同学习群体的行为差异,诊断学习活动中的关键学习事件并预测学习者的学习绩效。学习分析视域下包含描述性、诊断性、预测性三个维度相结合的在线行为模式分析框架,以Wiki协作知识建构行为模式作为案例研究对象,采用滞后序列分析等方法对Wiki编辑行为模式进行三个维度的分析与探索。结果表明,Wiki协作知识建构过程中学习者倾向于不断精炼、补充修改内容;在知识联结、知识收敛等高层次的知识建构行为中,高低成就组存在显著差异。研究进一步凝练关键行为序列建立预测行为序列集,为预测高水平协作编辑绩效提供支持。整合描述性研究、解释性研究和预测性研究的学习行为模式分析框架对于探索在线学习行为规律与预测学习绩效具有重要价值。

Traditional online learning behavior analysis methods usually divide learning activities into independent behavior units.It is not conducive to analysis of the characteristics and the understanding of online learning.From the perspective of learning analysis,the behavioral pattern analysis focus on the correlation and time series of learning behaviors,reveal the behavioral characteristics of individuals and learning groups,identify the behavioral differences among different learning groups,diagnose the key learning events in learning activities,and predict the learning performances of learners.This study proposed a research framework for online behavior patterns based on descriptive,diagnostic and predictive nature.It analyzed and explored the behavior pattern of wiki editing by using lag sequence analysis from three dimensions.The results show that learners constantly refined,added and modified content in the process of wiki collaborative knowledge construction.Among the high-level knowledge construction behaviors,such as knowledge connection and knowledge convergence,there were important differences between high and low achievement groups.The study further summarized the key behavior sequence and established the prediction behavior sequence set to provide support for the prediction of high-level collaborative editing performances.The study also revealed that the learning behavior research framework integrating descriptive research,explanatory research and predictive research is of great value for exploring the rules of online learning behavior and predicting learning performance.