学习者在线学习水平的学习分析模型研究——临场感学习分析模型构建与方法探索  

Learning Analytics Model of Online Learning:A Study on Construction and Its Methods of Presence Learning Analytics Model

作　　者：冯晓英[1] 刘月[1] 吴怡君 HAN Xibin;CHENG Jiangang(Institute of Education, Tsinghua University, Beijing 100084)

机构地区：[1]北京师范大学远程教育研究中心,北京100875 [2]北京师范大学教育技术学院,北京100875

出　　处：《电化教育研究》2018年第7期40-48,共9页E-education Research

基　　金：中央高校基本科研业务费专项资金资助"基于学习分析技术的在线交互质量评价的研究"(项目编号:SKZZB2015004)

摘　　要：文章从学习分析的视角,旨在构建在线学习者临场感水平的学习分析模型,同时探索有效构建学习者模型的方法路径。文章选取了两个典型的在线课程作为研究案例,采用内容分析、社会网络分析、相关分析、岭回归、多元线性回归等方法分别对两个案例进行建模,并对两组模型和数据进行相互验证与比对。研究发现:(1)两组学习分析模型均具有较高的解释度和通用性,能够较好地综合表征学习者的在线学习水平;(2)A组模型适合我国典型的以"自主学习+教师辅导(师生互动)为主、生生交互为辅"的在线课程,B组模型则适用于以"自主学习+生生交互为主、师生交互为辅"的在线课程;(3)"相关分析+岭回归+多元线性回归"的建模方法有助于提高模型的解释度,利用多组数据建模并验证模型通用性的思路具有可行性。From the perspective of learning analytics, this study aims to build a learning analytic model for the presence level of online learners and explores effective methods and ways to construct a learner model. This study selects two typical online courses as research cases, adopts content analysis,social network analysis, correlation analysis, ridge regression and multiple linear regressions to build models respectively, and verifies and compares those two models and data as well. Research findings include：（1） two learning analytics models have a high degree of interpretation and universality, and can represent the learning level of online learners comprehensively.（2） Model A is suitable for typical online courses featured by ＂autonomous learning ＋ teacher guidance（teacher-student interaction） supplemented by student-student interaction ＂ in China; model B is applicable to online courses which focus on ＂autonomous learning ＋ student-student interaction supported by teacher-student interaction＂.（3） The modeling method by correlation analysis, ridge regression and multiple linear regressions is helpful to improve the interpretation of the model. Therefore, it is feasible to use multiple sets of data to model and verify model universality.

关 键 词：学习分析 在线学习 临场感 建模 方法 

分 类 号：G434[文化科学—教育学]