Using HMM to compare interaction activity patterns of student groups with different achievements in MPOCs. Issue 5 (18th August 2019)

• Record Type: Journal Article
• Authors: Sun, Di; Cheng, Gang; Xu, Pengfei; Zheng, Qinhua; Chen, Li
• Is Part Of: Interactive learning environments. Volume 27:Issue 5/6(2019)
• Page Start: 766
• Page End: 781
• Publication Date: 2019-08-18
• Subjects: HMM -- compare -- activity patterns -- different achievements -- MPOCs; Educational technology -- United States -- Periodicals; 371.33
• DOI: 10.1080/10494820.2019.1610780
• Languages: English
• ISSNs: 1049-4820

Abstract

ABSTRACT: With the development of online learning, LMSs accumulated huge amounts of students' interaction data. Unfortunately, with the support of LMSs data, few researchers put a sight on interaction research in MPOCs. Particularly, comparing interaction activity patterns of different achievement student groups and in different course processes in MPOCs has been paid less attention. This paper generates hidden Markov models to identify frequently occurring interaction activity sequence patterns of High/Low achievement groups in the Learning/Exam processes under MPOCs settings. The results demonstrate that High-achievement students especially focused on content learning, assignments, and quizzes to consolidate their knowledge construction in both Learning and Exam processes, while Low-achievement students extremely did not perform the same. Further, High-achievement students adjusted their learning strategies based on the goals of different course processes; Low-achievement students were inactive in the learning process and opportunistic in the exam process. In sum, the comparative analysis implies that certain interaction patterns may distinguish the High-achievement students from the Low-achievement ones, and students change their patterns more or less based on different course processes. It is also helpful for students, instructors to adjust their strategies and make decisions, and for developers, and administrators to build recommendation systems based on objective and comprehensive information.