Examining the Effects of Learners’ Background and Social Network Position on Content-Related Interaction via the MOOC Platform
Open Access
- Author:
- Li, Qiyuan
- Graduate Program:
- Learning, Design, and Technology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 09, 2019
- Committee Members:
- Priya Sharma, Dissertation Advisor/Co-Advisor
Priya Sharma, Committee Chair/Co-Chair
Roy Clariana, Committee Member
Simon Hooper, Committee Member
Runze Li, Outside Member - Keywords:
- MOOCs
Interaction
Machine Learning
SEM
Social Network Analysis
Online Learning - Abstract:
- This study aimed to investigate the relationship among MOOC learner’s background, social network position and level of interaction. More specifically, it has two goals: (a) to find, at the individual level, the relationship among each learner’s background, social network position, and interaction; and (b) to find, at the thread level, how the level of diversity among learners in a single thread affected their interaction with one another. To achieve these goals, this research compared the magnitude and significance of factors influencing interaction, including background and social network position and their indicators. Prior research has largely focused on the effects of background and social network position on quantitative features of interaction, such as the number of views, replies, and votes, as well as the duration of threads. This research expanded the scope of the investigation to consider qualitative features such as cognitive engagement level (ACI score) and sentiment polarity in interaction. It also evaluated the extent to which each of these factors influences interaction. In practice, background and social network position could be associated with interaction, but MOOC instructors should be aware of the conditions that enable this to occur. Instructional plans encouraging learners to connect with their peers may lead to constructive activities but not interactive activities, the latter of which are based on peers’ contributions. High levels of cognitive engagement in interaction are generally predicted by negative sentiment. Therefore, negative words are also significant components of interaction in higher-level cognitive engagement. This study reminds MOOC instructors that the posts and comments constructed by negative words are also worth noting. The English-language proficiency levels of individual learners strongly predicted interaction level. MOOCs usually attract a large number of learners whose native language is not English; these learners comprise a subgroup that merits greater attention. This study demonstrated that diversity of background only weakly contributed to a decrease in interaction level for learners. Although negative, the effects of social network position diversity on interaction were generally weak. The top 20% of threads even showed a positive contribution of social network position diversity to interaction. Thread level diversity did not significantly impair interaction level.