Discussion Analytics: Identifying Conversations and Social Learners in FutureLearn MOOCs

LAK 2017 FutureLearn Workshop | Vancouver | 13th March 2017

Discussion among learners in MOOCs has been hailed as beneficial for social constructive learning. To understand the pedagogical value of MOOC discussion forums, several researchers have utilized content analysis techniques to associate individual postings with differing levels of cognitive activity. However, this analysis typically ignores the turn taking among discussion postings, such as learners responding to others’ replies to their posts, learners receiving no reply for their posts, or learners just posting without conversing with others. This information is particularly important in understanding patterns of conversations that occur in MOOCs, and learners’ commenting behaviors. Therefore, in this paper, we categorize comments in a FutureLearn MOOC based on their nature (post vs. reply to others’ post), classify learners based on their contributions for each type of posting, and identify conversations based on the types of comments composing them. This categorization quantifies the dynamics of conversations in the discussion activities, allowing monitoring of on-going discussion activities in FutureLearn and further analysis on identified conversations, social learners, and types of comments with an unusual number in a course step.

Slideshare URL:

Event URL: