Setting the pace: examining cognitive processing in MOOC discussion forums with automatic text analysis

Abstract

Learning analytics focuses on extracting meaning from large amounts of data. One of the largest datasets in education comes from Massive Open Online Courses (MOOCs) that typically feature enrollments in the tens of thousands. Analyzing MOOC discussion forums presents logistical issues, resulting chiefly from the size of the dataset, which can create challenges for understanding and adequately describing student behaviors. Utilizing automatic text analysis, this study built a hierarchical linear model that examines the influence of the pacing condition of a massive open online course (MOOC), whether it is self-paced or instructor-paced, on the demonstration of cognitive processing in a HarvardX MOOC. The analysis of 2,423 discussion posts generated by 671 students revealed the number of dictionary words used were positively associated with cognitive processing while analytical thinking and clout was negatively associated. We found that none of the student background information (gender, education), status of the course engagement (explored or completed), or the course pace (self-paced versus instructor paced) significantly influenced the cognitive processing of the postings.

Exploring the relationship between clout and cognitive processing in MOOC discussion forums

“…In a study on pacing within MOOC discussion forums, clout was negatively associated with cognitive processing, perhaps due to the showing of more confidence not fostering as much openness for discussion (Moore, Oliver, & Wang, 2019). However, one of the limitations of that study is that clout is a nontransparent summary variable of LIWC.…”

“…To expand our understanding of the relationship between clout and cognitive processing, we designed a new study using the same dataset from the study by Moore et al . (2019) to examine what role the pacing condition and other predictor variables had on the relationship between cognitive processing and clout. Considering the previously found relationships between the cognitive processing dimension and clout, this study provides a more focused approach to delve deeper in the specifics of those relationships.…”

“…Shah (2018) estimated that as of 2017, nearly 81 million students had registered for a MOOC. These courses are either self‐paced or instructor‐paced (Moore et al ., 2019) and are primarily classified as either a cMOOC or an xMOOC (Bonk, Lee, Reeves, & Reynolds, 2017). Most MOOCs are xMOOCs and these have the familiar course structure that focuses on instructor‐generated content, discussion forums and self‐assessments (Adams et al ., 2014; Bonk et al ., 2017).…”

“…In MOOCs, much of the course content engagement and student‐to‐student engagement happens within the discussion forums (Gardner & Brooks, 2018; Martin & Ndoye, 2016; Moore et al ., 2019). This creates a rich data source that can provide unique opportunities to study learner engagement and experience within these massive courses (Gardner & Brooks, 2018).…”

Informing Makerspace Outcomes Through a Linguistic Analysis of Written and Video-Recorded Project Assessments

“…data mining technique to reveal outcomes of interest. Linguistic Inquiry Word Count (LIWC) is one popular data mining package that outputs different linguistic variables to represent a text (Moore, Oliver & Wang, 2019;Pennebaker, Boyd, Jordan & Blackburn, 2015). We analyzed the presence of four LIWC summary variables in makerspace assessments (analytical thinking, authenticity, clout, and emotional tone) as these variables were seen as related to the four TLDF dimensions advocated for makerspaces and might efficiently provide evidence of the dimensions since a TLDF instrument was not otherwise available.…”