The Viability of Topic Modeling to Identify Participant Motivations for Enrolling in Online Professional Development

Autor: Heather Allmond Barker, Hollylynne S Lee, Shaun Kellogg, Robin Anderson
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: Online Learning, Vol 28, Iss 1 (2024)
Druh dokumentu: article
ISSN: 2472-5749
2472-5730
DOI: 10.24059/olj.v28i1.3571
Popis: Identifying motivation for enrollment in MOOCs has been an important way to predict participant success rates. But themes for motivation have largely centered around themes for enrolling in any MOOC, and not ones specific to the course being studied. In this study, qualitatively coding discussion forums was combined with topic modeling to identify participants’ motivation for enrolling in two successive statistics education professional development online courses. Computational text mining, such as topic modeling, is a learning analytics field that has proven effective in analyzing large volumes of text to automatically identify topics or themes. This contrasts with traditional qualitative approaches, in which researchers manually apply labels (or codes) to parts of text to identify common themes. Combining topic modeling and qualitative research may prove useful to education researchers and practitioners in better understanding and improving online learning contexts that feature asynchronous discussion. Three topic modeling approaches were used in this study, including both unsupervised and semi-supervised modeling techniques. The three topic modeling approaches were validated and compared to determine which participants were assigned motivation themes that most closely aligned to their posts made in an introductory discussion forum. A discussion of how each technique can be useful for identifying topical themes within discussion forum data is included. Though the three techniques have varying success rates in identifying motivation for enrolling in the MOOCs, they do all identify similar themes for motivation that are specific to statistics education.
Databáze: Directory of Open Access Journals