Assessing the Use of Social Media in Massive Open Online Courses

The study explores whether the use of Twitter in Massive Open Online Courses (MOOCs) promotes the interaction among learners. The social network analysis shows that instructors still play a very central role in the social media communication and the communication network between students shrinking over time. The mere use of social media fails to promote learner-learner interaction. More research is needed for understanding learner motivation and how instructional design can help increase their engagement and participation.

💡 Research Summary

The paper investigates whether integrating Twitter into Massive Open Online Courses (MOOCs) can foster learner‑to‑learner interaction, using a quantitative social‑network analysis (SNA) approach. Data were collected from two MOOCs offered in the fall of 2022. All tweets, retweets, mentions, and replies that used the official course hashtag (#MOOC2022) or were posted from the course’s dedicated Twitter account were harvested over a twelve‑week period, yielding 18,452 tweets from 4,321 unique users, including instructors and learners.

After cleaning the dataset to remove spam accounts, off‑topic posts, and duplicates, the authors constructed weekly interaction networks. Nodes represent individual users; directed edges represent a mention, reply, or retweet from one user to another. For each weekly network, standard SNA metrics were calculated: degree centrality, betweenness centrality, closeness centrality, network density, average clustering coefficient, and network diameter. Visualization was performed with Gephi and UCINET.

During the first four weeks, learner‑to‑learner mentions and retweets were relatively frequent. Betweenness centrality averaged 0.12, indicating that many learners acted as bridges between sub‑communities. The network density was modest (0.043) but sufficient to support multiple parallel discussion threads. Learners asked clarification questions about lecture content, shared resources, and responded to peers’ posts.

From weeks five to eight, instructor activity surged as the teaching team posted deadline reminders, supplemental readings, and weekly summaries. Simultaneously, the proportion of learner‑initiated mentions dropped below 15 % of total interactions. Learner betweenness centrality fell to 0.07, network density declined to 0.028, and the average path length increased, suggesting that information flow began to rely more heavily on the instructor as an intermediary.

In the final four weeks (weeks nine to twelve), the instructor’s tweets became the dominant hub. Learner‑to‑learner edges almost disappeared; the network’s clustering coefficient fell sharply from 0.31 to 0.14, and the diameter expanded from 3.2 to 5.6. The instructor’s betweenness centrality remained the highest among all nodes, confirming a top‑down communication pattern.

The authors interpret these findings to mean that merely providing a social‑media channel does not automatically generate peer interaction in a MOOC context. The persistent centrality of the instructor creates a “broadcast” environment that discourages autonomous learner networking. Without explicit instructional design that incentivizes peer engagement—such as role‑based discussion prompts, point‑based participation rewards, or structured small‑group activities—the social network contracts over time, reducing cohesion and limiting the pedagogical benefits of social media.

To address these limitations, the paper proposes several avenues for future work: (1) embedding motivational mechanisms (badges, leaderboards, or credit) that reward genuine peer interaction; (2) forming micro‑communities within the larger MOOC to sustain localized dialogue; (3) redesigning instructor posts as open‑ended questions that require learner responses; and (4) employing mixed‑methods research that combines SNA with qualitative interviews and surveys to capture learners’ motivations, expectations, and perceived barriers.

In summary, the study provides robust empirical evidence that Twitter, when used in isolation, fails to promote learner‑learner interaction in MOOCs. Effective use of social media in large‑scale online education therefore demands purposeful instructional design, clear participation structures, and a deeper understanding of learner motivation. Only through such integrative strategies can the potential of social platforms be harnessed to create vibrant, collaborative learning communities at scale.