If you own a sizable Moodle site that’s been gathering data about student behavior for a while, what can you do with such rich information? The following data science research using data from Moodle sites could let you know if you’ve been sitting on a gold mine all this time.
Dive deep into desirable activity patterns through clustering
A Java-based implementation of a “General Learning Index Manager” was built to access the Moodle database in search of history of interactions and communications between peers, which are later matched to final scores, also from Moodle. It suggests that, in addition to quantity and quality of interactions, a consideration should be made regarding “relevance” of data. Depending on the subject, a “clustering” technique to divide the interactions according to its contribution on specific courses and topics would help understand what constitutes desirable behavior and interactions in each context. Network effects, for example, are beneficial in Forums, but not in Quizzes or Assignments.
Graph browsing patterns to quantify student personality
It sounds obvious in hindsight, but one of the learning activities most encouraged by Moodle and similarly rich environments, is “browsing.” On the dashboard or any given page, the user has dozens of links to choose from. Could the specific sequence of navigations amount to a fingerprint of sorts for each student’s learning style? A preliminary analysis considered 5 navigational choices of 63 students, evaluated in terms of their expertise or decisiveness of the path taken, then contrasted to the Kolb Learning Style Inventory. It seems to highlight the value of paying attention to browsing styles, ranging from linear to open-ended, in the design of courses and platforms.
Improve Moodle architecture for faster data retrieval
In modern user-centered systems, including LMS, performance issues should be considered part of the User Experience. The principles of information architecture are only recently beginning to understand that users should be the focus of the experience and must be served before frameworks, and their needs addressed properly, where a specific structure is no longer an excuse for delays or limited functionality. The authors suggest specific data management practices that would make Moodle more responsive to the needs of users, such as key limits on file upload, decision-tree-based cache optimization, and even basic data mining techniques that quickly enhance the structure of courses.
This Moodle Practice related post is made possible by: eThink Education, a Certified Moodle Partner that provides a fully-managed Moodle experience including implementation, integration, cloud-hosting, and management services. To learn more about eThink, click here.