This is the first of what I hope to be a series of postings talking about data mining the learning management system at CQUni. It’s a project where I’m attempting to analyze previously untapped data that is held in the LMS database. Data like hit counts, resource utilization, discussion participation, uptimes etcetera can be obtained by applying some simple scripts to the LMS backend database. According to Heathcoate & Dawson 2005 this data can be utilized to aid in the reflection of pedagogical practices in alignment with usage statistics while at CQUni it could potentially be used to inform some aspects of the LMS replacement project.
Up until recently this data has been hidden away in the Blackboard database and had to be accessed via the Blackboard user interface, which is difficult to interpret and often doesn’t work at all.
So what? What use is this data?
One simple example that I’m currently working on is looking at the relationship between student hits on the course and discussion board and their final result for the course. This link appears to be, at first glance, very strong. Another interesting result is the pass/fail rates of courses in generally seems to be linked to the contributions of the teaching staff during the semester and in particular their contributions to discussion boards.
This work is in its infancy although I hope to progress it further over the coming months. The grand plan is for this to form part of my masters which I’m hoping to start next year.
The following graph shows the hit counts of students grouped by grade. The different colours represent the different years the course was delivered. The large increase in 2008 was due to a course redesign that was engineered to promote student engagement. The vertical range is hit count while the horizontal is grade.
I have a lot of work left to do on the scripts alone but more importantly I need to look at prior research into data mining the LMS and talk to some folk from around the Uni to determine their requirements. More to come.
Heathcoat & Dawson. (2005). “Data Mining for Evaluation, Benchmarking and Reflective Practice in a LMS.” E-Learn 2005: World conference on E-Learning in corporate, government, healthcare and higher education.