Trifold Benefit of Educational Data Mining and Learning Analytics

 

My focus for blog reflection this week is exploring the different ways that learning analytics/educational data mining can

 

a) assist individual self-regulating learning 

 

As described by Heffernan (2014), the ASSISTments ecosystem is a powerful platform that uses data mining to "assist as it assesses", which gives it the ability to provide feedback in real time, as students need it. As a platform, it can quickly deploy a wide range of content and experiments, based on student performance. The platform allows for easy construction of scaffolding questions, which helps support student problem solving, such as breaking a complex topic into smaller steps. For example, geometry scaffolding questions might include questions on each of the following: congruence, perimeter, equation solving and finally the substitution (p 475). While working through these problems, students would receive hints and feedback to facilitate the self-regulated learning. Over the course of this semester in fact, the ability to provide feedback on a granular level has been a constant in AI-supported education. In order to effect mastery learning, the question bank needs to be robust. Another support for self-regulation was the Automatic Reassessments an Relearning System (the rather unfortunate acronym of ARRS) which spaced out practice to foster better learning transfer, say, over 7, 14, 28 and 60 days.

 

b) facilitate successful collaborative learning

 

The ASSISTments learning ecosystem also allows for classroom collaboration, between teacher and students, and amongst the students themselves. The teachers were invested in the platform because they had helped feed the question scaffold into the database and were excited be a part of its implementation. Parents, too, are able to monitor (and encourage) student learning.

Tissenbaum (2018) also demonstrates just how robust this collaboration can be using real-time software agents and data-driven orchestration. The design of their tablet application and its ecosystem is rooted in the Knowledge and Community Inquiry framework, in which learning is encouraged at the class, community (and world) level in addition to the student-level learning. Students were able to co-construct the knowledge base, to advance the community cause in addition to their own learning. This was really key to their entire approach. Their application depended on data driven orchestration and real-time agents (again, feedback, scripts, orchestration) to foster such collaboration. As well, student groupings could be managed in a way that pulled from prior student responses and made the groupings more meaningful than just random assignment.

 

c) Assess the state and flow of a classroom in real-time

 

In fact, one of the threads that wove the readings this week was the focus on the teacher for such designs. Heffernan notes the importance of designing for the teacher, not the instructional designer, so that these tools are easy to use. Moreover, unlike a cognitive tutor, he notes it is the teacher who decides when the students are ready to advance on the ASSISTments platform.

Tissenbaum's tablet provided data analytics to the teacher that simply would have been too difficult to carry out without software agents. In fact, its primary focus was on the delivery and management of the entire curriculum, well beyond the student learning and interactions. In this case, the teacher was actually the co-designer. I also though Tissenbaum's model was a refreshing contrast to the earlier reading of McLaren (2010) in which it seemed their intention was to supplant the role of teacher through AI.

 

References

 

Baker, R. S. (2016). Stupid tutoring systems, intelligent humans. International Journal of Artificial Intelligence in Education, 26(2), 600-614.

Heffernan, N. T., & Heffernan, C. L. (2014) The ASSISTments ecosystem: Building a platform that brings scientists and teachers together for minimally invasive research on human learning and teaching. International Journal of Artificial Intelligence in Education 24, 4, 470–497.

 

McLaren, B. M., Scheuer, O., & Miksáko, J. (2010). Supporting collaborative learning and e-discussions using artificial intelligence techniques. International Journal of Artificial Intelligence in Education, 20(1), 1–46.

 

Tissenbaum, M., & Slotta, J. (2019). Supporting classroom orchestration with real-time feedback: A role for teacher dashboards and real-time agents. International Journal of Computer-Supported Collaborative Learning, 14(3), 325-351.