In a recent release, Moodle has shared a slim taste of the new enhancements coming for its Learning Analytics engine. (Formerly Moodle Inspire.) These new features are expected to become available on the open source version of the LMS’s 3.8 release, scheduled for November, 2019. They are an effort to make the powerful open source Machine Learning engine more useful for everyday users.
Content platforms, LMS included, have been key to spearhead the practice of Analytics and the use of technologies to track the behavior of website visitors. Commercial purposes, online marketing in particular, have been key drivers of the technology as well as widespread practice. Not all has been favorable for learners, thought. While general purpose tools and important concepts such as “Digital Maturity” can help organizations understand their data analysis needs, the concerns about applying Analytics lessons meant for buyers into online learning design could still go unnoticed.
What is coming in Learning Analytics with Moodle 3.8
Moodle has been a steady supporter of Learning Analytics since early on, even if more user-friendly tools haven’t always been a priority in the development roadmap. The latest announcement might be seen as a correction to it. The Learning Analytics team, featuring top developers and researchers such as Elizabeth Dalton and David Monllaó. Both are part of Moodle’s LMS team under the helm of Sander Bangma.
The main new Learning Analytics features in Moodle come in the form of insights designed to be actionable for teachers:
- “Students who have not accessed the course recently”: In addition to warning teachers about at-risk students, a new dashboard will allow quick access to their Moodle activity and send them a message from the same page.
- “Summary report”: Tied to renovations to the Forum, it will display activity and behavior statistics for each student, including total posts, replies, views, word and character count.
These join some of the other forecasts already available in Moodle 3.7 and earlier version:
- “Risk of dropping out”: For teachers, it lists students who are less likely to complete a course.
- “Upcoming activities due”: For students, it notifies them about activities soon to be due.
- “No teaching“: For admins, this model informs if no teaching activity is likely to occur for a given course.
- For developers, tools to build engagement indicators, event notifications and an API to connect the Learning Analytics engine with plugins or custom development.
Learning Analytics in the 2020s: Maturity or Adolescence?
The concept of “Learning Analytics” has been with us for more than a decade, experiencing cycles of optimism as early as 2004. It would not be considered as properly taking off before 2009 or early this decade. Naturally, the understanding, expectations and practices on the field have evolved in dramatic ways over the last decade. The question is: Has its evolution being a steady progress, or is its past history a series of missteps and dead ends?
The global EdTech community is still enthusiastic about Learning Analytics, both its potential to enhance learning experience but also a drive to understand its fundamentals at a deeper, theoretical level. A clear illustration to this effect was last October’s launch of the International Journal Of Learning Analytics And AI For Education or “iJAI.” The initiative, by the International Association of Online Engineering, abides by the Open Knowledge principles, providing full access to the peer-reviewed research.
While academia reports progress on the institutional front, by establishing networks and standards, news on practical applications with impact and ability to scale remain scarce.
On the commercial side, ventures in the practice could well predate this century. Practically no private company who today could be claimed as market or thought leader existed 5 years ago. While many academic have decades-long tenures on the field, research has not followed a steady path. The explosion of data science and related technologies could help explain this historical chasm. Machine Learning has definitely disrupted the field, to the point that the lead LMS are almost expected to incorporate it to some capacity. Blackboard Data has been one of the company’s most heavily promoted initiatives. A similar example is DIG by Instructure, Canvas LMS parent company. According to the latest earnings call of the publicly listed company, 15 to 20 institutions are piloting DIG and an “early customer phase” was announced for next year by Instructure CEO, Dan Goldsmith.
In conclusion, there seems to be a consensus about the disruptive potential of Machine Learning and AI in general, but one that is yet to be materialized. In all likelihood the look and practice of Learning Analytics will transform in the next decade.
Free-licensed Moodle 3.8 scheduled for November 18
Moodle 3.8 is expected to hit the virtual shelves on Monday, November 18; as well as the physical ones, as it coincides with the beginning of the first MoodleMoot Global in Barcelona.