Escape from the walled garden

In the past few weeks alone, my conviction that I need to get out of the LMS (Learning Management System) has grown much stronger than ever before. The LMS has been a wonderful way to disseminate PDFs to students (as opposed to the way things were done “back in my day”), an easy way to keep track of grades, and a handy platform to have students engage with each other in online discussions- or at least try to have them do that.

Because the LMS has been vetted by the College for compliance with FERPA, it is a Good Place to keep student information: grades and personal information are safe. Because of the password protection, copy-righted materials are accessible to course participants only, so nobody has to worry about legal consequences as long as we stay in the LMS. It is a nicely walled-off section on the internet. I’ve worked with Moodle, Blackboard, and Canvas. Of these three, Canvas is the most nimble and the user interface from a teacher’s point of view is the best. (If you’ve never used anything else, you may be surprised at that statement.)

Unicorn tied to a tree in a small pen, on a background of lush grass and vegetation. Tapestry from 15th century
The Unicorn in Captivity, in a “Hortus Conclusus” or closed garden. Met Museum

I have in the past year moved away from using Canvas for my course schedule and syllabus. Part of the reason is that I just enjoy experimenting. Canvas did not quite give me the flexibility I wanted to link materials, make space for more creative assignments like timelines and shared maps, or allow for students to explore. Its cut-out cardboard template works well for some things, but not if your mind or courses work in a non-linear fashion. How to share assignments between students, or go gradeless, was also not intuitive enough for me to figure it out, so I explored the wider world of websites and its possibilities. At least this way, I know that if something doesn’t work, it’s my fault, and not a change in the LMS that snuck up on me (as for instance seen in this post and this one).

Recent changes in Canvas make me very suspicious of using the LMS for much more than just parking PDFs of texts I scan for courses behind a password wall. The most important reason has nothing to do with any of what I mention above, and is all about student privacy. Last year, at Digital Pedagogy Lab 2018 in the Digital Literacies track, I learned how to look out for my students’ interests, and make sure to create as much as possible an online environment conducive to learning, rather than a course that did all the cool things just for me. We learned enough about tracking, data collection, and what makes some groups feel safe online or not, and we learned to read a ToS and Privacy Policy and judge the merits and drawbacks. Forcing students to take part in twitter discussions (or they get knocked in the participation grade) is for instance not a good idea. Giving students a chance to opt-out, on the other hand, is. When I decided to work with the students of HST107 China’s Magical Creatures on an Open Education Resource (OER) textbook, we (students and I) signed a Memorandum of Understanding, and they could opt out, or use an alias.

All this to say that although I know that digital redlining is a thing, and that most of the time am forgetful enough not to put preventive measure in place to stop Facebook or Twitter from tracking me after I’m gone from the sites, I know just enough to understand how online data can really, really mess up your life in unexpected ways. In particular “predictive analytics” is something that makes me very uncomfortable: the idea that you leave enough scraps of information out there for a machine to gather them all up and predict with some degree of accuracy what you’d like to buy next, or rather: can be nudged to buy next.

So colour me horrified when I heared a few weeks ago about the plans of Instructure, the company behind Canvas LMS, to turn to big data, machine learning, and AI, with something named DIG. This is their own form of predictive analytics. Here is a quote from CEO Goldsmith with an example how Canvas can help students:

What’s even more interesting and compelling is that we can take that information, correlate it across all sorts of universities, curricula, etc, and we can start making recommendations and suggestions to the student or instructor in how they can be more successful. Watch this video, read this passage, do problems 17-34 in this textbook, spend an extra two hours on this or that. When we drive student success, we impact things like retention, we impact the productivity of the teachers, and it’s a huge opportunity. That’s just one small example.

(source: https://eliterate.us/instructure-plans-to-expand-beyond-canvas-lms-into-machine-learning-and-ai/)

That seems handy, but we all (should) know that these systems do not know anything unless they’re fed enormous amounts of data. I’m sure I did not consent to that knowingly, and I doubt my students did either, so until I get time to read the small print, I’ll refrain from putting more data in there.

But besides informed consent (or more likely: informed refusal!), there is something else that bothers me much more. The idea that we can have a machine determine to such a precise degree what a student needs, for instance. Does the machine know about the student’s financial precarity, requiring her to work multiple jobs to make ends meet? Does it know about the role the student has as main caretaker for his ill parent? Does it know about the micro-aggressions on campus that make it impossible to concentrate on classwork? Does it know that behind the data there is, in fact, a human being with emotions and a life outside of the LMS? It won’t unless the system is connecting with even more data, and I hold my heart for when the moment comes some genius decides to integrate Canvas with all the other data we already hold on students, from financial aid, admissions processes, and (why not?) campus safety on where you’ve swiped that ID card for meals, access to buildings and borrowing books in the library. Surveillance society is here, some assembly still required.

I’m sure more thoughts, and more clearly formed ones will come as I spend more time reading through the articles and blog posts that were recently doing the rounds among twitter’s usual #digped #ethicaledtech voices. I will update this post, or create an additional one, as my thoughts take more structure. For now, I just want my students to know why I am shying away from Canvas. There is a big, huge monster in the walled garden. Don’t feed it.

"Keep calm and don't feed the monster" Red background, white letters under a crown
Source: The Keep-Calm-o-matic

Leave a Reply

Your email address will not be published. Required fields are marked *