Given the recent hacking of Canvas and the dislocation that occurred as a result on campuses across the country, many must be asking themselves - was this a Black Swan or a portent? If instructors assign a non-zero probability to the latter, they then will want to know what they can do in advance to mitigate the consequences of the next event. This post is motivated by those concerns.
Now it's time for the disclaimer, my usual one and another that is specific to the content of this post. I write these posts primarily to provoke the reader's thinking on the matters of concern. My proffered solutions may very well be off base. I want to admit that up front. Nonetheless the post can be useful in getting the reader to consider other possibilities that would be better. That value of the post, then, is in getting readers to reflect on the issues in a serious manner, rather than to just have them worried but with no strategy on how to self-protect. I also need to note here that I am a has-been. I last taught in fall 2019, so have no experience as to the consequences of AI on instruction. Further, I retired back in 2010 from doing ed tech work, first for the Campus at Illinois, then for the College of Business. At the time, I might have been considered to be an informed insider. This commentary which initially appeared in then Educause Quarterly, Dis-Integrating the LMS, reflects my mindset from back then, and which continued into the time when I was teaching the one course a year on the Economics of Organizations that I did in retirement. The rest of the post leverages those experiences, even if they are now out of date.
There is an obvious form of self-protection an instructors might take - use an alternative host (or set of hosts) other than the campus LMS, hosts that are less likely to experience an outage at a critical juncture in the course, perhaps because those hosts are used mainly for functions other than teaching and learning. But I can't leave well enough alone, so I am going to take a kill-two-birds-with-one-stone approach. Once alternative hosts are on the table, might that readily enable an improvement in pedagogy as well? Some of my suggestions below will be along those lines.
Make Much of the Course Website Publicly Available
This is first and foremost to reflect an ideal from 20+ years ago - Information wants to be free. I will try to illustrate some of the learning benefits from doing so in what follows, but let me start with this one. As this use of an alternative host is done by instructor opt-in and not as a requirement by the instructor's academic department or the campus, it is a way for the instructor to demonstrate some pride in how the instructor goes about teaching. Further, once having made such a commitment, it places an implicit obligation on the instructor in sustaining the quality of the online content that is posted on the course Website.
For the sake of illustration, here is my course site from a decade ago. This is the penultimate page on the site, as I made a few posts there after the course concluded that are less relevant for illustration purposes. The page that is displayed shows the end-of-semester posts. Let me also distinguish the course site itself from course content that I created. As to the latter, go to the tabbed page called Useful Links. The first of these was to Moodle, which the College of LAS supported as an alternative to the Campus LMS. I used it sparingly, but I did use it. The next link is to my PowerPoint files (and some PDFs as well) that are in my Box.com account. These files are publicly available, but are probably not discoverable without the Useful Links page. Because I subsequently required my students to make "Virtual Elevator Speeches" as part of their course projects (and because some students would only add the course during the second week of the semester), the file First Class Session.pptx is done in that style. I encourage the interested reader to download it and have a look. The slides themselves have images to illustrate the ideas and only a limited amount of text. There is discussion in the Notes pane. The presentation autoplays in slideshow mode, and then has musical accompaniment. This is both to convey a sense of motion and to time the entire presentation to coincide with the duration of the musical selection.
The reader might like the effect created, but balk at the idea of creating similar such presentations, as it is quite time consuming to do so. Let me get back to that issue later in the post. Likewise, the next link to my Excel files also requires some explanation, which I want to defer till later as well.
I want to note that there is also some video content of my own creation, which was produced to illustrate the economics in the Excel files. Each of those videos is embedded in a post on the class site. Here is an illustration of that. I also embed a variety of content that I've found on the open Internet and that is relevant for course purposes. This gets me to the next point.
A significant sticking point with having an open class site is the use of copyrighted material, where the instructor does not have permission to use the content in an open manner. This will happen, for example, if the instructor utilizes content supplied by the textbook publisher. The reader will note that in my class, while I had a textbook, it was more or less a reference book for the students and I didn't teach to it at all. I'm on record that we need to be moving away from textbooks in college instruction and toward inquiry based methods. If wishing would make it so. My hope is that self-protecting against the campus LMS failing might encourage a step in that direction.
Let's push on. Anyone looking at my course page will notice the Student Blogs in the left sidebar. I want to give a bit of background on this and how I used student blogs in this course. I first taught with student blogs back in fall 2009, in a course for students in the Campus Honors Program called Designing for Effective Change, so it wasn't an economics course. My experience there informed my later use of student blogs in my economics course as I will describe below. I wrote about that initial experience in a piece for Inside Higher Ed called Teaching With Blogs.
The first lesson, one I hadn't figured out in advance, is that the student blogging constitutes formative thinking on their part. They are okay doing that, provided the environment is safe. Here, safety refers to whether a potential future employer can see the student posts. They were afraid that might happen. Having students blog under an alias addresses this concern. So, in the economics course I assigned each student an alias, the name of a famous economist concatenated with the course rubric and the semester the course was being offered. This way, the aliases remained unique, even as I reused the names of the famous economists from one year to the next.
The next lesson is about the prompts I gave to the students on which they based their blog posts. (They were also free to blog on a topic of their own choosing that was relevant to the class, but that happened only rarely.) I wanted them to build connections between the various economics ideas we were studying and their own experiences, as I thought they would better learn the economics that way. This bringing in their own experiences also served to differentiate their posts somewhat. As I read each post and commented on it, I needed to maintain a freshness of perspective. I tried to have my comments situated in what the student wrote rather than comment about the intended lessons from the underlying economics. This gave some novelty to the comments and made commenting more enjoyable and less drudge work for me. Also, I required each student to respond to my comment with their own comment, to acknowledge what I had said. Sometimes other students would comment as well. This student blog provides a good example along these lines.
And while it is not possible to observe this, the expectation is that students would have done considerable pre-writing before making their posts. I referred to the pre-writing as mental puttering and wrote this piece to explain that it is tied with making connections among the ideas but that forming such connections doesn't happen in such a straightforward manner. As such, it is time consuming, which brings to mind the question, can this sort of approach survive now that AI is so prevalent? My guess is that it can't for the following reason. Many students view coursework as a series of obstacles, which they try to get past as quickly as possible. The thought of mental puttering is alien to such students and they would use AI to produce a tolerably good post in as short a time as possible. That there are aspects of the post that pertain to their own experience might slow them down a bit. But would that be sufficient? I doubt it. Before considering what might work let me make the following observation.
As I believe that most students are way too grade conscious, I did not assign grades to individual blog posts. This is in accord with the view I articulated here. Instead, I borrowed an idea from the Art and Design folks and embraced portfolio grading. While students would get a small credit for each post they submitted, halfway through the course I'd give one grade on all the posts till then. At the course conclusion I would give a grade for the remaining posts. The reality is that even for the earnest student it would take a while to get comfortable with the blogging, so one would expect to see some improvement in performance simply based on that. Then one might hope thereafter that the posts would become a bit more nuanced. For the better students that actually happened.
And I'd like to note a learning benefit from having the student blogs made public. Students who are finding it difficult to write posts or who have received commentary from me that they view as somewhat harsh can read the posts of other students and my comments on those. They might then emulate in approach the posts they found that were well received. This is a way of learning to write, by reading what other writers have done and then basing one's own style on the reading that was most enjoyable. Further, they might come to understand that my comments, even if critical, are intended to improve their writing. The comments, then, are a form of coaching. Even if it is awkward at first to receive such coaching, I would hope that eventually the students would want that, as they can see it is helpful for their own learning.
Can the approach be modified sufficiently that it remains effective even in the presence of AI yet produces similar outcomes to what I was able to achieve 10 years ago? I've been mentally toying with various tweaks that might be tried to achieve this end. One of those is to have students, either in pairs or groups of three, have online video chats that are recorded. During these chats part of what the students would say is similar to what they would have posted in their blog in the past, including their personal experience that is relevant to the economics under consideration. But, as listeners, they'd also then react to what the other student says, asking clarifying or follow up questions, so the chat very much becomes a conversation. The instructor would get to see the video itself, but to preserve a sense of anonymity for the students, only the transcript of the video would be made public. If these chats were viewed as giving a performance to be seen by the instructor, then students might feel it necessary to prepare in advance for them. Such preparation would be akin to the pre-writing I mentioned above, even if AI were utilized as part of the preparation. Perhaps this can work. I'd like to encourage others to try it to see if it can.
Have students complete weekly assignments where the students bring in their relevant personal experiences that that are published on the course Website under a student alias. The instructor should comment on these submissions, but only grade submissions in batches, once or twice a semester.
There may be subject matter that is too disembodied for prior student experience to matter. In my class the pure economic theory students were to learn fits in this category. The standard way to present this theory is via mathematical modeling. Yet undergraduate students who are neither math nor engineering majors, most econ majors are in this category, may be uncomfortable with abstract presentation of the math modeling. Is there a way for these students to visualize the economic theory without them finding the presentation off-putting? The Excel homework I designed for the course was aimed at addressing this question. Those workbooks are examples of what I had earlier referred to as Dialogic Learning Objects.
To get a sense of this, go back to the Useful Links tab on the Course Website and this time choose the Excel Files link. Select the Excel Tutorial file and download it. This is a tutorial about how Excel was to be used in my course. Enter some short textual content in the NetID box and then choose an alias from the pulldown menu. You need to include that information for the completed tutorial to generate a "key" which can then be submitted to get credit for completing the tutorial. Like the individual blog posts, the tutorials are credit - no credit. All the questions must be answered correctly to get credit. But what is submitted is only the key.
You should be able to complete the rest of the tutorial without difficulty and can submit your key if you'd like via the form link that is provided. I used Google Forms for that function. I included a comment box on the form and, surprisingly, a fair number of students used it to make an impromptu comment about the homework. That was a pleasant surprise.
If you are ambitious, you can try some of the real homework. The first of these in the course, one that doesn't require too much prior knowledge of microeconomics, is the one on Efficiency and Equity Concepts. The one on Math of Risk and Risk Preference is interesting in that each time a question is answered correctly part of the graph is drawn, so this really should help the students visualize what is going on.
The feedback that is given is generated by the IF function, along with conditional formatting. And many of the questions are parameterized in a way that students will see different numbers for the same question, with their alias determining what does show up. This is one step in the direction of preventing the students from cheating on the homework. Of course, a real math nerd could figure out the underlying parameterization and give the answers to other students that way. Fortunately for me, I never had that happen.
One might want to know whether other types of dialogic learning objects might be made that don't rely on Excel but do require AI in their construction. Posing this question is easy. Answering it, however, is beyond my expertise. If it is possible, might the instructor make a few as means of illustration and then have students make others, as their course projects? It is said that you really learn the subject matter deeply when you teach it. Might that also be true in designing learning objects to illustrate the subject matter?
I have been fascinated with this idea for some time. And then these learning objects that were course projects one semester might become learning tools for other students in future semesters. (As there is a copyright issue here, the students who made the objects as their course projects would need to give their permission for the reuse of these objects.) Alternatively, if a similar course were being taught at some other campus, might the courses share such learning objects?
All of this is meant as background for discussing the issue of instructor time devoted to instruction and how that time should be best spent. In my case, I was retired when I taught the Economics of Organizations Course and only taught that one course in the fall each year. I was still working when I taught that CHP class, but it was comparatively small with only 18 students. The Econ of Organizations class typically had fewer than 30 students, so providing the comments on the student blogs wasn't too much of a time commitment. And the design of learning objects, whether Excel files, or PowerPoints, or videos would typically be done before the course started. What about other instructors, who have larger classes, a full class load, and other obligations that make them far from time abundant? Let me offer a first pass at answering this question.
If you go to the url for the course Website and change the year, say from 2016 to 2015, you will find another course Website similar to but not identical to the one that we've considered. Indeed, you can do this back to 2012 as well as forward to 2017 and 2019 (I didn't teach in 2018). The point is that there were changes in the course from one year to the next, as I liked to experiment to address issues that arose in the teaching, but a large part of the course was done according to a template that was fixed by the second time I offered the class. So, with that, I suggest the following.
Instructors should develop a rolling plan about how to teach their course with a Website that is out in the open, one that balances their idealistic goals for instruction with the realistic limitations placed by the time they actually can devote to the course, recognizing that they will teach the course again, make further changes then, and develop a better overall approach.
Let me wrap up. I think I got carried away in writing this, giving a full spiel about pedagogy even while that wasn't what the title of the post seems to be about. Maybe the post will serve to generate posts by other authors who stick more to the topic. Or maybe some will be inspired to also write about pedagogy out in the open, but come up with many other different ways to proceed. I would be delighted to read such pieces, should they come to light.