In my current position at Middlebury, I’m more involved in software purchasing and worrying about how things play out across the institution than I have been in a number of years. Back in my HCPS days, I did a lot of this. At VCU, not so much. Turns out things haven’t changed a ton. Naturally, a small higher education institution is a bit different than a 50,000 student k12 system, but there are still a lot of fundamental issues that are the same.
I find the lines around this kind of work to be blurry by necessity and only growing blurrier. Not only is it hard to separate out what kind of technology might qualify as academic, but creating change in these areas is a mix of cultural change, workflow processes, and an increasingly complex set of security, privacy, and accessibility considerations. All that and we’ve just bought the technology, now we have to think through how we’ll know if it was worth it.
I’ve been talking people to lots of people at Middlebury with various ideas.1 I also wanted to better understand how this works/doesn’t work at other places. As a result, I’ve been reaching out to some people to talk through how they do it at their institutions. I’ll write that stuff up as I get a better understanding, but I thought it’d be useful to write up how I see things at the moment.
What is academic technology?
What makes a piece of technology “academic” as opposed to some other category? Initially, you could ask if it’s used primarily for teaching and learning. That’s a pretty solid start. Some categorization is easy. That piece of software that is used solely by Geography 101 students? That’s academic . . . but it can blur quickly. What happens when it’s used by faculty for research? Does the institution also use it for planning projects? My knee-jerk reactions is to tie academic technologies to learning goals. It’d be nice because it would help create review patterns and better associate the technology with particular departments. It’d also be a pretty big shift for us (maybe for anyone) and it opens a Pandora’s box of additional problems and quite a bit of work.
We also have software that starts as academic but becomes a solution across the institution. Our Canvas install started as purely academic software, but now we use it to meet a variety of institutional needs (student groups, committees, etc.) in ways that complicate things. Currently Zoom is under academic software, but it is also our enterprise phone system and how we do all our online meetings. The fact that some academic technology becomes a broader institutional solution is natural, and probably a good thing, but it makes governance of the software far more complex. I often pose the question to myself “What would happen if all the academic people decided to change or end a particular service? Who would it impact? How big would that impact be?” If I end up with a big list and many of those people are outside the academic realm, the technology is probably enterprise.
I also try to consistently use the term academic technology in an attempt to keep our consideration of interrelated systems open. Software is often involved, but it’s tied closely to hardware in some cases. Think about what is involved in providing streaming in a classroom.2 Zoom (the software) is just the tip of that iceberg. You also have to worry about cameras, possibly microphones, probably room computers, Crestron panels, wifi concerns . . . and the media services team it takes to support all that3 Looking at your Zoom software subscription in isolation will lead to a gross misunderstanding of the cost.4 I don’t mean to say that the network is part of academic technology, but rather that we have to have a better idea how these systems tier and think about the big picture when adding major elements to the global environment. It feels like these discussions need a constant zooming in and out to understand their place in the larger ecosystem.
Saying “software”5 also tends to get people thinking of programs that are installed on computers. Old purchasing models where you bought software and ran it without needing an upgrade for several years are mostly gone. I’m stating the obvious, but it’s much more likely you’ll subscribe to software on a yearly basis and generally the costs will increase a percentage each year. This puts software in the operating budget (which is harder to fund) and those “escalator clauses” result in software budgets that increase every year without fail. The SaaS model also makes tracking usage more complex. It was much easier to see if students were using software installed in computer labs. Now trying to determine usage is much more complex and often reliant on the data provided by each individual vendor. I have one service that emails me an HTML table as an attachment each week.6 This data can end up being inflated in various ways and the vendors are obviously motivated to make usage seem higher. In some cases just activating a 3rd-party-add-on in Canvas creates accounts that make it seem like those students are active users whether they ever actually do anything or not.
Online textbooks and certain digital resources in the library might meet even a fairly restrictive definition of academic technology. This matters because I’m trying to figure out what options we have and what amount of money is being spent. It gets harder and hard to point people towards our resources. It’s increasingly difficult to support them technically or pedagogically. Costs continue to grow alongside general complexity and the attendant privacy/security/accessibility concerns.
What is increasing the drama?
COVID has accelerated what was already happening. It also complicated things by pushing a bunch of one-time money into the system in combination with temporarily free-access to various products. As a result, many institutions are running into scenarios where technology has been built into the curriculum, where it has become an expected option by faculty and students, and where the price of maintaining the status quo is increasing while budgets are being limited or reduced.
You can also see increasingly complex technology being offered as a solution to increasingly complex problems. AI will claim to find students who are cheating. You can use handwriting to annotate digital documents. You can do crazy things in immersive 3d environments. There are 7000 different tools to help you process different types of data. Each of these examples comes with a lot of complexity. Some of that complexity is technical,7 but there are plenty of concerns that fall into the realm of ethics, privacy, equity, and accessibility. These areas are, rightfully, getting increased attention by faculty and students. Hacking attacks on schools are increasing. Additional specific expertise is needed to navigate each of these fields and most places do not have that expertise in the depth and sophistication needed to keep up. Even Educause agrees.
How people buy technology and why it matters
In many places, how academic technology gets purchased further complicates things. Often you have some academic technologies that are purchased through centralized IT or the central learning group. I tend to think of this as the highest level. In some places those purchases are restricted to “enterprise” solutions, but often it’s a mix of things. There may or may not be a pretty well structured review process. In the cases where a review process exists, it comes at some cost in terms of time and people. It comes at a cost in responsiveness. On a positive note, centralized money tends to be recurring unless budgets get cut.
Other departments may also pay for various technologies directly. You see that kind of purchasing with various learning or innovation groups. Libraries often purchase educational technologies. Individual academic departments will also purchase directly for their faculty/students. Review processes tend to get blurrier at this level, particularly at private institutions where rules about who can buy what are pretty loose. These technologies might end up being essential to the curriculum or they might be pilots. There might be money to fund the same thing next year or it might be one-time money. Central IT/Learning may or may not be aware of the purchase.
You also get purchases at even more granular levels. Grants might be used to purchase academic technologies. Individual faculty funds (startup funding, internal awards, etc) might fund things. You also see students being asked to purchase/subscribe as part of individual courses.
This matters because it’s nearly impossible to figure out what your academic technology environment looks like. Risk grows. Complexity grows. Costs grow. It’s hard to see what costs are getting passed to faculty and students. It’s difficult to figure out where you might save money by consolidating purchasing. You may not be taking advantage of better pricing that is available through consortium membership. These funds also have a tendency to go away from year-to-year causing unexpected problems. And almost all of the less centralized purchasing is going to happen without even minimal security, privacy, and accessibility reviews.
At the same time, I wouldn’t want to restrict faculty academic freedom or create processes that are rigorous, but unsupportable.8
Zooming out again
A lot of these issues are unsolvable in any absolute way. I’m just aiming for improvement. This is process work, cultural change, long-term, slow stuff. The goal is less about a solution and more around an increasingly better understanding of what’s going on and how we might focus our resources towards some specific goals. I want to improve our understanding of the academic technology environment as part of the larger institutional environment. I hope to standardize academic enterprise processes in ways that are similar to what we do with institutional enterprise systems, but with academic leadership. To do some of this, we will need to better define our categories. We will need to evaluate our purchasing and review patterns. There’s quite a bit to do just to get started.
All this is pretty much the opposite making a website or helping someone with a workflow. There I have an audience I can make happy. I can feel like I did something real and solid. I know from previous experience that I have to keep some balance in these things or I’ll get frustrated and depressed. It’s a difficult balance to strike because the organizational change work will eat all of your time and then some.
Anyway, lots of stuff to think about, but I’ve got a presentation on this to make.
1 Ask them. I’m sure they’re annoyed. Or at least bored.
2 Is this academic or functional? It’s certainly used for teaching purposes, but it’s not going to get tied to any specific learning goals. You do teaching stuff in it some of the time, but also presentations. It feels more like a core service for the enterprise with academic considerations.
3 I’ll avoid, for now, all the challenges behind whether we should provide streaming access to classes and what it takes to do that well.
4 You did figure out your position on video transcription and who pays for it, right?
5 I’ll probably mess up my terminology even in this short post. Language here needs to be precise and defined but it’s hard to get it right and to come to common definitions (and then usage).
6 I set up an automator action that watches a folder on my desktop and then uploads any file I save there to server folder. There a PHP file builds an index to allow people to browse usage data. If time ever allows, I’ll set up a scraper to make a better dashboard. But this is clearly insane.
7 I spent an inordinate amount of time last year troubleshooting written annotation in the Canvas app. I ended up spending a large number of hours documenting the issue and eventually working directly with their developers to even prove there was an issue. We eventually reached a solution but it took quite a bit of technical work to even get the right data to the company.
8 Some signs that the process isn’t working – it takes a very long time, volunteers are performing the work, the work require expertise that the people doing the work don’t have.