Learning Analytics and the Future of Change in the Classroom
How does a university take the inordinate amount of data it collects and somehow make sense of it to build strategies for driving change in the classroom? And is it worth the investment?
According to Andrea Deau, senior director for higher education programs and partnerships at 1EdTech, while there are significant barriers to implementing learning analytics — including potentially significant resource requirements, as well as cultural issues — the benefits to students outcomes and to the institution itself can be enormous. Plus, as Deau sees it, in the near future, some of the resource barriers may be alleviated with new AI tools potentially taking on a good chunk of the workload.
Deau herself has worked in a variety of capacities in higher ed and now works closely with 1EdTech member institutions — universities and colleges — "to meet the challenges they face in the rapidly growing and evolving digital teaching and learning landscape." She was assistant vice provost for online lifelong learning at the University of Wisconsin-Madison and launched UW's first online undergraduate degree programs, as well as holding other roles in higher ed. 1EdTech is a nonprofit focused on digital transformation in education, including higher ed and K–12.
Deau will be presenting in a higher ed-focused session called "Learning Analytics Is a Journey — We'll Show You the Way" (along with representatives from University of Notre Dame) at the upcoming Tech Tactics in Education conference, being held Nov. 7–9 in Orlando, FL.
Campus Technology had a chance to sit down with Deau in advance of the conference and discuss her insights into developing learning analytics, which she will be discussing in person at Tech Tactics, as well as her outlook for the future of learning analytics.
Campus Technology: Let's start by talking about where we're at with learning analytics in higher education. How widespread is it? How far does it have to go? And how has learning analytics evolved over the years to this point?
Andrea Deau: I can speak to that in the context of our membership, [which is] probably very reflective of where higher education is overall. I would say that it is really on the spectrum, and I would look at it as a maturity model.
By and large, there are only a handful that are really doing expansive efforts with learning analytics at scale. But a lot depends on the resources they have that they can commit to this because it is resource-intensive. And there's a lot of infrastructure to do learning analytics at scale — and staffing requirements.
I would say a lot of our members want to do it. The challenge is, there's a lot of barriers to doing it — the resources, the infrastructure, the buy-in, the change management, and trying to get build capacity. If you're a large institution, you really have to manage that conversation across multiple units on campus. It is an interdisciplinary and cross-departmental collaboration that has to take place to make that happen.
CT: What are the opportunities for the ones who haven't done it yet or may not be using it to the greatest effect?
Deau: Early identification of at-risk students. You know, the small to mid-range universities are really dealing with enrollment challenges and making sure they show value to their student population. So helping them through the process and identifying early interventions are key.
Personalized learning, modifying curriculum, improving the curriculum, optimizing learning resources, and enhancing teaching strategies really across the board. Student success initiatives, identifying trends and patterns and just overall continuous improvement and ROI for the institution.
CT: What do IT leaders need to know about data architecture and governance in the context of learning analytics?
Deau: We're an organization that promotes common frameworks and standards to help improve the experience and the digital learning that happens in a university or college. The key to that is data integration, getting all these ed tech tools to talk to one another, to have a common framework for the data that you collect; to be able to scale it is key. In the end, the institutions themselves need to think about that data privacy and security, meaning the ownership and access along with all of that key infrastructure. But that is going to be so unique to each institution that we try to focus on the things that we can control, which is working across the market with our members to make sure they're getting the data in a common framework that they can really use once they start collecting it. Because as you know, their data can come in a wide variety of formats, and you need to have a common thread that ties them all, which our standards do.
CT: Can you talk about where learning analytics is heading? And also, I'd be really interested to hear your take on how you think AI is going to be impacting learning analytics. Obviously AI has gone through a transformation itself recently.
Deau: Having been an educator — I'm a trained teacher — I've always leveraged technology for helping my students learn. And I see that in the kind of work that I do today. I'm not in the teaching field, but I am definitely in the teaching mindset. Seeing how we can leverage technology for learning has really been a personal passion of mine. So personalized learning at scale is, I think, an area where AI is going to contribute greatly. And that is something that our organization thinks about. In fact, one of our leadership imperatives is personalized learning, and thinking about how we can make learning accessible for every individual. That's one key.
Real-time analytics for getting that data back in real time has been challenging because of the infrastructure, but I think AI is going to change some of that and make it a lot easier to collect and analyze that data as it comes in.
I think also in the immersive technologies. This is kind of in the personalized learning thread, but immersive technologies are going to benefit from AI in the analytics element of it, automating assessment and feedback. In fact, there's been some experiments I think, with generative AI as a teaching assistant. I think that's kind of fascinating. It comes with all sorts of challenges and concerns. But certainly, people are starting to try that right now.
Also in our organization, we think a lot about upskilling and the future of work and making sure that students can represent their learning in meaningful ways and skills and competencies. And we think that AI will likely play a significant role in helping learners, whether they're in a higher ed setting, or just continuously learning new skills and adapting to the workforce.
But then, you know, you talk about all those really wonderful things. And then you get to the sort of the cautionary tale in there. And that's ethics in AI and really ensuring that we're understanding what's underneath those algorithms and addressing bias and fairness in the use of these because that can impact different populations in significant ways. And so we need to do our best ethically in the use of these technologies.
That also goes for data privacy and security. Those are going to be key in the future and where it's headed.
Also, kind of under the under the covers of it all is that integration with digital learning environments. I think AI is going to help us connect these environments. Even today, with all the work that we do as a community to try to promote a connected digital learning environment, sort of a seamless experience so learners can learn, it's still kind of clunky. I mean, not because of the standards that we put in place that are really trying to get it to work together. But there's a lot of effort in going into making that happen and getting people to adopt and getting suppliers to adhere to the standards. I think the key is, though, AI will help us make those integrations more seamless in a way that learners can just learn and faculty can just teach, and they're not going to have to navigate all those technological complexities as they once did and may even end up as you've got your own personal assistant that helps you with some of those menial tasks of the work that you do, freeing you up for more of those higher order thinking parts of your role.
One other area I've heard one of our members talk about: With AI, ChatGPT goes out and crawls the entire web. And some of that content is dubious. I can do the same search on ChatGPT multiple times and get different answers. And it will cite articles that don't even exist or that are not really academic articles. And so some of our members have talked about a future where there are walled gardens of content, basically, that you can leverage with a ChatGPT or other generative AI. So you're basically walling off the content so that you can have trust in the results that you get. So I think it'll be interesting.
But it's exploding, as you know, and I do think that AI is going to help our members accelerate their work with analytics because there are going to be new tools available and new capabilities, so that with the budget limitations they may have, this type of initiative really allows them to move it forward in significant ways without having to tap enormous resources. So we'll see, but I think that that efficiency will play a role.
CT: What would you like to leave us with about 1EdTech?
Deau: I think one of the things that I really value about the work that we're doing is the community. While Notre Dame and I are going to be at Tech Tactics, I think you should be aware that there are many, many other members that work diligently to further this work. Our organization is really about working together to solve problems for members. And in the analytics space, that's a big area of our initiatives. (We have four key initiatives, and analytics is one of them.) So we know our members want data; we want them to be able to make use of the data so that they can really get into solving problems for their students. So we work together to do that.