Using Data and Analytics for Student Success

Gerry Bayne: Welcome to EDUCAUSE Exchange, where we focus on a single question from the higher ed IT community and hear advice, anecdotes, best practices, and more.

If you look at the EDUCAUSE 2020 top 10 IT issues, number five is student-centric higher education, and issue number six is student retention and completion. And one of the tools that technology is providing on the path to student success is data. Data is allowing institutions to better understand students, rethink systems, and create predictive analytics and early alert mechanisms to help students complete their degree. But this use of data has its own challenges for instructors and IT departments. In this episode of EDUCAUSE Exchange, we ask, what is your best advice for institutions trying to harness data for student success?

Connie Johnson: Pick your data. Once you get into the data landmines, you can live in data.

Gerry Bayne: That's Connie Johnson, provost and chief academic officer for Colorado Technical University. She says finding actionable data is the most important way to make positive change for students.

Connie Johnson: And what we try to do is look at what's the most important data. So for example, DFW rates, persistence into the next course, what assignments are students getting hung up on? And we use specific targeted data for action. And that would be my other advice. If the data doesn't lead to analysis and action, then you may not need to review the data. Because when you start working with dashboards and different kinds of data, you can stay there. And my recommendation is make sure that you have a decision point that you take that data and you act upon the data.

John Campbell: Yesterday, I was in a discussion with an institution that spent 18 months building this dataset, asking questions after questions after questions. And then in the end, they had not even gotten to the point where they have taken actions.

Gerry Bayne: John Campbell, vice-provost at West Virginia University echos Johnson's concern for finding an actionable purpose in the data.

John Campbell: You get so focused on the data. Particularly if you get data junkies in the mix, they become enthralled with the data and all the possible questions. And so at times, you can end up far down the road and actually not accomplish anything. And so that's why I keep focusing on what actions do you want to take? What do you want to learn so you can take an action? That's where we have to continue to look at.

Gerry Bayne: Another important aspect of using data for student success is ownership, according to Stacy Morrone, who serves as associate vice president for learning technologies and dean for IT at Indiana University, Purdue University, Indianapolis.

Stacy Morrone: I believe that harnessing data for student success, as institutions of higher education, we have to first and foremost own that data. So any vendor that says, "Well, this is our data, and you can have access to it, but it's really ours, and we might charge you to get access to that data," for me, that's the end of that discussion. It really has to be that our trusted vendor partners are ones in which they are saying, "This is your data and we're not going to charge you for this data. This is yours to do with it what you will." So things like the Unizin Data platform provide this opportunity for data from our partners to be put into a place where researchers can look at this data and say what does this really mean?

Gerry Bayne: Formed in July of 2014, Unizin is a consortium of like-minded universities interested in shaping the digital teaching and learning landscape. The Unizin data platform offers data products and services for learning analytics, application development, research, and business intelligence. And by operating on common platforms, Unizin members can leverage analytics of learner outcomes, both within and across institutions.

Stacy Morrone: So we've got mountains of data, but we first have to own it and make sure that it's ours, and then think about thoughtful questions that will ultimately help us evolve in higher education.

Brian Davis: My advice for institutions trying to harness data for student success is to build a dedicated team to reach out to those students. Start in their freshmen semester.

Gerry Bayne: That's Brian Davis, director of information and learning technologies at South Carolina Technical College System. He thinks the use of data should lead to a deeper high touch relationship with students throughout their career.

Brian Davis: Give them the skills that they need to be successful, help them transition from that high school, possibly to a college environment, and then to ensure that you follow up, make sure that you have both academic and freshmen advisors to stay in touch with those students, to keep up with their current progress, and to make sure that you award them for their achievements, you reach out, congratulate them, stay in contact and make sure that they're succeeding as they should.

Linda Jorn: I think a lot of institutions and ourself, initially, we focused kind of on the technology, but it's very important to support concurrent processes. So when you're initiating a learning analytics initiative on a campus, you need to address technical issues, you need to address policy issues, cultural issues. And for us, cultural issues, we mean like professional development for faculty, and then also workflow issues.

Gerry Bayne: Linda Jorn is retired from the University of Wisconsin-Madison, where she served as associate vice-provost for learning technologies and the director of academic technology. She says that planning and getting teams, processes and standards all on the same page is essential to using data for student success.

Linda Jorn: So if we're going to pull data from many different systems like student information systems, Canvas, we need to have common data standards, common data dictionary, integration standards, such as IMS is helping us with. So it's really important to work on all of those things at once, because if you get a tool, but you don't have the policies in place, questions about ethical use of policies will come up.

And it's really critical that you engage the campus leaders that need to be engaged in a learning analytics initiative. And for us, that meant including our institutional research leader, our registrar, our campus learning assessment director, our chief data office, our advising director. And that's really a key part to make this sustainable and scalable, is that the leaders who may have data in that area, they may be the data stewards for data you need, they may be directors that are working with key constituents on campus, such as advisors. So it's all about concurrent processes and working on those parallel and together.

Gerry Bayne: Jorn says that using data for student success is an ever-evolving endeavor and it can hinge on the reliability of the data.

Linda Jorn: Right now, the way we're getting data out of systems, we can trust some of the data more than others. So as that evolves and we trust the data more, I mean, I think it's like any other innovation. We don't know quite yet how we'll use it. We think we know, we have certain questions we're asking, but I think it will be exciting to see how once we get this data into faculty's hands and students' hands.

Gerry Bayne: Speaking of trust in data, when you're navigating the spectrum of data types available when using it for a question or application for student success, the attitude towards governance can vary pretty widely.

Kathe Pelletier: Some people are like, "Oh, thank God we have just got to get a structure. We have to know who owns the data. We have to understand where it lives, how we make decisions, and yes, we just need to figure out a system for governance." And then other people are like, "Oh, governance. Man, that's really getting in my way. I just want to ask a question and answer it using data."

Gerry Bayne: Kathy Pelletier is director of the Teaching and Learning Program here at EDUCAUSE.

Kathe Pelletier: One term that I love is dirty data, knowing when to really roll up our sleeves, look in that dirty data and know that it's not scrubbed yet, we haven't confirmed that it tells us what we think it tells us, it'll take time to do that and to validate the quality of the data, but it might be some data that we could use immediately to help a student right now. And as long as we're not doing harm, having access to and leveraging that dirty data can really improve student success.

On the other hand, we don't necessarily want to use dirty data when we're reporting to IPEDS, or when we are looking at our institutional KPIs, the kinds of things that the institutional research department would want to make sure that we've looked at and scrubbed and looked at all sorts of different ways. And there's a spectrum in there where knowing the amount of structure to apply to the decision-making, to the ownership, to the access, depending on the type of data that's being accessed and the use that it's being used for, I think that that can help to solve a lot of those kinds of tensions on campus where there is a desire to use data to make decisions, but you want to make sure that you're applying the right data to the right type of decision with the right people at the table to make that decision.

Gerry Bayne: If you'd like to read more discussion by higher ed IT leaders about data and student success, check out the EDUCAUSE review article entitled Student Success Perspectives on the EDUCAUSE 2020 Top 10 it Issues. You can find that article at er.educause.edu/studentsuccessperspectives. I'm Gerry Bayne for EDUCAUSE. Thanks for listening.