5 steps for using advanced analytics to transform students and schools

Information about your information will illuminate holes in policy as well as practice
Fred Weiss, Othot
Fred Weiss, Othot

It’s an honor to work with institutions of higher education. They have the honor and responsibility of expanding and transferring knowledge and awareness to individuals who are hungry to learn and grow intellectually.

Sometimes, however, I am stymied by our inability to leverage the detailed knowledge about individual students, the challenges they’ve faced on the path to college, and what opportunities lie ahead.

Data exists on just about everything we and our students do, everywhere we are—including in educational settings. Right now, colleges and universities have an ocean of data about student behaviors and circumstances, course interaction, assessment delivery, the efficacy of student support systems—and even more information than is possible to catalog.

I find, however, that our institutions are not cataloging this information or if they are, the information is siloed and unable to be combined with data from across the institution. Not using this information—or in some cases not even being aware of it— is something we need to change.

More from UB: Dept. of Education unveils college COVID-19 Handbook

Data management and the coordination between departments to secure that data to the highest degree is complicated and fragile. Fortune 500 companies struggle with it right alongside institutions. And yet, the information we already have on hand has the power to be truly transformative to our institutions in general and certainly for individual students, too.

As just one example of this, good data practices can help schools intervene early to support students at risk of falling behind or stopping out of college. In nearly every institution of learning, a school counselor can see, for example, that a student is getting progressively lower grades in subsequent assessments or perhaps has stopped visiting the writing center.

These data points come from different departments of the university but when combined, indicates a risk that we can not only pinpoint early, but we can know what types of interventions are most likely to work. Not only can we know which intervention —tutors, remediation, reminders, financial supports, mental health intervention—we can also know how best to deliver those resources—text message, call, personal meetings— because of multiple data points from throughout an institution.

This is not the future of data, this is now. All of this and more is knowable and actionable today, in nearly every school in the country.

We don’t need to collect a single byte more. You can always collect more, but now is not the time. Instead, proactively manage and use what you already have.

To be more helpful, the following are five steps institutions can take to make positive, life-changing decisions with the data resources they already have.

1. Prioritize data integration and awareness

This sounds easy but if the answer is to declare it a priority, nothing will happen. Good data utilization takes an actual commitment that must come from leadership.

Schools that use their data the best have someone in charge of the process. That does not mean they have someone assigned to it; it means that good data use has a champion, an advocate.

The first step is to find the person or people who care about this and empower them—give them the power to set deadlines, create deliverables in other departments. Also, make sure the highest levels of institutional leadership approve and will stand behind the process all the way through.

That may feel challenging but the reality is actually pretty clear—using data to improve efficiency and student outcomes won’t be a luxurious option for some schools. Every school is going to get better at it—the only questions are when and how well. Followed by, how far behind will you be if you wait?

2. Identify the most important intervention questions

If you try to “use data,” you probably won’t do it well. Instead, have the end goal in mind.

Instead, I suggest asking, “What is the most important problem we can solve and where is the data we need?” Or “what’s our institutional priority right now?” Is it retention? Persistence? Recruitment? Changing pedagogy or assessment? Using financial aid more efficiently?

Once you settle on what you’d work on first, then find out what data you have that can inform those decisions and solutions. Once there, build a roadmap to bring that data in and design ways to analyze, understand and use it. That does not mean start small, it means start specific.

3. Get expert help.

Data is complicated and, as mentioned, many large, highly resourced organizations struggle with it. So get help.

This does not necessarily mean hiring an outside expert, though that can be beneficial as they can add insights free from institutional biases and may have experience on the very issue you’re trying to address. It does mean finding people who know how to excavate, standardize and analyze data sets.

Often, experts in this can already be on staff—either teaching or lecturing. Be cautious however about tasking these projects to the IT department. While they certainly need to be engaged early and are generally your best resource for information security, IT departments are often backlogged and overworked and answering daily needs and emergencies.

More from UB: For 12-year-old NASA hopeful, free tuition to ASU Online

You don’t want a project of this size or impact to be simply something else someone—who may have less knowledge of departmental business needs—has to manage. Those who take it on need a bigger picture view of how different departments interact to support individual students.

4. Communication: Talk too much!

Everyone is rightly concerned about their data—who is collecting it and for what? Rather than address these issues while the effort is underway—and concerns and rumors are spreading—design the data project in public, in full view of students and other stakeholders.

If the goal is to design supports to reduce dropouts and academic delays, say so. Be clear about the goal as well as what data may be accessed. Provide a place people may go to ask questions or join the team. Provide timelines and status updates as often as possible.

5. Standardize your data collection and integration.

Inevitably, as you begin to work on a single data project aimed at a specific use, you’ll learn a great deal about the state of your information.

That information about your information will illuminate holes in policy as well as practice. You’ll learn that standardizing information upfront can save countless hours and dollars on the back-end. You’ll learn ways to understand your institution and your learning community that you never imagined.

But like the previous suggestions, don’t plan to “standardize data.” That’s obtuse and too big. My advice is to start on a project and prioritize and address what you find as you progress. Doing so will make the second data intervention easier and faster.

Fred Weiss, the CEO of Othot, is a CASE Laureate with a long history of working with advanced data analytics in higher education.

Most Popular