Colleges and universities generate vast amounts of data every day, spanning from the research findings of their scholars to the log-in times of their students to learning management systems. But institutions of higher learning have fallen behind businesses and government when it comes to putting this data to use, according to a new study in Science.
According to “Data blind: Universities lag in capturing and exploiting data,” although colleges may have lots of raw information, many are “data poor” in terms of using it to their advantage or “data blind,” reluctant to discuss data governance. These deficiencies represent what the authors called “an invisible tax on an organization’s efficiency.”
The study was based on interviews with 12 university leaders, including provosts, vice presidents for research, librarians, and chief information officers from a variety of types of institutions across the country. The findings were surprising to Dr. Christine L. Borgman, distinguished research professor in information studies at the University of California, Los Angeles, one of the study’s authors.
“We were hoping we would find some really good ideas for how they were tackling larger infrastructure questions around data. And we didn’t really find as much as we expected,” said Borgman. “We did find small-scale innovations, many from student projects or prizes across campus. But nobody really stood up and said, ‘Yes, we have a plan. Here’s five years. Here’s the governance model that works.’”
The study described many uses for the data that universities collect. Provosts could make better decisions if they had better data on faculty research, career interests of prospective students, and research funding trends. Faculty could be freed from the burdens of populating personnel files and reports. And the pandemic could have been different.
“People were drowning,” said Borgman. “They didn’t have good inventories of anything from technology to people to resources.”
There are also impacts for DEI efforts—demographic data about students and faculty could be better used to see whether attempts to help them succeed are working. But the study found that opportunities were often being lost.
One of the major culprits was a silo mentality—an unwillingness to share information between different parts of an organizations. Different parts of universities seek to control their own data. One provost interviewed said that his university had many “data czars and czaresses.” This led to information networks that cannot coordinate with each other, and the duplication of effort.
Dr. Jonathan Gagliardi, vice president of economic mobility and social impact at Northern Arizona University, knows what this is like.
“Much of what the authors found really resonated with my own experiences at a number of institutions,” he said.
Gagliardi has found that silos are often more emblematic of a cultural issue than a technical one.
“It doesn’t matter what solution you get if folks become entrenched,” he said. “If people are concerned that data are going to be used in a punitive way rather than a positive way, new walls can pop up very quickly.”
Getting buy-in is a delicate process.
“You can say you want to integrate data better; that’s cool. But unless you back that up by bringing the campus community around the table and being clear, open, and honest about how you intend to use data, people are going to be skeptical,” said Gagliardi. “It’s an exercise in trust culture and shared governance.”
Universities will also have to manage the privacy concerns that come with a deeper use of data. The study found that issues of privacy, while valid, are often used in a broadly—like a talisman, as the authors put it—to short-circuit conversations about data sharing.
“Folks are right to point out the real concerns that exist [about] data privacy,” said Gagliardi. “But that should not be a full stop. There’s definitely risk associated with data privacy. On the other hand, there’s also lost opportunity in terms of being able to leverage data effectively. There’s the opportunity cost of not being able to best serve your stakeholders.”
Some experts questioned whether the situation was really as bad as the study describes. Dr. Karen Webber, professor emerita at the Louise McBee Institute of Higher Education at the University of Georgia, pointed out that, ironically, the study lacked data.
“This is a sample of 12,” she said. “A conversation with 12 individuals may not adequately represent everyone.”
She and Dr. Amelia Parnell, vice president for research and policy at the National Association of Student Personnel Administrators, thought that things were headed in the right direction.
“I want to give a little bit more credit to campuses in terms of their progress,” said Parnell. “I see it in the restructuring of divisions and adding new roles. I see it in intentionally trying to configure data systems to connect with each other. I see it in the more elevated role of institutional research.”
These moves are in line with the study’s recommendations, which include that schools should invest in their data management capacity and investing in knowledge infrastructure, the networks of people, artifacts, and institutions that interact with data.
Ultimately, Parnell believes that universities will have to improve because the use of data is only becoming more important.
“We got enrollment challenges that are coming. We have funding challenges. We have this larger debate about the value of the college credential. We have all types of things that are pressing campuses to show with data that students are getting the most valuable skill set,” she said. “As long as we have those types of things, institutions will make their use of data a priority.”
Jon Edelman can be reached at [email protected].