Kent Wada and Christine Borgman

UCLA continually produces significant amounts of data about the members of its community, as do third party providers such as publishers and educational partners. These data can be used for purposes as diverse as instructional enhancement, accreditation and governmental review, and increasing the efficiency of daily operations. Yet without careful consideration, data may be misused, causing harm to individuals.

The expanding demand for data about UCLA students, faculty, and staff inspired the creation of a joint Academic Senate–Administration data governance task force. The task force is co-chaired by Kent Wada, UCLA Chief Privacy Officer and Director, Strategic IT Policy and Christine Borgman, Professor and Presidential Chair in Information Studies. It is charged with recommending a campus governance mechanism by which new, potentially risky uses of data about the UCLA community can be holistically considered.

Why Now?

UCLA is the steward of increasing amounts of data about its students, faculty, staff, alumni, patients, research subjects, donors and volunteers. In order to keep a realistic scope, the task force is focused exclusively on the appropriate use of data about faculty, students, and staff.

These data can be aggregated, analyzed and mined in any number of ways: to build comprehensive profiles of individuals, to predict what will happen in the future, or to prescribe courses of action based on these predictions. Such abilities are of interest not only to UCLA, but to many external entities who would like access for purposes ranging from collaborative research to marketing and advertising opportunities.

“When we had paper records, it was easier to control who could get access to data about faculty and students,” Borgman said. “Now we have automated information systems, and we are generating vast profiles on individuals.” In the digital age, the need for a task force to control and restrict access to institutional data became clear.

A Tag-Team Effort

“The reason we have this joint Academic Senate–Administration task force is because you can find very different opinions and sets of ethics on both sides. It is important to get everyone in the same room to come up with something we can all live with,” Borgman said. In their first meeting, the task force articulated many of the data-related issues facing the campus, including information security, privacy, appropriate data use, and third party partnerships.

“We can assume that data always need to be properly secured. We can assume that data about individuals – whether Social Security numbers, information about health, or grades – must be surrounded by the appropriate privacy controls,” Wada explained. “So the task force is focusing on how the campus considers new uses of these data that have potentially negative consequences. We need to ask ourselves,’What benefit is there to our community? What are the risks? Is the use consistent with the values of the institution?’ And even, ‘Is it creepy?’ The question we need to answer is not ‘Can we?’ but ‘Should we?’”

Big Data: Friend or Foe?

Indeed, big data has many potential benefits – it has allowed people to solve difficult problems and identify important patterns that would have otherwise gone undetected. The implications for public health are especially huge, with big data enabling people to quickly locate organ donors, sequence and analyze the human genome, and track global epidemics. Big data can also be harnessed to address challenges like climate change, poverty and crime.

Yet while data mining is producing some amazing results, it also has the potential to cause considerable damage. In 2012, media outlets broke a story that Target had correctly predicted that one of its teen-aged customers was pregnant based on her purchasing patterns and that her father first heard the news from the retail giant. When the story came to light, reaction varied. Some were unsettled not only by how much the company knew, but what they could predict. For others, it merely confirmed this kind of analysis was routinely practiced without people’s knowledge.

Universities Dive into Big Data

Colleges and universities, like private companies, have reason to want to capitalize on the promise of big data. They would like to increase student learning outcomes, improve degree completion times, and reduce student debt load. Analysis of large aggregations of data about students may well help identify ways to achieve these goals. But there is an unresolved issue about benefits and balance with respect to student data when students are not explicitly consenting to being researched in this way – even if it may directly benefit them.

In addition to posing a threat to individual privacy, big data can also be contaminated or misinterpreted, leading to decisions that negatively affect people’s personal lives. For example, using big data, university administrators can easily count the number of publications and citations scholars have and make hiring and firing decisions based on that information. But numbers do not tell the whole story.

“The way those citations are generated varies greatly from field to field. In some sciences and parts of medicine, one might publish 25 six-page papers a year. However, in the humanities, one might publish one book every couple of years,” said Borgman. “And yet administrators love numbers, and once numbers become public, they cannot be pulled back. The only thing you can do is make sure the numbers that are released are accurate and that you have ways to document the information so that it is less likely to be misunderstood.”

Beyond UCLA’s Walls

Many external parties are interested in collaborating with UCLA on “big data” projects. “Private companies are coming to UCLA saying, ‘Let us take your data, mine it for you, and give you research analytics and reports on the strengths and weaknesses of your university.’ UCLA needs to determine what rules are appropriate before we release any of our data,” said Borgman.

“The data we have has tremendous value, though it may not always be quantifiable in dollar amounts,” Wada added. “We’d like to find a way to enclose our data in a ‘bubble wrap’ of our values so that we continue to have a voice in their use and trust that our expectations will be respected. This is most crucial when the missions of two collaborating organizations differ significantly.”

A Pioneer in Information Policy

The charge of the task force is to provide recommendations for campus data governance back to the Executive Vice Chancellor and Provost and the Chair of the Academic Senate. In concrete terms, it will recommend a governance structure that can consider new data use issues and will create a set of principles that can help guide the decision-making process.

“A lot of good work already exists that we hope to build upon,” said Borgman, “such as the UC-wide Privacy and Information Security Initiative and the Code of Fair Information Practices. We have to understand what is needed beyond what is already there that will meet the challenges of the big data world.”

“Many universities are asking what is ethical and appropriate in the online learning space because they know it can be invasive of privacy, but our goal is to generalize the ability to consider these new uses of data beyond learning, It’s a very exciting time, full of new opportunities … and pitfalls,” said Wada.

Borgman added, “In my recent visits to information policy and law centers, both in the U.S. and in Europe, it is apparent that no other university is this far along in the conversation about data governance. We’re sticking our necks out. UCLA is a pioneer in information policy, yet again.”