Is student privacy erased as classrooms turn digital?

Privacy researcher Elana Zeide says schools need more transparency about how student data is used by educational publishers and software companies.

Privacy

When parents and school officials in New Jersey discovered that educational publisher Pearson recently monitored students' Twitter accounts during standardized testing periods, an uproar ensued. Parents were alarmed and one superintendent called it "disturbing."

But Pearson maintained it was doing what was necessary to make sure students weren't cheating on the Common Core test it administers. 

The debate highlights the evolving nature of student privacy in the Digital Age. As schools rely more apps and educational software inside and outside the classroom, massive amounts of data are being collected on students. More often than not, parents are in the dark about how this data is being used – and how their kids are being monitored – by schools and software companies. 

I recently spoke with Elana Zeide, a privacy research fellow at New York University's Information Law Institute, about privacy concerns in education. Edited excerpts follow.

Selinger: Parents have become very upset about how schools are approaching privacy issues. What are their main criticisms?

Zeide: Parents are critical of how much information is being collected on students, how securely this data is shared and stored, and how education records may be used to serve noneducational interests. Parents are concerned that schools and the vendors they’re sharing student data with aren’t adequately protecting potentially sensitive records from hackers and identity thieves.

Another set of concerns involves commercial use of student information by schools or outside entities that provide services like e-mail or run buses and cafeterias. Many parents feel it’s inappropriate to serve students advertising based on user profiles. Others feel that companies who use information generated by students to turn a profit by, say, attracting more advertisers or developing new products are exploiting their children. There is particularly vocal resistance to the idea of private parties selling student information for “commercial” purposes.

Parents even worry about how well meaning actors will use student information. All 50 states now have state longitudinal data systems that store student level information over long periods of time – frequently from early childhood through entry into the workforce. Private companies are also creating profiles that track students over time. The adaptive learning company Knewton, for example, creates knowledge graphs of concepts students have mastered to help teachers track their progress and differentiate instruction accordingly. Parents fear that granular data collection and long-term retention will create the modern day equivalent of the proverbial permanent record, and that inaccurate, outdated, or unrepresentative information may curtail future opportunities. They wonder if a behavioral problem in sixth grade will hurt their child’s chance of getting into a good college.

Selinger: Why do you believe some of the parental criticisms are misguided?

Zeide: Parents often apply the term “student privacy” to a broad array of issues concerning information collection and use in education. Many object to increasing reliance on standardized tests and the implementation of the Common Core. These policies often mean that students take more tests on digital platforms, generating much more information about their performance than previously possible in a world of paper records. It’s understandable, then, that these policies prompt parental concerns about student privacy.

However, the education policy issues of how to teach and assess student performance should be examined separately from the information policy decisions that determine the appropriate ways to store, share, and use information that may be generated by these tests. Even if schools stopped giving these kind of tests tomorrow, they would still collect a considerable amount of potentially sensitive information about students – as would independent education platforms like online tutors. There would still be a need to address information policy issues about how schools, educators, apps, and the companies they rely on to provide services should handle student data. That’s the purview of student privacy.

And yet, in a sense, “student privacy” actually is a misnomer. Student privacy issues are more accurately about educational privacy. Information about students reflects on their teachers, schools, and administrators – all of whom have a stake in student information policies. Still, student privacy has a deeper emotional appeal, and perhaps that’s why it’s the term nearly everybody uses.

Selinger: Recently debates have erupted about companies like Pearson monitoring students' social media accounts to ensure they aren't sharing test questions over the Internet. Is the parental outrage appropriate, given the surveillance activities that are occurring?

Zeide: The Pearson case is interesting because it shows how unexpected information practices can generate strong emotional responses even in the absence of noncompliance, ill intent, and articulable harm. Pearson was monitoring publicly available social media accounts to fulfill its obligation to prevent cheating. Doing so is legal, and might be required under its testing administration contracts. It’s also hard to point to a particular harm students suffered as a result of this “social listening.”

However, news of Pearson’s practices prompted parents’ outrage. My sense is that this emotional response is less about the potential consequences of the company’s test monitoring than the disruption of parents’ expectations about information practices. Parents understand that people might read their child’s tweets, but that casual observation doesn’t seem as intrusive as the idea that a company would monitor and analyze them systematically. Parents are acutely sensitive to the loss of privacy by obscurity. 

The incident highlighted the degree to which students are under surveillance, both within and outside of traditional school environments. Social media is generally seen as a separate domain from school. Twitter seems more like “off campus” speech – so that Pearson’s monitoring is more like spying on students’ conversations in carpools than school hallways.

Selinger: What are the broader implications of the Pearson controversy?

Zeide: This example illustrates how new information practices may put the interests and preferences of parents at odds with the overall interests of the educational system. While Pearson’s monitoring may make parents uncomfortable, it also ensures a fair testing environment that gives all students an equal opportunity to succeed. This tension between the interests of individual students and those of the broader educational system underlies much of today’s student privacy debate.

Selinger: What changes in policy or outlook would you recommend to make things better?

Zeide: We need to address issues about how to handle the student information that has been and is currently being collected, regardless of the underlying education policies. One way to start is with better transparency about the type of information schools collect, who can access it, and what can – and cannot – be done with it. This won’t solve privacy issues, but providing parents with more information about may go a long way in building parents’ trust. Without this openness, it’s easy for rumors and misinformation to proliferate, as happened during the controversy that brought down inBloom, the Gates Foundation-funded data repository.

The conversation also needs to shift from focusing on individual interests and harms to take the broader consequences of information practices into account. Schools and vendors need to stop dismissing parents as Luddites simply because they can’t articulate a specific and immediate harm to their child. Parents, in turn, need to understand that schools can’t defer to all their privacy preferences because there are also collective interests at stake that affect the entire educational system.

Policymakers need to be wary of making already overburdened schools responsible for investigating and overseeing third party practices, which is currently the case under the primary federal student privacy law, the Family Educational Rights and Privacy Act. New state regulations and proposed federal privacy reforms take a better approach by regulating the actors parents are concerned about – online operators – directly.

Evan Selinger is an associate professor of philosophy at Rochester Institute of Technology. Follow him on Twitter @EvanSelinger.