Mining Student Data
Balancing Research and Privacy
Technology can improve lives, but it creates new challenges, too. So it is with smart tutoring systems and other powerful new educational tools.
By recording every keystroke, these instruments show us, in real time, how a student grapples with a math problem, applies a scientific concept or constructs a sentence in a second language, revealing where understanding may break down. That data, along with information about grades, attendance and behavior can be tracked throughout a student’s school career. It can be combined with data about other students, helping principals and superintendents weigh the effectiveness of different programs.
And with today’s enhanced computing power, it can reveal patterns and trends that may not be apparent to the naked eye. Is a policy hurting a particular minority group? Do a program’s benefits persist beyond a few months? Is a strategy cost-effective?
"We have an unprecedented opportunity to test theories of learning and give parents and policymakers what they have rightly been clamoring for: evidence that methods work." - Susan Fuhrman
In short, we have an unprecedented opportunity to test theories of learning and give parents and policymakers what they have rightly been clamoring for: evidence that methods work.
The new technologies do pose risks. Personal data could be misused by marketers, students could be vulnerable to identity theft, and learning disabilities or past disciplinary issues could penalize those applying for college or a job.
Recently, a coalition called Student Data Principles has proposed some excellent guidelines. Students’ personal information must be shared with service providers for legitimate educational purposes only. School systems must create policies for overseeing this process. Schools and their contracted service providers — including researchers — must establish clear, publicly available rules for data collection, use, protection and destruction.
Yet researchers must add their voices to this discussion. To that end, the National Academy of Education (NAEd), of which I am immediate past president, is exploring additional questions. Would requiring eventual data deletion undermine longitudinal analyses? Who should own and control access to data — and how to link key databases if individual identities are protected
Ultimately, we all want the best for schools, teachers and students, and we all want to ensure student safety and privacy. With intelligent, inclusive discussion, everyone can win.
BY SUSAN FUHRMAN
Published Wednesday, Nov 4, 2015