What Will Happen to ‘Big Data’ In Education?
By Anya Kamenetz
Yesterday, a $100 million startup lost its last customer. According to a Politico article, the state of New York, inBloom‘s last remaining client, will delete all student data on the repository due to privacy concerns.
InBloom’s company spokesperson told Politico the nonprofit was “pushing forward with our mission,” though at the moment there are no known state partners.
InBloom’s trajectory has shined a spotlight on the public’s sensitivity around what happens to student data. When it first began as a mammoth ed-tech project in 2011 by the Council of Chief State School Officers, the Bill and Melinda Gates Foundation and the Carnegie Corporation called the Shared Learning Infrastructure, the purpose was to provide open-source software to safely organize, pool, and store student data from multiple states and multiple sources in the cloud. That included everything from demographics to attendance to discipline to grades to the detailed, moment-by-moment, data produced by learning analytics programs like Dreambox and Khan Academy. An API — application programming interface — would allow software developers to connect to that data, creating applications that could, at least in theory, be used by any school in the infrastructure.
In February 2013, just a little over a year ago, SLI relaunched as an independent nonprofit named InBloom. The company had nine state partners, Colorado, Delaware, Georgia, Illinois, Kentucky, Louisiana, Massachusetts, New York and North Carolina, representing 11 million students. At SXSWEdu, they made a splashy public debut the following month, hosting parties and panel discussions as an official sponsor, a gathering focused just as much on business as on education.
“Our purpose is to remove the friction in the deployment of technology in the classroom,” CEO Iwan Streichenberger told me at the conference. “It’s not very exciting, but if you don’t have plumbing you can’t have appliances.”
InBloom’s offering is not uncommon — in fact, similar technologies exist all over the internet, and consumers largely embrace them. For example, an API allows you to sign in and comment on the Washington Post or rent a home on AirBnb using your Facebook profile information. You can sign in to personal finance site Mint.com and see all of your bank account information in one place, or for that matter, buy products on Amazon with one click thanks to back-end security protecting financial transactions.
But student information, like electronic health records, remains much more sensitive than other kinds of consumer information, and the public response since InBloom’s launch has been equivocal at best. In state after state, parents and other education activists raised concerns that student data would be exploited for financial gain or stolen by hackers. In the words of Leonie Haimson of Class Size Matters in New York City, one of inBlooms’ staunchest critics, “For-profit vendors are slavering right now at the prospect of being able to get their hands on this info and market billions of dollars of worth of so-called solutions to our schools.”
THE FUTURE OF STUDENT DATA
So what does the demise of inBloom mean? It doesn’t mean that student data is safe, either from marketers or hackers. According to a recent study by Fordham University Law School, 95 percent of schools and districts already use a hodgepodge of third-party cloud providers for data storage and internal data mining. Fewer than 7 percent of these arrangements actually restrict the sale or marketing of student information by vendors, and parents are generally not informed of how their children’s data is stored or used.
Last November, a $5 million class action suit was filed against the makers of the ACTs and SATs for selling personal information about millions of high school students. And the Universities of Maryland and Indiana have recently suffered well-publicized data breaches exposing student Social Security numbers and other sensitive information.
It also doesn’t mean that the use of student data will come to a standstill. A small startup called Clever, launched in 2012, uses a single freestanding API to connect the vast mishmash of student information systems with third-party applications. It doesn’t store the data itself; it just makes it easier to get the data out of silos where it can be accessed by educational software developers — the exact same purpose announced by InBloom. So far, 10,000 schools have adopted it, with no public outcry.
It does mean that, at least when it comes to student data, when tech businesses roll out product pitches to schools, they need to understand the extent to which the concerns of parents and teachers will affect their business plans.
In our conversation, Streichenberger talked about taking down “barriers to entry” and creating an “attractive market for entrepreneurs.” Silicon Valley may generally take it for granted that what’s good for its business is good for everyone else, but for those wary of the privatization of the public school system, these are fighting words.
The same critics who are skeptical of the Common Core State Standards (also Gates-seeded) are skeptical about inBloom because of the underlying notion of “interoperability” in education practice. Standards and interoperability are core tenets of the web. It’s because of shared communications protocols such as HTTP that anyone can create a web page or application that is usable by virtually anyone else. Similarly, proponents of the Common Core argue that having a shared understanding of priorities will allow all schools to improve by sharing best practices.
The principle of interoperability is now coming into conflict with a longstanding tenet of American public education — that of local control. The ultimate outcome is anyone’s guess, but it’s now clear that the transition to a big-data educational future probably won’t come from a centralized mandate or single coordinated effort — at least not now.