Judging by this recent headline in the Washington Post, inBloom is anything but.
Back in March 2013, inBloom had a big coming-out party at SXSWedu, with a posh room filled with suede white couches and promises of a game-changing data warehousing tool for U.S. school districts. The nonprofit had lined up nine state partners and was expected to spend the subsequent year building secure data services while wooing customers and edtech application providers. Optimism for the program was squelched, however, when Stephanie Simons of Reuters dropped an article that raised questions--and stirred concerns--about the non-profit, aptly titled “K-12 student database jazzes tech startups, spooks parents.”
“Spook” turned out to be a gentle way of putting it. InBloom wound up spending much of the past 12 months battling a bonafide mediastorm. Education reform critic Diane Ravitch suggested that the non-profit is engaged in identity theft. Leonie Haimson, a parent advocate of 20+ years and founder of Class Size Matters, spearheaded efforts to get districts to cut ties with inBloom.
The bad press has proven costly. Six of those original nine state partners have cut official ties. Massachusetts is still deliberating over whether to use inBloom’s services, according to inBloom representative Adam Gaber. New York and Illinois (with the exception of Chicago) are moving forward in their partnerships with inBloom.
What gets lost in this he-said/she-said brouhaha is that data-tracking and third-party data collection companies are anything but new. And they won’t cease to exist, regardless of what happens to inBloom.
What is inBloom, and what does it actually do?
Known as the Shared Learning Collaborative back in 2011, inBloom is an Atlanta, GA-based nonprofit that is building a repository for student data. It has been funded predominantly by the Gates Foundation and the Carnegie Corporation. As a nonprofit, inBloom has used this money mainly for building its full-time staff, which currently numbers 32.
According to Chief Product Officer Sharren Bates, inBloom’s mission is to “provide a valuable resource to teachers, students, and families to improve education” by offering a data service for schools districts to collect, manage, and reflect on student data--hopefully informing, and improving, existing instructional practices.
Here’s how inBloom is designed to work, according to Bates and the inBloom website. inBloom provides a Secure Data Service (SDS) for schools and districts to upload whatever student data they collect. Teachers, administrators, students, and parents have varying allowances of access to that data, depending on state and district laws. The primary purpose for providing this SDS to districts is to assist administrators (mostly at the district level, where the majority of public school system procurement of edtech tools takes place) to choose whatever best supports their instructional goals with their students’ data in mind.
The inBloom API helps districts to collect and store student data, which can be used by schools and district internally to track and personalize learning for students. For example, a special education teacher might analyze their students’ data to generate IEPs, or Individualized Education Programs.
Alternatively, districts can use that data to craft products tailored to their district’s needs--typically with the help of outside vendors.
Districts may already work with some content and tool providers or software companies. InBloom has a set of partner providers that can work with districts, including Wowzers, iResult, and CPSI, and offers a list of the kinds of tools and services that districts may choose to build out with student data as a reference point.
Example applications that schools can build include a Learning Map Authoring Tool (where educators can browse through standards like Common Core to develop maps linking back to student performances), and a Student Insights Tool (where teachers can “red flag” specific conditions and track students who meet those criteria). These tools are meant to be starting points, which local districts can further develop to use and apply student data in instructional practices.
Some of this is indeed happening in New York and Illinois, the two states still engaged in the inBloom partnership. inBloom has supported the EngageNY project with its data platform over the last year, and is currently supporting Illinois’s Shared Learning Environment.
However, the fine print is where things get messy. As reported in the Reuters article, the repository can also hold sensitive information such as test scores, attendance numbers--and even addresses and Social Security numbers. (An inBloom rep later made note of an inaccuracy related to SS numbers, saying, "It is our policy to not store social security numbers on our system.")
What’s more, the information can be authorized for third-party company use. According to Bates, “school districts have that responsibility and authority to load information and approve applications, in order for those application providers to use the API.”
In layman’s terms, this means that any district could opt to share that data with commercial vendors, something they might consider if they are looking for content and tool providers to design tailored solutions.
The critics
Cue the ensuing call-outs of potential privacy breaches by parents. According to a recent Common Sense survey, 90% of adults are “concerned about how non-educational interests are able to access and use students' personal information.”
In New York, Haimson rallied parents who shared her frustration at what they felt was a lack of communication from inBloom on how student data would be used. “The idea of parents having any kind of say was way beyond anything they ever thought about. There was no consideration of having parental input or parental consent,” she charged, in an interview with EdSurge.
inBloom’s privacy and security policy statement (which has since been revised) added fuel to the fire by using classic boilerplate language that looked ominous in the context of students: “inBloom, Inc. cannot guarantee the security of the information stored in inBloom or that the information will not be intercepted when it is being transmitted.”
The debates became acrimonious and bitter, with assertions flying fast: “inBloom set out to collect as much personal student information from as many states as possible,” Haimson charges. “They wanted to reach out to vendors with money offers, and try to lure them into building products around highly sensitive data.” (Bates strongly denies this allegation.)
Following blog posts and townhall meetings, states began to back away from inBloom. Louisiana tipped the domino chain on April 19 when State Superintendent John White agreed to pull student data out of inBloom. By August 1, 2013, five of inBloom’s state partnerships were kaput.
But isn’t everybody doing it too?
It’s worth noting that inBloom’s privacy statement is something that many software companies in other industries use. And while inBloom has the potential to interface with millions of student records, it is only one of many companies involved in the practice of collecting and sharing federally-mandated student data.
Bill Fitzgerald, a former educator who has followed inBloom closely, says collecting student data collection isn’t “new by any stretch.” LearnSprout, eSchoolNet, Clever, Ed-Fi, and Pearson are just a few of the companies that collect and analyze data to help schools improve instructional practice. They may not function exactly as inBloom does, but as Fitzgerald describes, you can see the similarities in these products: “Ed-Fi is another data sharing solution, also based on Common Education Data Standards. Like inBloom, Ed-Fi also highlights how their tool supports data collection and sharing with third party vendors,” says Fitzgerald.
Schools have systematically collected data for years. “For 15 years, school districts have implemented learning applications transferring data within FERPA regulations,” says Bates. “In this context, the collection of data is not separate from the continuous improvement of achievement in the classroom.”
And there’s the big bomb--federal law. Following the Department of Education’s FERPA changes in 2012, schools could disclose student directory information. Going even further back in time, No Child Left Behind federally mandated K-12 data collection--specifically, using statewide longitudinal data systems. As Fitzgerald describes it:
“Federal law mandates that schools, districts, and states collect data on individual learners in the name of "accountability;" this ensures the need for data management. If inBloom, Ed-Fi, Pearson, etc were all to disappear tomorrow, there would be another round of datastores popping up because collecting and storing data is--according to federal law--a required element of public education.”
Where inBloom misstepped
So how did inBloom manage to become the face of the student data controversy?
Bates concedes that “this is part a larger conversation with lots of different facets and we didn’t expect for it all to be pointed at us.” At the end of the day, she stands behind inBloom’s mission and denies allegations that the system openly provides student information to third-party providers. “We don’t do any of that. That is not in our mission, that’s not what we do--in fact, we’re forbidden from doing that legally. School districts are responsible for managing disclosures to [companies].”
But she concedes that inBloom didn’t communicate well. The result: parents felt in the dark about where student data could wind up. “I don’t think we took advantage of those customer scenarios that were presented [last March] at SXSWedu,” Bates explains. “If I had to do it over again, I would advocate for a stronger push on customer statements. I wish we could have made clear that this is about school districts solving real problems with teachers, students and families.”
In speaking with Bates, Fitzgerald and Haimson, it’s clear that several factors led to the avalanche of controversy.
1. Communication and lack of transparency
As Bates mentioned, inBloom could have done a much better job at its communication efforts to ensure that all “customers”--districts, schools, teachers and parents--were involved from the beginning.
“InBloom's launch was pretty tough to watch,” Fitzgerald tells EdSurge. “They appeared to focus initially on the appeal to vendors, almost to the exclusion to the benefits to districts, schools, and learners. With the communications around launch pitched to the vendor community, they fed into the narrative that the whole purpose of inBloom was to shovel student data into the hands of corporations. Then, when pushback started, the PR response was muted and ineffective.”
“InBloom allowed their story to be written by others,” he adds, “and arguably they have been playing catch-up ever since.” He also cites the false charge that vendors paid inBloom for use of data as a specific example of something that received a great deal of attention and confused parents, teachers, and advocacy groups in New York City.
“Had inBloom been able to articulate a business plan, it might’ve helped,” Fitzgerald suggests.
2. Timing
Additionally, inBloom launched its delicate data initiative at a particularly busy time.
With the amount of newly-minted initiatives in 2013--Common Core and changes in nationwide testing, in particular--the public was already weary of another large-scale effort that could affect student learning. On January 24, 2014, EdWeek blogger Catherine Gewertz released a post chronicling the efforts of school chiefs in 34 states to resist sharing personally identifiable student data with the federal government.
Couple this with Edward Snowden’s revelations about government snooping in June 2013, which escalated public concerns around data privacy, and inBloom was headed for a collision. Fitzgerald explains:
“The Snowden revelations helped more people become aware of how seemingly innocuous data points could be used to create a pretty complex picture of what a person is doing. More than any event I've seen in the privacy space, Edward Snowden got people talking about data, metadata, collection of information over time, and how that information could be misused. The effects were definitely seen in the education space.”
3. The Gates-Murdoch-Klein conundrum
Fitzgerald also calls out the involvement of Rupert Murdoch, Bill Gates, and Joel Klein as highly combustible tinder for critics.
“When people hear ‘inBloom,’ they hear, “Bill Gates hired Rupert Murdoch to allow Joel Klein to build a system where they can see my child’s data,” explains Fitzgerald, having outlined the involvement of each name in another one of his blogposts. It “hurt inBloom’s chances at a fair hearing,” he says.
Bates stands behind the Gates sponsorship, explaining that “to build a secure, production-ready data service that is ready for hundreds of districts in the country to solve real problems takes tens of millions of dollars.”
Takeaways
InBloom got caught between two powerful but surprisingly divergent societal tides: a top-down, take-control approach and a grassroots personalization movement.
Data can better inform students, parents and educators about the learning process--and it can also help companies build more effective--and profitable--products.
It's important, however, to tread carefully. Individual empowerment and personalization are in vogue. InBloom hopes to deliver on it.
But the top-down, take-control approach that inBloom took raised concerns that data and technology will trump the individual efforts of teachers and schools to innovate and personalize learning for the students--and breach deeply-held notions of privacy and security.
Are we willing to give up some individual power and privacy for the sake of better edtech products--and potentially improved learning outcomes?
That is a question that ultimately the education world as a whole, not just inBloom, needs to answer.
