inBloom

inBloom: Welcome to the Big Data Backlash

For those of us who live, eat, and breathe technology, the cloud is something we take for granted. At a high level, the cloud refers to a collection of technologies for deploying applications and storing data across multiple servers and computing appliances. There are public clouds from which businesses can essentially rent capacity and private clouds on which organizations host their own data but leverage the technology and inherent flexibility of a cloud infrastructure. The cloud, whether public or private, has enabled all sorts of innovative applications, lowered computing costs in the Internet Age, and is a proven technology that is absolutely here to stay.

Unfortunately, New York announced on Wednesday that it was pulling out of an ambitious effort to apply cloud-based storage and applications to managing, mining, and using student data to improve educational outcomes. New York had joined a consortium of other states to centralize student data ranging from test scores to attendance in the inBloom data warehouse. inBloom is a non-profit largely funded by the Bill and Melinda Gates Foundation. Parent and lawmaker outrage over the inclusion of identifying information in the database prompted the New York state legislature to prohibit the state Department of Education from giving student information to any entities that “collect and store data for use in a data dashboard or portal”.

Just what was inBloom going to do with all of this data? inBloom provides infrastructure for third-party companies to hook into the data and deliver personalized learning, response to intervention, teacher reporting tools, and more. Thus, had New York or any of its school districts contracted with a learning management company or other educational service provider, that company could have simply developed a plugin to inBloom and delivered their services easily and securely to any state (or districts they contain) in the consortium.

This is where the cloud comes in. Infrastructure like inBloom’s can be deployed in the cloud, accessed securely, and reduced costs for both providers and consumers of educational technology services. The cloud offers incredible economies of scale and, more importantly, the ability to capitalize on the promise of so-called Big Data. Centralizing data in the cloud means standardized fields and tables for which vendors can write one set of code, promoting faster delivery of services and new, innovative approaches previously hampered by the fragmentation of student data within states and across the country.

I’m not unsympathetic to parental privacy concerns. I’m a parent and obviously don’t want my children’s personal information sprayed around the Web. But the people “outraged” by inBloom need a dose of reality and education about the power and possibilities of the cloud instead of the fear, uncertainty, and doubt that detractors used to get this legislation passed in New York. Schools and districts across the country (and around the world) regularly store data with LMS providers, many of whom are using the cloud to host data and their LMS applications. Teachers use countless cloud-based tools like Edmodo and ClassDojo to improve their interactions with students and parents. Web-based tools that students have mastered specific skills and standards and help teachers tailor lessons based on student needs store data in the cloud.

The same parents and lawmakers use the cloud everyday when they buy something on Amazon, trusting that the online retailer won’t share their credit cards, addresses, and phone numbers with the world. And the same approaches to big data analytics that suggest other purchases based on their purchase histories can be turned around to develop powerful new learning applications that leverage student data to genuinely improve learning outcomes at scale.

A final argument that opponents to inBloom and data-driven education bandy about is the idea that we will be using computers and test score results to replace teachers. Unfortunately, the cost of reducing class sizes to 10 or 12 students is out of reach for any but the most elite private schools. New tools that take advantage of extraordinary advances in data analytics, fueled and supported by cloud-based infrastructure are the only way to empower teachers to differentiate instruction in classes of 30 students.

No, inBloom isn’t evil. Looking for ways to use big data and improve our ability to teach our students, identify their strengths and challenges, and give them the best possible educational experience in a flawed and overcrowded system doesn’t sound terribly evil either. Let’s look at the reality of how the cloud and big data pervade our lives and how we can best apply them to rapidly and dramatically improve education.


Picture: “New York – Brooklyn Bridge” by Daxis via Flickr

Christopher Dawson is EDUKWEST's Editor at Large. He is also Senior Contributing Editor at Ziff Davis and The Channel Company.

  • Paul Smith

    Excellent perspective Chris. I I’ve read and reread the NY bill and it seems to boil the ocean. http://goo.gl/HJae19

    Here’s what the legislation specifically bans:

    “SHARED LEARNING INFRASTRUCTURE SERVICE PROVIDER” OR “SLISP” SHALL MEAN ANY ENTITY THAT COLLECTS, STORES, ORGANIZES, OR AGGREGATES STUDENT INFORMATION AND CONTRACTS WITH OR ENTERS INTO AN AGREEMENT WITH THE DEPARTMENT FOR THE PURPOSES OF PROVIDING STUDENT INFORMATION TO A DATA DASHBOARD OPERATOR FOR USE IN A DATA DASHBOARD. PROVIDED THAT THE TERM SLISP SHALL NOT INCLUDE BOARDS OF COOPERATIVE EDUCATIONAL SERVICES OR REGIONAL INFORMATION CENTERS OPERATED BY BOARDS OF COOPERATIVE EDUCATIONAL SERVICES OR OTHER PUBLIC ENTITIES.”

    So, eScholar has been in NY since 2004, aggregating PII at the state level and providing data dashboards. I suppose, that if their dashboards are only available at the local level then it’s okay? But I don’t see how that’s different from what inBloom was trying to do. If inBloom was doing state-level dashboards, couldn’t they just kill that part of the service for NY? Why delete all the data?

    IMO, I don’t see the need to aggregate PII at the state level. However, I do think it would be extremely valuable to answer questions like:
    - What is the chronic absenteeism rate across the state and by district/school?
    - What is the suspension rate rate across the state, by district/school and by demographic subgroup?
    - What is the current final grade percentage for Algebra for all freshman across the state, by district/school and by demographic subgroup?
    - Etc., and how have these numbers changed over the past 3, 5 and 10 years?

    We don’t need any PII to answer questions like these.