What Big Data Means for K–12

Want to see what K–12 Big Data looks like? A good place to start might be North Carolina's Charlotte-Mecklenburg Schools system and its 141,000 students.

Multiply that figure by the ­number of data points about each of those students that the district captures — names, addresses, ­sibling relationships, ethnicity, ­attendance, disciplinary issues, and performance on ongoing formative assessments and high-stakes tests, among others — and it's easy to see how processing, integrating and analyzing so much information could become challenging.

"CMS has always had Big Data sets because we have student, staff, transportation and food service data for thousands of youngsters," says Dr. Valerie Truesdale, chief information and transformation officer for the 159-school district, one of the 20 largest in the country in terms of enrollment.

And those data sets are getting bigger all the time. For instance, the upcoming school year will bring even more student data to CMS in the form of item and performance task test results from the Smarter Balanced Assessment Consortium, a state-led group that's developing assessments aligned to the Common Core State Standards.

The goal for CMS and other ­districts isn't simply to build a ­goldmine of data, but to mine that data for golden insights that will help tailor learning to individual ­students. "To individualize learning is really to customize learning for each child so he or she has a unique experience that [suits his or her ­precise] learning needs," Truesdale says. "Technology increases the ­possibilities for customization."

Indeed, data-driven decision-making to inform instruction "is the heart and soul of what we're about," says Keith Krueger, CEO of the Consortium for School Networking. "The more feedback we can provide to the students, the better we can individualize learning and move ­people at their own pace."

Front to Back

Of course, it's important to do the proper legwork so that all the data that's being accumulated can be put to good analytical use. Doing so requires strategizing around design, data collection, presentation and governance, Truesdale says. "One of the big keys for us with large data sets is how to design a storage system that's robust enough to handle the volume," she explains.

Today, CMS relies on storage area network, local server and hosted storage to hold its student and ­business systems data. But as North Carolina school districts, including CMS, transition from the state-developed Window of Information on Student Education (NCWISE) system to a vendor-provided one, they — and the North Carolina Department of Public Instruction, which hosts some of their data — are analyzing opportunities around cloud-based storage.

Collecting data in a way that will provide methods for analysis with both short- and long-term ­outcomes, by capturing upfront the key data fields or elements that are required for ­educational needs, must be part of the plan, Truesdale continues. "If you have a large data set and want to know how many seventh-grade, ­previously retained Latino boys did well in math, what codes do you need on the front end so you can extract that on the back end?" she asks.

Linking disparate data sets so that decision-makers can sort, analyze and view information around their specific data queries, rather than just be handed pre-defined reports of charts and tables, is critical. So, too, is establishing protocols to protect and secure the data for viewing only by authorized parties — in other words, determining "who gets to read what and why," Truesdale adds.

In fact, student data privacy and security have been getting more ­attention with the debut of Big Data–based services that promise to help schools support personalized learning. Such services need student information in order to build the ­databases that will drive the creation of new educational products.

But many educators are cautious about wading into such pools.

"We are highly proprietary about student data," says Linda Mariotti, assistant superintendent of teaching and learning services for Salt Lake City's Granite School District, the third-largest district in Utah with 90 schools and nearly 68,000 students. "We own it, we keep it, and anyone we partner with doesn't share it."

Whether Big Data startups for the K–12 market prove helpful remains to be seen, says Sue Derison, director of information systems and support for Forsyth County Schools in Georgia. But she is certain that the work the 36-school district is doing to develop a cloud-based adaptive learning platform based on data that students generate in their online and offline work will ultimately help them gain mastery of Common Core standards.

"The important thing with the data as we see it is this: How does it improve instruction in the classroom?" Derison says. "The trick is to be able to combine what I call 'autopsy data' of what has happened with the child, with what goes on currently in class, with formative and summative evaluations on an ongoing basis."

In addition to longitudinal data, scores from online work or assess­ments scanned in from offline work go ­immediately into the platform, and its predictive ­analytics engines go to work to ­develop recommendations to help the student get up to speed. "This helps in a lot of ways," Derison says. "Teachers don't continue to teach things their students already know. It gives just-in-time feedback of what to pay attention to now, ­before students get so far behind that they can't catch up."

The platform, dubbed EngageME – P.L.E.A.S.E., will roll out to the ­district's 39,000 students in August. Initially, only the eight schools that have been testing the system will have access to its recommendation engine.

One thing is certain: As education becomes more Big Data–driven, ­educators and IT leaders must ­remember that human judgment matters too. "You have to pay attention to whether the data resonates with what the teachers know to be true about a student's performance," Truesdale cautions. "There's no ­substitute for authentic analysis."