K–12 Interoperability Moves from Reports to Insights That Inform Instruction

Data interoperability should be invisible. Perhaps we’ve grown too accustomed to paying for purchases with a mobile app, or booking reservations online, or using Google to plan our daily commutes. These conveniences, which we often take for granted, require data to pass through different systems and servers.

In education, however, data interoperability between applications is sometimes missing. Data about a student’s background, their benchmark test results, or their school schedule sometimes remain fragmented across different systems. Yet as teachers and students interact with a growing collection of digital tools and services, making data systems “talk” to each other has become a priority.

The demand wasn’t always there. Shortly after the turn of the century, efforts to support data interoperability in K–12 technology systems were driven largely by compliance. Schools and districts needed a way to standardize data about their students’ backgrounds and achievement levels to meet state and federal reporting requirements. This was one of the needs behind the development of Ed-Fi, a set of education data tools and standards, in 2011 with support from the Dell Foundation.

As today’s education technology tools are used more frequently to drive instruction, expectations have grown for how data can support teaching and learning. Digital content and curriculums capture a richer set of data that can help teachers see how students interact with content and grasp key concepts, and whether they need additional support.

“Giving every child a rich individual experience is the ultimate goal of data interoperability,” said Troy Wheeler, president of the Ed-Fi Alliance. “To do that today, we need to equip schools with sophisticated systems, taxonomies for data, and common standards.”

Reports to Insights

As the demand for interoperability increases among schools, some companies are taking notice and working with their product development teams to make change — but the actions are small and the changes come slowly. How that information from edtech tools is stored, exported, and made accessible to educators remains inconsistent. Many schools still manually export and import spreadsheets to move data between systems. Some companies share data through secure file transfer protocol — but only if their customers ask. Others will build their own APIs to communicate with school systems.

One of the most visible efforts around K–12 interoperability so far may seem wholly mundane: rostering, or the ability for schools and vendors to create and manage student accounts and login credentials for different software programs. Yet as efforts from companies like Clever have effectively illustrated, it is an often ignored but critical step to letting teachers and students use online instructional tools. (You can’t use online tools if you don’t have an account or can’t remember your password.)

Solving the rostering problem just scratches the surface of how interoperable data can support learning, said Wheeler. “If all we can do for teachers and students is ensure they have single sign-on capability and one password, that’s kind of a travesty. We’re probably doing a disservice by thinking of interoperability as a convenience and not a necessity.”

Efforts around interoperability “need to move from one of data transport, like, ‘How do I exchange information between different platforms?’ to contextualizing data, which is what really matters,” Wheeler continued. He noted that many vendors have created dashboards to display data from their own tools. But educators should also be able to reference information from other products. Can a student’s activity on Khan Academy offer any clues into why he or she did not do well on the problem-solving section of an online benchmark math assessment? Is the student an English language learner who does best with visual and verbal cues not included on the assessment?

Contextual information about a student already exists in the student information systems (SIS) used by districts. Knowing whether a student is an English language learner or has been chronically absent from school can give educators insights into why that student might be struggling. However, pulling and analyzing the data, and combining it with instructional and assessment data, can be wickedly complicated. Paul Smith, a former product and marketing manager for PowerSchool, has detailed extensively all the functions SISs perform that are critical to a school’s day-to-day operation, as well as the diffuse market that numbers more than 100 vendors (some of which exclusively serve specific regions or states).

Despite the complexity, all this data can and must be leveraged to “provide a 360-degree view of the student,” implored Wheeler. That means using all the data at one’s disposal — attendance, behavior, formative and summative assessments, courseware performance, digital projects — to help educators tailor their instruction for each student. But it’s not as simple as pulling the data. To make meaning of data and use it to drive instruction, teachers need to develop skills in data analysis or have someone at their school who can support them — and they need the extra time it takes to actually look at the data.

Vendors agree that interoperability has become a fundamental necessity for how software can support teachers’ instruction. “I think of data interoperability as the skeleton, as the infrastructure required for ideas like ‘adaptive’ or ‘personalized’ learning to become more than just a buzzword or marketing ploy,” explained Julian Miller, co-founder and CEO of Learnmetrics [https://learnmetrics.com], a company that helps schools aggregate data from different tools and translate the information into actionable insights.

Higher Ed Parallels

The desire and efforts to merge functionalities previously siloed across instructional, assessment, and school operational systems have parallels in the higher-ed community. Data, after all, is age-agnostic and can support educators and students through making content more easily shareable, or by leveraging insights from online activities and assessments. Many big ideas (or buzzwords) in education technology — from adaptive learning and personalization to the flipped classroom — transcend K–12 and higher education.

A white paper calling for a Next Generation Digital Learning Environment (NGDLE), published by EDUCAUSE in 2015, proposes a framework for how an ecosystem of digital tools, including the learning management system (LMS), can better support instructional needs, beyond recording grades and sharing syllabuses and links to course materials.

Interoperability is the “linchpin” of NGDLE, wrote the authors, since the “ability to integrate tools and exchange content and learning data enables everything else.” Key to this work is the adoption of common data standards — a daunting technical feat. IMS Global, a nonprofit focused on these efforts, is expanding from its roots in the higher-ed community into K–12 (as well as corporate and and government sectors).

Schools Drive Demand (and Lead the Charge)

Promising efforts are underway in the K–12 community to sync up data between different systems. Digital curriculum developers may use results from assessment providers [https://www.nwea.org/blog/2015/resources-for-putting-map-assessment-data-to-work/] like the Northwest Evaluation Association (NWEA) to determine the appropriate activities for each student. School systems are leading the charge in other cases. InnovateEDU, a Brooklyn-based nonprofit, has developed Cortex, a tool that aims to combine the functionalities of an SIS and LMS to empower teachers to tweak their instruction for students. The tool ingests instructional, assessment, and student information data and lets teachers build customized “playlists” of digital content, activities, and simulations, all of which deliver data back to the system. At Achievement First, one of the first schools to pilot the tool, Cortex allows students to engage with the materials at a different pace and depth depending on their competencies.

Connecting the data pipes to support such interoperability efforts is no simple feat. Technical hiccups abound for both developers and the schools trying new tools. At Providence Public Schools, another district that tried Cortex, human and technical resource constraints delayed the rollout. Cameron Berube, the district’s director of curriculum and instruction, told GettingSmart: “Our technology office hasn’t added staff, nor have we increased the size of our research planning and accountability team.”

Learnmetrics’ Miller believes it’s a positive sign to see schools driving the demand for data interoperability. “A few years ago, when a company said, ‘We don’t share the data across systems,’ sometimes school districts [would] just suck it up. But ‘no’ is no longer an appropriate answer. If products can’t be flexible and meet districts where they are, they won’t survive.”

Tony Wan is managing editor of EdSurge.