Blended learning and the promise of data

The reasons given for the increase in online and blended learning often include improving access to a wide range of courses and high-quality teachers, meeting the needs of a wide range of students, and personalizing learning. Another reason for the move to blended learning, which appears to be gaining momentum, is the recognition that courses and schools with an online component are much better situated to generate, report on, and use student data. The data issue is, of course, closely tied to individualization. Personalization happens because the teacher gets data about student progression and achievement much more often, and data that are higher quality, than the data that would be generated in a traditional classroom. The investments that Khan Academy is making in data generation and back-end systems—building on the ubiquitous videos of Sal Khan explaining seemingly everything—are symptomatic of the emerging focus on data generation. Khan Academy version 1.0 was focused on providing instructional materials that students could access from anywhere if they had an Internet connection. Khan Academy 2.0 is transitioning to creating the data that tracks student progression.

Data generation and usage can work well in places like Khan Academy, closed systems like math programs such as Dreambox and ALEKS, and blended schools that use a single learning management system, such as those run by Connections or K12 Inc. It’s a much harder transition for schools that are using existing student information systems, and that often have to tie those systems to the state data repository. At the state level, the challenge is even harder. While states have invested in data systems in the last decade or so, and those systems are better than ever, the ability of educators and policymakers to use the data in effective ways to improve student learning still seems to be years away.

This last point is made exceptionally well by a November 2012 report from the Data Quality Campaign:

“States are making progress in supporting effective data use, but the hardest work remains. Although states collect quality data and have enacted policy changes, they have not yet focused on meeting people’s needs.

• States have laid the foundation to link P–20/workforce (P–20W) data systems but lack governance structures with the authority necessary to share appropriate and limited critical data. This deficiency impedes their efforts to empower stakeholders with critical information to ensure that students stay on track for success in college and careers.
• States are producing reports and dashboards using longitudinal data but are lagging in ensuring data access by stakeholders such as parents; there is more work to do to meet all stakeholders’ needs.
• States are increasingly providing training to help stakeholders use data but have not done enough to build the capacity of all education stakeholders to effectively use data.”

Finally, and perhaps most importantly: “Building data systems is easier than changing how people value and use education data. Creating a culture that supports data use for continuous improvement takes policymaker leadership.”

The DQC report focuses on the need for leadership among policymakers. We believe that same leadership is necessary at the district and school level as well. The culture of data use has not yet extended to most classrooms. Until that happens, one part of the promise of blended learning will remain unmet.