The Age of Data-Driven Assessment

Seems like we should know more about knowing more.

GUEST COLUMN | by Troy Williams

CREDIT Macmillan New VenturesIn today’s society, a strong focus is placed on getting a good education. We are taught from a young age that working hard and doing well in school adds considerable value to our lives. Parents know the paramount importance of this to ensure positive outcomes for their children, and thousands of families move each year simply to get their kids into better schools. Education is, in short, one of the most widely recognized of social goods.

The line between online and offline learning is blurring, resulting in a need for more data-driven tools in the classroom.

Given that education has been a high priority for centuries, and that just about everyone recognizes its value, why don’t we know more about the best way to educate a particular child at a particular point in time? Despite widespread and earnest efforts—along with billions of dollars of investment in education technology—the typical student is not learning much more or much faster than a few decades ago.

The line between online and offline learning is blurring, resulting in a need for more data-driven tools in the classroom. These tools are in their adolescence and have become the topic of healthy debate. Are data-driven tools necessary, effective, or fair? And how can instructors maximize learning outcomes with these tools? That’s what we need to find out in order to give the next generation of learners their best shot at success.

The way students are taught hasn’t changed much over the years, despite impressive advances in technology. Students proceed swiftly from one grade to the next each year, regardless of what they actually know or may have forgotten over the summer. Summer learning loss needs to be considered in our education system—the long break interrupts the rhythm of instruction, and requires a significant review of the material from the previous year when students return to school in the fall (Cooper, 2003). Today’s students can easily cram for exams and pass their courses, but often fail to master the concepts or retain the material as they advance. The concept of the big test is antiquated—we need to be assessing students at every moment in real-time in the background and adjusting learning through the process.

To combat this issue, educators and entrepreneurs are working together to create data-driven learning environments that can assess each student after every activity and compare the student’s current performance to their past performance. This technology offers the prospect of individual learning without placing new burdens on teachers. If the student demonstrates that they have mastered the concept at hand, that student can be pressed onward to tackle the next concept in the subject they are learning. If they have not yet mastered the concept, they can be provided with another activity that goes over the material again in a different way, and allowed to move on only after have proven comprehension. If, after repeated attempts, they cannot seem to make progress, a teacher can intervene to help the student over the hurdle. By using data-driven assessment in the classroom, instructors can accommodate the needs of each individual student with the assistance of technology.

In data-driven education, each teacher, administrator, and parent would have a dashboard of data for each student that provides an accurate view of the student’s knowledge at that moment in time. More importantly, it will also show the gaps in that student’s knowledge and compare them to their class and the national average. This system will also generate reams of data on which learning techniques are most effective for what kind of students with which types of learning style.

Ultimately, a data-driven educational system would emphasize competency rather than time spent in a classroom. A student would waste less time on concepts that they have clearly mastered, and spend more time working on concepts that are within or slightly beyond their current capabilities. Students who fail to achieve competency in an area will have an accurate assessment, which instructors can use to resolve the issue.

Technology has transformed just about every aspect of our economy, but so far it has had virtually no impact on education. As data-driven education technology is refined and utilized over the next few years, we expect to see dramatic progress in learning outcomes. If that happens, students will be able to move through the educational system more quickly while gaining a better understanding of the materials covered. However, these assessments can only produce their intended effects if supported by a willingness to change instructional practices in the face of new information (Halverson, 2007). Given the importance of education and track record of technology in other sectors, it is certainly worth a try.


Halverson, Richard; Grigg, Jeffery; Prichett, Reid & Thomas, Chris (2007). The New Instructional Leadership: Creating Data-Driven Instructional Systems in School. Journal of School Leadership, Vol 17-N2, p159–194.

Harris, Cooper (2003). Summer Learning Loss: The Problem and Some Solutions. Retrieved July 17, 2015 from

Troy Williams is President and General Manager of Macmillan New Ventures, where he is responsible for identifying emerging technologies and trends that will have a major impact on student performance and outcomes. Follow @MacmillanNV


    Leave a Comment

    %d bloggers like this: