Pop quiz: you’re a student in a classroom of 30 students, and you’re stuck on a math concept. What’s worse, you just arrived at this school mid-year and you’re actually two grade levels behind in math.
Which approach will help you figure out your problem the fastest?
- Wait for your teacher to come around to help you;
- Use a computer-driven instructional system that adjusts difficulty according to your performance and shows you videos explaining the concept;
- Ask your personal math coach.
For too long, we’ve assumed the answer was either A or B - wait for the overloaded teacher or rely on the (elusive) adaptive platform. But what if we combined the best of both and brought live online coaches into the classroom?
Online coaches can be as proactive and responsive as teachers, and far more flexible and effective than expensive technology systems. The key: empowering coaches with technology and data so that they can point students in the right direction. Without technology, this kind of warm but precise intervention is almost impossible to achieve without spending thousands of dollars per student; with diagnostic technology and online coaches, we believe the price can be as low as $1000 a year for an entire classroom.
Tutoring: Effective But Costly
Tutors are incredibly effective, as Dave Frey of Skooli and Arthur Vander Veen of Compass Learning have laid out here. They and others point to landmark research conducted in the 1980s by Chicago education professor Benjamin Bloom, which found that the average tutored student performed 98 percent -- or two standard deviations -- better than traditionally taught students.
Since then, Bloom and others have wondered how to achieve these benefits without the prohibitive cost -- the so-called “2 Sigma Problem.” But we believe many of these attempts are missing the mark.
Frey posits that such success will come from online tutoring services’ ability to vet effective tutors and connect them with students. (Venture capitalistsseem to agree.) However, it’s not the availability or even the quality of tutors that limit student access and progress: it’s the challenge of connecting coaches directly to students at the moment they most need help.
Meanwhile, Vander Veen and others put forth adaptive content, digital instruction, and intelligent tutoring systems as the linchpins to learning gains. But in our experience, these technologies are useful only in the hands of the right teacher. On their own they are rarely any more effective than textbooks or worksheets -- particularly among students who are the furthest behind.
One key reason it’s so much more more effective to bring together tutors and technology is that students’ misconceptions are complicated to diagnose without technology, and yet almost impossible to correct with technology alone. It’s a far better use of time for coaches to let the software do the diagnostic work and place students on the right standard, and to focus on helping students identify and correct their misconceptions. Real live people working this way are ten times as effective as the best computer program.
Marrying Coaching and Algorithms To Solve Bloom’s 2 Sigma Problem
In other words, humans are vital, but they need both information and the ability to connect with students most in need of help. Moreover, these coaches work best when they’re integrated right into the flow of everyday instruction.
In schools working with Zeal’s algorithmic and live coaching combination, teachers set concrete math learning goals, provide the initial instruction, then assign students to work through practice problems. Our algorithm finds the topics students are ready to learn. Meanwhile, Zeal coaches – all of whom are former math teachers and tutors – monitor student progress. When coaches see a student struggling, they initiate a tutoring session that includes a shared whiteboard and audio connection, right there in class.
And unlike most offline tutors and even online tutoring services, we work closely with classroom teachers. In addition to helping students and capturing data, our coaches send teachers follow-up notes highlighting areas where individual students or groups could use more focus. Often, these are pre-skills that precede the skill or goal the teacher wants to work on -- resulting from several years’ worth of gaps in instruction and learning that the teacher and student might never have otherwise noticed, let alone overcome.
A Math Coach For Every Student
To make this accessible and affordable, technology is used to the best of its ability: automating the diagnostic placement work, not the instruction. When the software figures out where students are falling behind, and then places students on the right standard, coaches can then step in for short 3-5 minute sessions and make rapid progress.
Indeed, people matter much more than we have been led to believe, even in the age of smart software systems. “Using good software, humans begin to chip away at the big issues of the day," says Ben Stern, an education account manager at Dropbox. "Software might free up time to focus on the issues, surface data that elucidates the issue, and empower teachers and learners to grow in new ways. But humans do the important work.”