Brian Greenberg, recently the Chief Academic Officer of Envision Schools, listened to all the hype about blended learning and, like a great edu-preneur, decided to judge for himself.  He took a group of students in grades 9 through 11 who had failed Algebra I and randomly assigned them to two summer school classes, one class exclusively used Khan Academy and the other was a traditional math class.  The same teacher led both classes.  Envision’s staff blogged about their mini-experiment at and we recently caught up with Brian to get his thoughts and reflections.

What were you trying to accomplish with the pilot?  We wanted to test the assumption that blended learning, and specifically Khan Academy with Chromebooks, can do as well or better than traditional instruction.  If we gave teachers this seemingly incredible tool that allowed students to access a lot of content themselves, then what would happen to the teacher’s role?  What would happen to student learning?  It took a leap of faith for all of us, including the teacher, to try it.  But it started to work right away.  We could instantly see the students being successful.  Of course there are limitations and we know that.  The pilot only ran five weeks, the sample size was small, and the pre-and post-assessments were not comprehensive enough to truly compare the gains of the two groups, but we decided to experiment and put it out there in an open source way and have other people learn from it.

How does a traditional school “get started” with Khan Academy?  We talked to the Khan team and the folks who had worked with the Khan pilot in the Los Altos School District, and they were very clear – start at the beginning.  Let kids learn the architecture of the system.  Let them make progress.  Let them experience success and get the rewards quickly – what will happen is they’ll find their way to where they get stuck.  This is exactly what happened.  Most of the students that were having trouble in algebra ran into deep numeracy gaps before the algebra content even began.  So for students to master the material they really needed, it meant letting them work on multiplication, division, fractions, decimals and percentages.

Out of a two-hour block, students would work on Khan for roughly 90 minutes and the remaining time was used for warm-ups, mini-lessons, breaks and wrap-up.  We felt it would be hard to tell students, “We want you to go at your own pace, but only for 15 to 20 minutes each day, and the rest of the time the teacher is still in charge.”  We were much more interested in turning the reins over to students and giving them control.

In order to master a standard on Khan Academy, students have to get a “streak” of ten problems in a row correct and you can earn badges and recognition for your accomplishments in a game-like way.  While we thought the videos where the “a-ha,” it turned out the opportunity for students to go at their own pace, work on differentiated content, and receive immediate feedback was more powerful than the videos.

So how did learning change for students?  It’s really interesting because students feel that math class is now for them instead of being done to them. We recently interviewed a student who is by all accounts a disengaged student.  At the beginning of the summer, she told us how much she hated math and she has repeatedly experienced failure in the subject.  But she’s been really into the computer, the self-pacing, and being able to push herself.  She’ll say, “I’m not leaving until I get this streak of 10 problems right.” But we noticed she was mastering standards without watching videos, so I asked her how she learns to do these problems.  She looks at me whispers, “Don’t tell the teacher, but I ‘take the hint’ before I even start the problems [so I don’t get penalized and have my streak reset].  Once I have the hint, I copy the [algorithm] on a sheet of paper really carefully and I use it to solve all the problems.  But don’t tell my teacher.”  Now this student, who has never been into copying algorithms, is now motivated to watch an example about how to solve a problem, study it carefully, and try to generalize it to other problems.  The best part is she thinks she’s “getting over” on the teacher.  We had to laugh because if we could just teach kids to write down an algorithm, understand how it works and apply those steps, we’d be halfway towards good math instruction.  If the teacher had given her the algorithm or put in on the board, it just wouldn’t have worked for her because she has a different motivation to do math.  Now it’s on her; she feels the responsibility in a way that she doesn’t in a traditional classroom.

And we see these types of behaviors manifesting themselves more broadly as students begin relating differently to math content.  At the beginning of a streak, students will go quickly on the first one or two problems.  But when they get to problems three and four in a streak, you’ll see them slow down.  They want to make sure they are getting it right.  Sometimes they’ll call a teacher to check a problem before they input it, because the cost of losing the streak is a big deal.

One other unexpected outcome is the students help each other a lot, which is easier to do when everyone is working on different problems.  We’ve seen instances where students write on the board, “I’m struggling with X,” or “I’m proficient with Y,” and they pair up and help each other.  The students say it’s really nice when they are all working different problems because there is less competition and stigmatization.

And what was the teacher’s experience?  The teacher is as busy as she is in a regular classroom but she’s doing different things.  She’s doing a huge amount of one-on-one conferencing with kids, which we don’t see a lot of at the high school level.  And I’ve seen her spend more time working through problems in her conferences because she knows there are supports for other students if she’s busy with one kid.  They can take the hint, they can watch the video, or they can pull a classmate aside.

Khan has real-time teacher dashboards which the teacher uses in class.  She can tell if students are on the site, if they are tackling engaging problems, how long they are spending on problems – so from a management point of view she can see what students are doing in that moment.  She’ll say, “OK, this student is red, I need to pop-in and ask why this student is stuck.”

I watched a student who was doing really high level stuff and was clearly a bright math student but had a few big gaps.  And he was stuck on negative integers.  All he needed was for someone to sit with him for three minutes to show him why he was having a problem.  Once he got it – boom!  The dashboard turned green and he could go on to the next challenge.  But under the traditional paradigm, he would not know he had that learning gap and it’s unlikely he’d receive this personalized, just-in-time remediation.  With Khan, the teacher can see these types of gaps on the dashboard and intervene.

She still does direct instruction to introduce big ideas, build culture or create momentum in the classroom, but about 75 percent of the time students are working on their own and she’s either watching the dashboard or doing one-on-one’s with students.

What were your biggest challenges? The first challenge was all the numeracy gaps that Khan exposed – do we plough ahead with the algebra content or do we first give them the foundational skills, knowing that we only have five weeks?  So the teacher identified the most important algebra standards and told the students where they could find that content on Khan.  And she used this strategy to try to accelerate them while they were also filling in their own gaps.

A second challenge was that some students languished on content where they were successful and did not always move to more challenging work.  This is why you need great teachers to help these students set goals and gain momentum – and keep them on track if they are straying.

I think we can also do more with small groupings.  While one-on-one conferencing is great, it’s not sustainable as the only strategy.  I’d like to see more of an elementary approach to our Khan classes where there are different stations and groupings of students working on different activities.

Finally, I’m curious if our students will exhaust the content in a year-long pilot, which is one of the challenges of letting them go at their own pace.

How do you think teaching will change in a blended-learning future?  I’m convinced that if students get a lot more learning in each hour because they are taking control and going at their own pace, then it frees up time for teachers and students to do more application, creative assignments, and strategic interventions.  These are all things that teachers want to do now but their time is dominated by a million other things, even in good schools.

As much as I think Khan Academy is great, I’m not sure I want all students, every day just being exposed to that.  The combination of algorithm plus application is really the magic we want to capture in the blended-learning movement.

Also, we are still learning what to do with real-time student data.  We have all seen its potential but today’s teachers are not trained on what to do with this data and the easiest thing to do is forget that you have it.

How do you see algorithm and application merging together with blended learning?  So my fantasy is having really great projects that are also on Khan or a similar type system, where kids master x, y, and z standards and that unlocks a really great challenge or project for them that requires them to apply those skills.  Or possibly flipping it and saying here’s a really great challenge and to solve this you’re going to need to master these standards; and students are motivated to do this work because they already understand the application that they are trying to solve.  The projects may be digital or hands-on and, either way, students can earn a badge or other recognition for this more challenging work.

I think this is what content 2.0 is going to look like.  The beta version was digitized textbooks.  The 1.0 version is this interactive software that is just now starting to emerge.  And the 2.0 version is this much deeper application of knowledge.

I hear lots of grumbling about education technology.  What’s on your technology wishlist?  My wishlist right now is

  • A platform that can plug in lots of different content streams into one integrated system that feels seamless to teachers and students.  We are going to consume a lot of content and this is a must-have.
  • A common standard for student data so we can merge student reporting from different content vendors.  Today we have to look at a different report from each vendor and it is unclear how to compare them.
  • A 1:1 computing environment in every school, because teachers will never change their practice if they don’t know whether they’ll have computers that day.  We’ve had the experience of Chromebooks turning on in 8 seconds and technology begins to disappear when it is working properly.

So after five weeks – what’s the verdict? Empirically, we can’t really make any definitive statements.  But we are encouraged enough that this teacher is planning a year-long pilot with multiple sections of 9th graders.  The student engagement in the Khan Academy classroom was off the charts and there were real incentives to learn.

I’m convinced that the combination of immediate feedback and differentiated content and pacing is a game-changer and allows students to take more control of their learning.  We had a student who failed algebra three times.  The teacher introduced linear equations and she wanted to see how fast students could get through linear equations 1, 2, 3 and 4.  This student happened to be the first kid on the first day to get to 4 and he stood up and high-fived his neighbor.  He was so pumped that he did it first.  And I’m thinking, this kid has failed algebra three times and he’s this excited about passing linear equations?  There’s something going on here beyond just a computer teaching him.

I start to wonder what happens when a school blends all the subjects.  Would students still be called 9th graders?  Or could we say, a student is at the 9th grade level for English and the 11th grade level for history.  I find myself starting to question the length of class periods, number of kids in the room, the architecture of the building – you begin to understand why there might be better ways to organize schools.

Where does blended learning go from here? Blended learning isn’t a solution where you just plug it in and you make schools great.  But I think it really is a valuable tool for good schools and good educators to get a whole lot better in differentiating and personalizing.

As educators, we need time to figure all this out. We need to experiment, to get it wrong, to iterate, and forge a path to better schools for the entire sector.  We need to do a better job of reaching every student, increasing engagement and ownership, and giving students a chance to learn more efficiently and with more depth than the current system allows.  And I say this as an educator who has put in my years in the current system.  Blended learning is not the silver bullet, but it has the potential to truly transform our schools.  If the best educators jump in and start experimenting in this space, I’m optimistic about the results for our students.

Alex Hernandez is a partner at Charter School Growth Fund, a venture philanthropy that provides growth capital for high-performing charter school networks. He leads CSGF’s “next-generation” learning investments in blended learning programs and is eager to talk to social entrepreneurs who want to re-invent schools. twitter: thinkschools


  • ccinstitutedev