Using Adaptive Technology to Increase Math Test Scores

I am John Danner. I’m the CEO of Rocketship Education, and our mission is to eliminate the achievement gap nationwide in our lifetimes. When we designed Rocketship, we really thought about what does it take for the most out risk children in order to catch up and then get ahead and succeed in college and in work. We found that the vast majority of programs were kind of constructed almost like text books where they were linear, and when we found Dreambox, what was kind of eye opening for us that for whatever reason the folks that designed Dreambox realized this exact problem which was if a child is stuck what you actually do is give them a little test, figure out what they know and what they don’t know, and then go to a lesson related to the thing that they need to learn next. We found that was incredibly unique. The Dreambox had figured this out. Nobody else is doing it that way.

Dreambox figures out where they should be and then it accelerates learning at that point because at the end of the day, it’s how much does that child learn per hour in the learning lab or in the classroom every day. When you net out what is Dreambox and why is it different, the foundation of Dreambox is this adaptive engine that knows what is related to what from a math -pedagogical sense in math so that when a child is somewhere and they’re not getting in it, it knows where to go. Dreambox figured that out early, got it right, and it makes the program much more coherent while it’s adapting like what the program’s doing underneath is figuring out, “Well, this is the most appropriate thing for this child to do right now.” And that’s kind of a magic in Dreambox.

One other things that we decided to do last year was to do a study with SRI to figured out how much learning was actually happening in our learning lab. What we were expecting was that we’d see some optic in the amount of learning the students got by using this 20 hour- what SRI found in fact was that the children that were on Dreambox for 20 hours increase the national percentile rank by 5 percentage points, so if you were at a 50 national percentile rank, you rose to 55 just during those 20 hours of Dreambox instruction. That’s relative to this control group that didn’t make that progress, so we’re sure that it’s based on the instruction that they got on Dreambox. That’s an outstanding result. You know, 5 percentile points is what schools work hard to get in an entire year of classroom instruction, and 20 hours of instruction getting the same result is just phenomenal.