Check out the exciting innovations for back-to-school. Learn more

Seamless Formative Assessment For Personalized Learning


I’m Cathy Fosnot. I was a professor for many years at City College of New York. I founded mathematics in the City in National Center of in service from math education. I’m the author of various books and articles on mathematics education including the context for learning series from Heinemann and I’m the senior content advisor with the academic team to Dreambox Learning.

One of my favorite things about Dreambox is the way that we deal with assessment. It’s not a drop here, and a step here, and a test here. It’s seamless. It’s formative assessment at its best in that we’re accumulating data every time a child sits down at the computer. We’re accumulating data every time a child manipulates one of the Digital Manipulatives because were coding what they’re doing with the manipulative.

This informs instruction in a way that’s dynamic rather than static. When school districts don’t have a vehicle like Dreambox, they tend to approach formative assessment by doing 3 or 4 tests during a year. And assuming that the first one gives them baseline data, and the second one a couple of months later tells how much progress kids were making, and the third one later on and here kind of prepares you for the fourth one that you going to use by the end of the year that is your outcome based summative.

Heraclitus once said “You can’t step in a stream in a same place twice”. I like thinking about that when I think about assessment. I think when we use a test, it’s a static form of assessment because all were doing is seeing what the child can do on that test in that moment on that day given the way we’ve crafted the task – the tasks the items. In no way we’re measuring children’s real understanding or real learning, because learning is far too complex to take static one shot glimpses.

Formative assessment, genuine formative assessment can only be done the way we’re doing it in Dreambox. Meaning that, it should be mining the data continuously. It should be collecting data and compiling it, so that we’re getting full picture of what the child actually knows and then sharing that with both teachers and parents, and that’s how we’re informing instruction and it’s truly formative assessment.

One of the things we tried to do in Dreambox is build multiple tools that allow multiple ways into content. And some children choose one over another like some better than others do better with some then they do with others and we also build repertoires that way of multiple ways to represent concepts.

I think exploration is absolutely critical for learning. Children need many opportunities to try things out, to wonder if they work, games can be used to support children. I think games should have many openings and possibilities that should not just be a game for practicing reinforcement but should be a game were kids are actually exploring and learning by that exploration.

I think when teachers are watching what children are doing in Dreambox, they are actually learning a lot of math themselves. I have a lot of teachers tell me, “I had no idea I could do some of those strategies for computation.” “Oh my goodness some of those strategies kids are learning in Dreambox are strategies I didn’t even know”. Many teachers will say the only way I notice subtract is to regroup. That’s the way I learned, and now all of a sudden, children are solving 71 minus 36 by saying is equivalent to 70 minus 35 I’m going, “Oh my gosh wait a minute. Is that kid right?” So teachers are learning a lot as they are watching what children are learning in the environment, but I think platforms like Dreambox are really intelligent adaptive learning environment that is going to absolutely revolutionized what we do in education.

Cathy Fosnot, Ed.D.