Individualizing Instruction Through Intelligent Adaptive Learning

Webinar Date: May 09,2013

Webinar Description

Providing truly differentiated, individualized instruction has been a goal of educators for decades, but new technologies available today are empowering schools to implement this form of education in a way never before possible. Intelligent adaptive learning software is able to tailor instruction according to each student’s unique needs, understandings, and interests while remaining grounded in sound pedagogy. In this web seminar, originally broadcast on May 9, 2013, research firm Metiri Group and DreamBox Learning shared how intelligent adaptive learning works, the role the technology can play in raising student achievement, and the research base required for districts to invest wisely in these new tools.

View Transcription

KED:    Hello, everyone and welcome to today’s District Administration Web Seminar. My name is Kurt Eisele-Dyrli. I’m the magazine’s web seminar editor and I’ll be your moderator. The title of today’s event is “Individualizing Instruction through Intelligent Adaptive Learning,” and it is being brought to you free of charge by our sponsor DreamBox Learning. Providing truly differentiated individualized instruction has been a goal of educators for decades but new technologies available today are empowering schools to implement this form of education in a way never before possible. One of these new technologies is Intelligent Adaptive Learning software and this is the topic above our presentation today. We’re going to hear the latest findings from Cheryl Lemke at research from Metiri Group about how Intelligent Adaptive Learning works, the role that technology can play in raising student achievement, and the research base required for districts to invest wisely in these new tools.

Before we begin though, some brief housekeeping notes for you. This is the WebEx platform and panels for communicating with us are located on the right side of your screen. Please note you may have to expand these panels. To do that, you just click the small triangle, that’s to the left of the panel name. If you’re having any trouble listening through your computer speakers or if you prefer to listen over the telephone, please just click the request telephone button under your name in the attendee panel. A pop-up window will appear with a phone number and access code and we’ll also post that phone number in the chat panel that you see on the middle right of your screen. Speaking of that chat panel, please use that to send a message to our host and producer, Kylie Lacey, about any technical issues you’re having. If you have a question for our speakers, you can ask them at anytime. We ask that you use the Q&A panel that you see in the bottom right hand corner of your screen. Please send your question to all panelists, which you’ll see is the default option. We’ll answer as many as possible during the Q&A session towards the end of today’s seminar. Also our speakers’ slide decks will be available for you to download, and the archived recording of today’s presentation will also be available for you to review or forward to your colleagues. I’ll tell you more about that a little later on.

So, with that, on to our program. Our presenters today are Cheryl Lemke, CEO of the Metiri Group. We’re also joined by Tim Hudson, he’s the director of Curriculum Design for DreamBox Learning. So, first Cheryl is going to start it off, so with that, I’m going to turn it over to her, Cheryl, welcome.

CL:       Well, good morning from Los Angeles, California, delighted to be with you today. I want to provide some context first of all for the work that we’ve been doing for DreamBox. They commissioned us to actually develop a paper on the theoretical underpinnings of Intelligent Adaptive Learning. In other words what does the research say about why US principals and education leaders would even want to consider an Intelligent Adaptive Learning System, and I must say that one of the things that we’re finding is that this is perfect timing, that with this next generation, use of technology, you really can do things that have not been possible before, and the timing is right for the Common Core State Standards because of course we’re all aware that we’re actually moving from more of an algorithmic approach to Mathematics which, you know, the former standards had a lot of that in it into a deeper learning of really understanding the theory behind and really having an intuitive understanding of mathematics, and this kind of technology, this next generation can really do that. So, let’s talk first of all about what we’re going to do today. We’re going to talk about the What—which is definitions. We’re going to talk about what the Learning Advantage is, and then we’re going to talk about the research basis for it and why you would even want to consider it.

I would also suggest to you that the US Department of Ed is actually advocating for this kind of learning that they’re recognizing that with big data and learning analytics, you can actually, behind the scenes, track what a student is doing and use that to inform the way that you present students work, that way that it’s paced, the way that it’s structured, the way that it’s presented. So, there’s some real opportunities to really take advantage of the data in informing what, where, when, and how we work with students to at the end of the day to improve their learning.

So, first of all, let’s talk about the What. I have a definition up on the screen for Intelligent Adaptive Learning and there’s three bullets there, and really that is the essence of Intelligent Adaptive Learning. One is, it immerses students in modular learning, and that means, you take the curriculum, that it’s broken up into chunks or it’s modularized and then it allows students to really go through that curriculum in kind of an individualized, personalized way. So no child really sees that curriculum in quite the same way, because the system adapts and adjusts and personalizes that learning. The second one is that, the system is actually listening to the student. It’s capturing the data with every keystroke that the student does, and then considers that in terms of a cognitive model that has, that it’s designed behind, and then it actually decides what the student should be doing next according to that. And the third one is just simply that it guides the student’s learning experiences for pacing and for path.

So, the essence of it is that, there are three words in the definition, intelligent, adaptive and learning, and we’re going to talk a lot about the learning. But, I think the adaptive is really important, in the sense that it adapts to the child and personalizes a child by sequencing in a way that is comfortable for the student and is enabling the student to be successful. It has multiple experiences, and so there are multiple ways to teach the same concept, and so if one’s not working, it can move into another one which is really powerful. It customizes the presentation and so some students are more oriented to text than they are to simulations or animation or visuals. Others have other preferences, and it can really customize according to that, and then finally it individualizes that pace, and it really enables a student to work at their own pace.

We also have the word intelligent at the front of that, and this one is probably the most important, that there is a cognitive model behind the Intelligent Adaptive Learning systems that we’re seeing out there. And one of the things that you’ll see today when Tim speaks is that the cognitive model that is really behind the DreamBox Intelligent Adaptive Learning is based on experiential learning and active learning and inquiry learning, and we who have been in education for a while kind of consider computer-assistive instruction to be more didactic—that just kind of pushes things at children. This is not your father’s CAI, this is very different, and there is a cognitive model behind. So, those of you who are interested in using this kind of work really need to probe when you’re looking at Intelligent Adaptive Learning systems to make sure that the cognitive model, the pedagogy if you will is in sync with your approach, and in this case you’ll see that it’s really an inquiry approach in an active experiential learning approach in the DreamBox piece. The second piece is that there’s an intelligent analysis of data, and so it collects literally a 100,000 pieces of data an hour for every hour that the student is on, and informs what’s next for that student and what the probes are and what the responses are to that student based on that. And finally it’s extremely interactive.

Why now? It’s interesting and I mentioned earlier that the timing is really right. I’ve got three ideas here that I think are really converging and allowing us as educators to do things that we’ve never been able to do before. The one is the cognitive modeling about how people learn. We know so much more than we did 10 and 15 years ago about how people best learn, and so we can use this cognitive modeling in these Intelligent Adaptive Learning systems. The second one is that we have affordable and sophisticated devices, and the prices have dropped so significantly that we’re really moving into that area of ubiquitous computing and enabling us to do this. And the third one is the big data that I mentioned earlier that you can actually collect lots of data behind the scenes, and inform the decision-making that’s going on. So, it behooves us to take a look at it, and, you know, we typically have one teacher with 25 or 30 students, think about the power of having an intelligent adaptive kind of tutor available for every child that’s there is. It’s pretty powerful.

I also, I’m not going to dwell on this one, but I wanted to bring to your attention. This wonderful model of Blended Learning that Heather Staker and Michael Horn developed and published in 2012. They actually have a model that talks about different variations of Blended Learning from mostly bricks-and-mortar, all the way over to mostly online learning. And I just want to point out that this, what we’re showing you today can actually be used in a variety of ways. It typically what teachers do is they use it as a supplement to teaching mathematics to kindergarten and first and second grade students. They don’t use that as, this is the only approach that you’re using, but it is certainly used as a supplement which is the rotation model within that Staker and Horn model.

So, let’s really get into the meat of this, first of all, the Learning. One of the things I think we’re all well aware as educators is that there is a zone of learning that we want to keep kids in, and what we want to do is we want to actually balance how complex the tasks are that we give to students which is the vertical axis to the level of skill that the student has and what we’re really trying to do, if you look at the bottom of that rectangle, that we take students where they are, and so, they come to us with a lot of prior knowledge and what we want to do is move them smoothly through this optimized zone of learning into the achievement of the learning standards. And so, we’ve got to balance that task complexity and the skill level, and what happens is students go through that optimized zone of learning in very different ways and the Intelligent Adaptive Learning Systems really even allows them to have even more complex pathways through that. But what we don’t want them to do is move out of that zone. And what happens is, if the task complexity is really too high for the skill level of the student, the student gets out of that zone and becomes very frustrated or confused, and on the other side, if the skill level is too high and the task, or the assignment that the teacher gives that student is too low, kids get bored and we know that we’ve got a lot of bored kids out there. And so, what our suggestion to you is that the Intelligent Adaptive Learning system really helps us manage that and make sure that students stay within that zone of optimized learning and as most of you are well aware of, it really is based on Vygotsky’s, the zone of proximal development, plus a lot of other things that have happened in the cognitive literature.

Let me just give you a kind of a bird’s eye view also a schematic of what Intelligent Adaptive Learning System is, and I want to point out that this is not taking over what the teacher does. It really is in the context of what the teacher is doing in teaching mathematics to students and experiencing mathematics with students in the classroom. So, first of all what it does is as I mentioned, it has a modularized curriculum, so students move through it in different ways. It has learning activities that are associated with that, and there’s multiple perspectives to that, and it also has kind of an embedded adaptive continuous assessment. So when the child first comes in, of course there’s a diagnostic test to see where that child is, and it places a student in the right module so that, you know, it’s balancing that complexity of the skill level with the expertise that the student has.

It captures data constantly and stores it in a database and then once it feeds that data into the cognitive model. So, it’s constantly doing that analysis, and then it feeds the data, the decision-making back into the system. So, if the system sees that the student is struggling, the system can actually say, let’s back up a little bit, or let’s go to an alternate activity that perhaps would speak to this student, and so it gets that, so does get feedback back to the student, and of course it gets feedback to the teacher. It really informs the teacher. So at any time the teacher can pull up some data analyses on-screen and see where his or her students really are. So, it’s really an interesting one, and again, we have to have students and teachers really using this in an intelligent way that really moves them forward.

So let me talk a little bit about the research and then we’ll get you in and actually give you a bird’s eye view of what the DreamBox application really looks like. Basically, what the research on Intelligent Adaptive Learning systems says, is there are five theoretical underpinnings to that, and I talked a little bit about this, but let me talk more specifically about each one of those and what the research says about it.

First of all, in tutoring, in the area of tutoring and feedback, I think most of us who are educators really recognize that we’ve known for, you know, for over 20 years that if you take a student and put him or her into a classroom and then you compare how they would do in that classroom to how they would do with a tutor, the research shows that there’s going to be almost twice as much learning happening with that one-on-one tutoring than there is in a 30 to 1 or 25 to 1 classroom. And so there’s been kind of a challenge over the years of could we actually use technology to close the gap and to enable students to really use Intelligent Adaptive Learning to do that. And I must tell you that that ceiling’s been broken. We’ve really broken the barrier with that, and the research now shows that Intelligent Adaptive Learning systems are as effective as personal tutors. Wow, I mean, think about that, if you could literally bring a personal tutor into your classroom for every child. It’s pretty amazing that we have actually just broken that barrier and really come into it. And the secret sauce in the tutoring is the targeted feedback. If you look at research out there, and one of the sources of research that we used extensively was a meta-analysis of 800 studies that was done by John Hattie, and if you’ve not seen that, you might want to take a look at this.

This is pulled out of the Hattie research and is a specific piece of it by William’s of 2009, but what they talked about is that there are three aspects of feedback that are really important. One is frequency, and that is important, although there’s only an effect size of about .34, which is significant, but it’s not huge. The real big effect is really when you get formative feedback, when you not only provide the student with some information about, did they get the answer right or wrong, but probe a little bit about why they did, and actually get the student to reflect back on it, then you get an effect size of about .9 which is huge. And then finally the questioning that we do is really important. So, not all feedback is equal and so we have to make sure that we do it appropriately.

This slide actually demonstrates that. It has from, there’s three bars there, the top bar is strong formative feedback that actually has that interactivity in the probing and the questioning with students versus moderate feedback which does some of that, but not as much, and the finally the weak formative feedback, just simply tells kids whether they’re right or wrong but doesn’t do much else. And you can see there’s a real difference in how students learn in order to really take a look at that. And finally, in this area of tutoring, the research shows that right now, students receive very little feedback in the classroom. You would be dismayed to hear how many minutes of feedback students get per hour with instructors or per week. And it also suggests that much of the feedback that they get is from peers and it’s often times incorrect, and we can do something about that. So, we can do some changing with that.

So, I’m now going to turn this over to my colleague, Tim Hudson, who will be showing you kind of examples of the tutoring aspect of DreamBox.

TH:       Thanks so much, Cheryl. I appreciate it. And thanks to everyone for joining us today. My own background, I was a high school mathematics teacher in suburban St. Louis and then was a K–12 District Math Coordinator, and came across DreamBox in that role, because, as many of you out there, I’m sure are struggling with, there’s a lot of prior knowledge gaps at any grade level in elementary schools, we have the challenge of differentiating, of remediating, of accelerating; a lot of complexity there and technologies like DreamBox are well equipped to support school and district and teacher goals to help reach all learners.

One key thing that, first key thing I want to say building off something Cheryl mentioned earlier is this idea of cognitive modeling, the idea of learning principles. There’s sometimes out there the sense of, well, math software is math software, you know, there’s a little bit of, little bit of explanation, and then a little bit of practice, as if we’re just trying to digitized a textbook, but that’s not all what we do at DreamBox. DreamBox is fundamentally different for a lot of different reasons, the intelligent adaptive learning approach using the principles that Cheryl discussed and we’ll continue talking about for the rest of this hour. The key thing is inquiry is at the core of what we do. It’s not computer-assisted instruction, because in DreamBox we don’t actually have explicit instruction. DreamBox is not a sit-and-get or a sit-and-listen kind of thing. The DreamBox experience as I’ll be showing you is about sitting and thinking. So students thought is the forefront, and to share a couple of learning principles that inform our work from Wiggins and McTighe’s “Schooling by Design”: “An understanding is a learner realization about the power of an idea,” and that, “Understanding can’t be given,” you know, no matter how much we try to explain and give kids understandings, it just can’t happen. We have to engineer learning so that learners see for themselves the power of an idea for making sense of things. And this is what great classroom teachers do, they engineer learning experiences and DreamBox follows that same design.

For example, Division with Remainders, something students struggle with, often many of us just were sort of shown an algorithm and long division for example, and then asked to practice it quite a bit. But, before we would give that to students in DreamBox, we need them to work with the ideas of division. So, here’s an example, we have a, we call it the Bag-O-Matic, and we show students, you know, thousands of gumballs. Obviously the can’t count them, but there is some element of concrete representation there. And the first thing we ask students to do is to start packing gumballs using friendly numbers, using their own intuitive strategy. So, building on what I said just a second ago, we don’t ever say, here’s the best way to pack these gumballs, we instead ask students to start thinking and give it their best thinking. They’ll see five red dots there on the green strip in the middle of the screen, that means kids can, you know, use this Bag-O-Matic five times and as they progress through the different lessons, they get fewer and fewer, sort of moves to make, because we want them to be thinking more strategically.

So, at this point, I’d like to real quick do a poll and give you all 15 seconds. So, the problem is 5,916 divided by 4. And if you were doing this, what would you pack first? Let’s see, just one more second, really. Okay. So, the poll has ended. But, we don’t actually give students multiple choices. We expect them to pack the gumballs using whatever number they want. Most adults tend to gravitate towards perhaps one of those four options. But in DreamBox, students would only see, only see the pack button and they’d have to create their own. Maybe they wanted to try packing 80 first, or 800. The key is, is that students. Oh, okay, we’ve got a lot of no answers there in the poll results, probably because the poll went pretty quickly, and not a lot of folks had a chance to respond. So, if were to be watching a student do this problem, just as a classroom teacher would if they gave their class this problem and then watched how they solved it. If a student were to pick 400, then, would create the equation 400 divided by 4 and they would actually see those 100 bags that they just created and see how many more gumballs they have left to bag. So, hopefully students at this point would be thinking more optimally and maybe they would try 4,000 next.

The goal being, we’re trying to help students realize ideas. We’re not trying to give them understandings, because cognitively that’s not effective. Here’s a different problem, but the same lesson, division with remainders. You see that this student, you know, packs 700 gumballs and 700 and 490, and in DreamBox, vocabulary is a key thing that we develop, but we only introduce the vocabulary like the word remainder here when it’s needed in context. We don’t front load vocabulary because there’s a good reason from a cognitive stand point to, for students to use a word and learn about a word when that word is needed. And we also ask students to interpret remainders which kids traditionally have a hard time doing. So, now, I’ll turn it back to Cheryl to take the second design element.

CL:       Great, the second one is really about sequencing. It’s about allowing students to go through this modularized sequence in the way that’s comfortable for them based on their prior knowledge, their skill levels, and to some extent, their preferences and they can go, in some cases they can actually go back and do some more practicing, and if they feel like they need it. So, they do have some choice within. But this is really powerful because, again, if you go back to that zone of optimized learning, one of the things that we’re looking at is the idea of personalizing, the idea of making sure that there’s a challenge in there, and you may get some kids to go through in this way, and some kids go through in another way and some kids go in a third way. So, it’s all, it’s really personalized and so the chunking or the modular part of the curriculum is really important and of course we all are aware that there is some sequencing that is required in mathematics, and so the system actually, certainly knows that. But often times there are ways that you can jump in, you know, kind of a parallel way to different things depending upon what the student is, and the reality is what we’re trying to do is keep them is that zone making sure they’re interested, they’re motivated, and they’re always provided a challenge, just like gaming, they’re always provided a challenge, yet it’s a challenge that is within their reach. So, I’m going to turn it back to Tim who’s going to talk a little bit about how DreamBox sequences.

TH:       Thanks, Cheryl. Yeah, the sequencing of lessons and the development of ideas as far as students coming to make sense of mathematics, coming to deeply understand mathematical ideas and develop true number sense, that’s the key thing that every classroom teacher knows, you know, what we’re learning today will be built upon down the line. What our technology enables, what Intelligent Adaptive Learning enables is to not make that as linear a 180 days sequence through classroom lessons, where, you know, we’re working with a whole group students on sort of moving along according to a pacing guide or something like that. So, an example of the key sequencing ideas that we, I guess like, a key way that DreamBox does sequencing for students is like to plan backwards from sixth grade for example. The distributive property is in the sixth grade Common Core standard.

So, if you are in a Common Core state, being, for students to be able to use the distributive property that’s an expectation for sixth grade, and many of us learn this in school, the sort of term foiling, if that brings happy memories for you in mathematics, the whole first outer inner last thing, but at the whole other conversation about why maybe using the word foiling isn’t so great. But this is a sixth grade expectation, but, I taught high school mathematics as I said and to simply introduce something in sixth grade that’s a key mathematical idea, is a little too late. So, in DreamBox, we actually develop this idea starting way back in third grade, where students are skip counting and using multiplication on the number line and here’s a screenshot saying that, oh, 2 x 6 plus 4 x 6 is the same as 6 x 6. It’s not just about memorizing the fact that 6 x 6 equals 36. You need to be able to break down the numbers and flexibly and fluently work with them. So that’s in third grade, we also develop the array model, where students are here trying to solve 9 x 11, as you can see on the left, and they’re given a piece already, 9 x 9, a 9 by 9 square, and students use the buttons and the rulers over here on the right to construct sort of the missing partial product and again, they’re enabled to create the pieces as opposed to us saying, oh, here’s how you go through and do it. Kids are always the ones that are exploring and testing out their ideas.

In fourth grade, we increased, you know, the size of the numbers. Here, students are trying to solve 12 x 24 and we asked them to unfold a map, and as you see here, they first unfolded it 12 times by 10 piece and a 12 by 10 piece and are about to lock in this last 12 by 4 piece, so, that’s partial products once again, same thing with the distributive property and kids are only dragging this in one direction. I saw a question come through, “Are the virtual manipulatives in DreamBox, drag-and-drop like the PARCC assessment will be?” And yeah, we, all of our lessons have quite a bit of interactivity that will help students either on the PARCC or the Smarter Balance assessments, because they are very interactive.

Moving on, again this is a sequence heading towards understanding the distributive property in sixth grade. After students have worked on those ideas, we actually give them a two-dimensional open array to work through and this is one that students rarely have an opportunity in classrooms, a lot of print materials and other software don’t focus on the connection between multiplication and the array. Let alone give kids a really engaging experience—these googly eyes kind of follow the cursor around—but, we empower students to create their own partial products to eventually solve 38 x 78 and we’re able to respond to their initial conception, they can create up to six arrays as you see there in the instruction, and you know, this student isn’t really using in optimal strategy, however, it is their thinking, and we don’t front load DreamBox lessons with how students should think about it, but rather empower them to think about it and make sense of it themselves. In later lessons we only give them four arrays, for example.

Here’s another lesson in this progression, where now, it’s not dragging out individual pieces, but it’s moving this one button that’s in the middle of the screen to simultaneously create four partial products, looking for an optimal solution, where over on the left they’ll have to add up all these partial products. Then we get into a place value sort of column multiplication situation where students multiply 8 x 500 then 8 x 10 then 8 x 7, it’s kind of just a different way of looking at the standard algorithm for multiplication, and really developing a strong sense of the distributive property and how things are related. Then finally, by the time students are in sixth grade and working on sixth grade content, when they’re ready, they get to work with this particular array in DreamBox. So, all has started building way back in third grade, so that students have a deeper understanding in mathematics and it’s not just a skill, a foiling skill that they’re shown in sixth grade and expected to really understand deeply. I’ll turn it back to Cheryl.

CL:       Great. Thanks, Tim. It makes me want to go back to school, and in fact, as a former mathematics teacher, this is so powerful because it really does allow students to get into the fray and play with mathematics, and get that intuitive idea of what it is to add and to multiply and what those properties are. The third area that is important in the Intelligent Adaptive Learning system is really the pacing, and we have really strong research. I mentioned on the screen here that Hattie mentions 377 studies of mastery learning, and part of the power of mastery learning is you really don’t go on in the sequence until you’ve actually mastered it at 80 percent. And the research is really clear that when students are in this kind of environment, they actually learn more. And we know that this is very difficult for students. I, as a former mathematics teacher tried to individualize learning for my kids and it was just, you know, it worked for a while, but I was staying up until two in the morning every day trying to do it. This one allows, gives the teachers the tool that they need to do, to do that individualization, and allowing kids to pace themselves and not to have to get into some materials that they’re really not ready for just because the class is.

So, we’re going to turn this back to Tim, who’s going to talk a little bit about how the pacing works in the DreamBox Mathematics K–2 environment.

TH:       Thanks, Cheryl. Here’s what DreamBox looks like to a student in grades K through 2. You see three options there, three lessons; one using the math rack and the 10 frame are some tools that are fairly widely used across the country and around the world to help kids make sense of base-10 system in early numbers, and the great thing is, you know, if you’re a classroom teacher using either of these tools, to help kids make sense of it, and maybe every student has one in their hand, it’s very difficult for a classroom teacher to get around to really talk with every student about the manipulative, about how they’re thinking about the manipulative, you know, sometimes kids have wandering eyes to take a look at what their neighbor has done or how they’ve solved the problem or built the problem. But in DreamBox you can be confident that every student is using those tools with specific questions that are designed to make them think and help them make sense of the ideas, and as a teacher you get the reporting that informs you about which students have make sense of these ideas.

There’s usually multiple options for students to choose from, which is that element of choice that Cheryl was talking about and it’s not at all linear. Students in the primary engagement can be working on second grade; they can be working on second place value and at the same time in third grade skip counting. So, usually students in DreamBox are working on ideas in multiple grade levels. For intermediate students in the upper grades, there are more options, more opportunities to as you can see here to practice things they’ve just learned. We have some concept videos that are fun to watch and sort of thought provoking, still not explicit instructions, just things for kids to think about, and then recommended lessons that are open or in progress that helps them make forward progress in grade level content. And back to Cheryl.

CL:       Thanks, Tim. The fourth area that we want to talk about that really, again has good research behind it is the regulation of cognitive load, that one of the reasons that we sometimes see students getting out of that optimum, you know, zone of learning is that we throw too much at them at once, and they really don’t have the capacity to handle it, perhaps their skill level’s not up to it. And there’s some good research behind it. One of the things that we know is that we have a working memory that all of us as humans have this limitation. We have two parts of working memory, we have one that’s verbal and text, and we have a second one that’s visual, and the two of them worked together. So, that working memory is where we think and you can see how those words that we really want kids to do, the sense-making, the schema development, that diagnostic, the integration, that’s really where we want our kids to be, and so they get information through the verbal channel which is talk, sound, and text.

They also get the information through the visual channel and one of the things that we know is that we can only hold seven, plus or minus two things in the verbal working memory. So, I don’t know if you’ve ever experienced this, but perhaps you’ve walked in, you’ve walked from one room to another in your house and you’ve gone into the room and you said why did I came to this room again? Well, it’s not because you’re getting old, it’s just because something distracted you. You already have a full working memory and if you wanted to engage with the new thing that distracted you, something had to go and obviously it was the thing that you were looking for.

Kids have that too, and we see that with children, for example, we see this with children that are learning to read, where you might have them, you might have them look at a passage and read a passage and perhaps they’re still in the decoding mode and so, they struggle with the decoding, but they get through the passage, and then you look at them and say, what did it say? And of course they have no clue, because their working memory was full up with that decoding piece and they had no place for the comprehension. And so one of the things that DreamBox and Intelligent Adaptive Learning systems try to do is make sure that the student, that what they put in front of the student, the student is ready for, and they have all of the things that students need in order to solve the problem on the screen. So, there’s a good instructional design, so the student doesn’t have that cognitive overload or the cognitive dissonance.

The other thing that it does is the visual working memory should be constructed, so that the visual works with the verbal and it doesn’t repeat it, but instead it adds new dimensions to it. And if you’re only using the verbal and the text and you’re not using visual, then you really underutilizing the working memory that students have, and so the way that we need to do this, the way we need to do this is we need to organize information in a way that it gives everything kids need on a screen, which DreamBox does. We need to organize visual and verbal together, and that’s visual and text as well, the verbal stands for that. So, the screen design is there and that enables kids to do that schema building that they really need to do, because if you want to use your working memory wisely, you have to automate some things. So, for example, back to the example of the student with decoding, once that student learned to decode that was automated, no longer had to take up working memory as they were trying to get on to the comprehension. So, back to Tim and he’ll talk a bit about how DreamBox does this.

TH:       Cheryl, that reminds me a lot. I took Latin in high school, and I could pick up some Latin text right now and read it, probably pretty fluently clear through, but would have no idea actually what it said, cause it’s so much dead vocabulary left for me, I haven’t had many people to speak Latin with since high school. So, there’s working memory. Problem 29 + 72 in DreamBox, one of the things that we do is we develop, you know, multiple strategies. Having number sense means that you don’t immediately reach for a calculator or a pencil when you see a problem, it’s you’re able to play with numbers and think strategically about numbers. So, for 29 + 72, we give students, what’s called compensation buckets, and you can, there’s counters in each bucket. You don’t see them, but you can use the up and down arrows to move the counters from one bucket to another. There’s a limited number of counters you can move because we know that when we watched kids test this out, they tried to move all of the counters into one bucket to solve the problem. But the goal is to get to friendlier numbers, and this is a very manageable, to Cheryl’s point. The cognitive load here is very minimal. There’s not a whole lot of things to attend to other than the numbers that, is really what we want to be the focus of the lesson. So, students could turn this problem into 30 + 71, and realize that the conservation of number, the compensation as an addition strategy is pretty valuable, but with limited visual and verbal cognitive load, where as a lot of adults look at 29 + 72 and grab a pencil, and say, all right, I’ve got 11 and I’m going to carry one. We’re helping students to think more strategically about the numbers and become more fluent.

Also for something like multiplication of fractions with, many students and adults struggle with mainly because it’s often just taught as a, here multiply the top numbers and multiply the bottom numbers, which is not something for students to understand, but rather something for students to regurgitate or parrot back. In DreamBox we wanted kids to realize, as I said before, we wanted to engineer our lessons so they realize the power of multiplication with fractions and have some a-ha’s about that idea deeply. So, what you see here is one screenshot of several problems that a student has worked through. What we first started with was a unit square over on the right and asked students to create 1/12 of 1/3 using the different pieces that you see here, dropping them into a drop zone looking at these little coin pieces, and the first thing you see is 1/12 isn’t an option, but students are able to compose that number by doing 1/2 of 1/2 of 1/3.

This lesson is actually pretty far through the progression of this particular unit in DreamBox. But the key idea here is that students build 1/12 of 1/3, and they can see visually that there’s 36 pieces there, 12 x 3, and we ask them to shade in different portions, so the second problem you see there is, well what’s 1/12 x 2/3, so instead of just shading in 1/3 a piece, you shade in 2/3 and then they ask what 5/12 x 2/3, and that’s actually the image you see here, that 5/12 x 2/3 is what’s shaded in the pink, what’s key here is a lot of adults don’t even have this sort of array model, this mental conception of what’s happening when you multiply fractions. It’s too much about the rule and not enough about the sense-making. So, kids are always required in DreamBox to create these pieces and if students get things wrong we reference, we have very short verbal feedback to give students if they’re wrong. We say things like, “Look at the total number of pieces, and look at the pieces you shaded in.” Very simple directives that don’t just give them the answer, but rather spur their mind onto additional thoughts.

And then lastly here with 13/12 x 2/3, with this problem, this particular point in the learning progression, students don’t actually create it on the array, you actually, you want students to be able to solve these problems without always using an array, without always having that concrete model. So, this is a transfer problem, but it’s still based on the similar, the same ideas that students have been working on before, oh, 13/12 x 2/3. Well, if I would have created that, how would that look, and that’s how we’re engaging students in transfer, and all the while a relative low cognitive load as this lesson progression plays out.

CL:       Great examples, and I particularly like the combination of the visual and the text on these screens. It really does use the full capacity of students’ working memory which is really smart. The last story that we’re going to talk about today is engaging through gaming and the I think we us educators are well aware that kids are fascinated with games, and that is, it actually has, if you look at the construct of most games, it’s really in essence a great learning strategy, that there’s a logical sequencing, in that there’s tiers and levels of award and progression through the game, that there’s novelty and variety, something’s always different, that there’s student choice, that the students make a lot of choices as they go through, they can go back and try it again and there’s no negative aspect to that, that learning, and that’s what learning should be. It should be the fact that’s its trial and error, but, and that, you know, you’ll get to the right answer eventually, but there’s a number of different ways to get to that answer. The intellectual safety, the fact that you don’t always have the right answer, there’s a lot of trial and error there, and finally the clarity of goals, that you kind of know where you’re going. And we do see a lot of research in the area of gaming, some of it’s positive, some of it’s negative, especially with, you know, the violent part of this. But if you look at the constructs here, I think you’ll see some aspects of gaming in Intelligent Adaptive Learning systems that fully engage students and enables them to do that trial and error and be motivated enough to come back and try things again.

Also, I want to give you a piece of research that really talks about kind of the multi-model aspect of simulations and modeling, etcetera. That if you look at the differences in using modalities in media, what you find is that when you use non-interactive, which some of the DreamBox really are, that, with students, and you’re addressing basic skills which we’re both in the DreamBox piece. You can actually get an effect size that’s significant, in fact increase the student learning by 21 percentile points, which is huge and in non-interactive also increases the higher order. If you get into simulations and modeling and kind of the more complex things, what we find is, when you’re teaching basic skills often times you don’t get any effect, so you might as well kind of stick to the core of the non-interactive, but for higher order thinking, the impact is huge, when you use simulations and modeling and that interactive kind of media, you really do get a huge gain in student learning. And again, theoretically, this makes sense and it really provides that foundation for Intelligent Adaptive Learning systems.

And here are the five principles of effective gaming that you can take a look at. It’s very similar to what I started with in terms of the cognitive modeling. And I’m going to turn this back to Tim, who’s going to talk a bit about the gaming features that are in DreamBox.

TH:       Thanks, Cheryl. So, one of the key things about learning games, especially games with mathematics is we often underestimate how complex numbers really are and mathematical reasoning really are. Sometimes people kind of think that, you know, learning your multiplication in fact is basically the same idea, the same sort of thing as learning the state capitals, and that’s really not the case when you are learning state capitals, they’re somewhat arbitrary. There’s not a whole lot of sense-making to be headed to why certain states have certain capitals, there’s historical context there for sure, but from a learner’s stand point, to know that Albany is the capital of New York, for example, there’s, yeah, there’s not a lot of sense making to get that fact or even a lot of transfer that’s going to happen later on with that fact. Whereas with math facts, knowing that 8 x 6 is 48, there’s a lot of relationships underlying that, you know, well, 2 x 24, and 3 x 16, all of these things are 48, plus I’m going to need to use that information 8 x 6 is 48 to know 80 x 6 or 80 x 60. These are, these facts are not things that are as arbitrary and sort of rote as state capitals are. So, when we design games in DreamBox, we realize the sense-making that needs to happen first and the transfer that needs to have it afterwards, and so, we let the math drive the game, rather than create a game and then overlay some math with it.

Here’s an example of a pirate ship sort of game using the number line where students use the plus and minus buttons to create equations to help pirates find the treasure. So, in this case, it’s unknown where they started, but they know that they moved a 134 units and landed on 965 units, so students, this student here decided to subtract 100 and subtract 30 and is about to subtract 4. So, that’s how they’re solving this problem, and it’s pretty motivating because they’re looking for something along the shoreline and it’s their performance that is directing the movement of the ship. At far as sequence challenges go, later on we use this same game, but now, we’ve taken away one of the buttons, and students can only move the ship using the plus button. We’ve also added decimals in here. So, what this teaching is how students can add and subtract negative numbers, and by only allowing them to use the addition button, they realize that I have to add negative numbers if I want to move left. And kids only of course see this lesson once they’ve shown proficiency in the earlier ones.

Lastly, as far as timely specific feedback, this is our miniature golf game to help develop fluency. Its features aren’t terribly adaptive because once students have reached this point of getting this fluency game, they already understand conceptually what’s going on. But we still have—to Cheryl’s earlier point about high order thinking—what we have here, is students are given the problem, the miniature golf ball starts at zero and you see the flag there, the hole at –8.3, and students have to choose two numbers whose sum is –8.3. But we have a windmill in the middle of the way, so, and we also have fractions and decimals on the score cards, so students have to choose two numbers, but they also have to be thinking very fluently about those fractional values and decimal values. Plus they can’t putt their ball to get it lost in the windmill, so there’s even a lot of thinking students have to do about, you know, what’s half way between zero and –8.3, how do I not choose a number first that’s going to get stuck in the windmill, quite a lot of good thinking and quite a lot of fun.

CL:       Great. And in summary, we want to get to a few questions at least during this session. In summary we just want you to think about the fact that this typically is used as a supplement to the regular curriculum, but it actually puts kids in, and some schools have used it, for example, where they put kids in for a 90-minute sessions for two days a week. There’s a variety of other models that people use and Tim can probably address that. But, in summary, you know, one of the things that we’ve noted is theoretically there is good research behind the elements of an Intelligent Adaptive Learning system. That it personalizes and it allows teachers to do blended learning in ways never before possible, that it uses data in really powerful ways and provides that data set to teachers who can actually use that data-driven instructional decision-making for other things in the classroom when they’re teaching mathematics that are really even outside of the system, and at the end of the day, we’re really trying to get a deeper learning.

The bottom line is that there’s good theory behind Intelligent Adaptive Learning systems, but, and there is, this is a big but, to date there are research studies that empirical research studies that are underway on specific Intelligent Adaptive Learning systems like DreamBox and there are others out there that really are not complete yet, that it just takes time and this is such a new thing in the fray, and so, within a couple of years we’re expecting that we’ll see empirical research in addition to the theoretical underpinnings. I’d also like to say that this paper that I did is broader than just DreamBox, that’s really about that whole genre of Intelligent Adaptive Learning systems and that will be made available to you, so, back to Tim.

TH:       So, in summary, in DreamBox, we have a pre-K through Grade 5 mathematics program that’s web-based and students can access it at school and at home. It combines rigorous elementary math, I’ve shown you several examples of that, it’s a motivating learning environment, we didn’t talk a whole lot about that aspect, we did talk about the gaming fundamentals and really the student-directed empowering type lessons that we create, and then the Intelligent Adaptive Learning Engine, that was really the key piece of our talk today that results in millions of individualized learning paths. It’s not a linear progression and it really helps meet all students, right where they are, tailored to their unique needs, to help differentiate, accelerate, remediate for every student in real time based on their strategies and their thinking.

For more information you can visit Tons of information there, whitepapers, including the whitepaper that Cheryl has written and I think now, I believe now, it might be time for some questions.

KED:    Yes, definitely great. Thank you so much, Tim and Cheryl. We’re going to get to some of your questions, but just a reminder to our audience, I know there was a lot of information there in the survey findings as well as a little bit about DreamBox and how that works. You will have the opportunity to download these slides and this event is being recorded as well, so, you’ll have the opportunity to view the archives—you can go back over everything at your own pace, if you want to review the materials that were in here. Let’s get to some questions. Again, another reminder, just submit a question in the Q&A box, you see in the bottom right hand corner of your screen there. Tim, you touched on, there was a question earlier about the PARCC assessments. I know that a lot of people are concerned and interested with the Common Core and the next generation of assessments. Perhaps you could touch on how DreamBox helps prepare students for those new assessments.

TH:       Sure. I think one of the main things to know is that most cases, certainly in many cases if you take a look at what the free the lessons that we have that you can see on our website and then compare them with the PARCC or Smarter Balance list items. One of the things you’ll notice is ours are quite engaging when compared to the assessments, and that’s kind of to be expected. The assessments are designed really to capture correct answers and incorrect answers and some elements of sort of, you know, preformed thinking with text boxes, it’s which we don’t currently have in DreamBox. But, because we build all of our tools to analyze student strategies as you saw there with the googly eyes, you know, how our students breaking apart the problem 73 x 28, because we build our tools to capture student thinking. They are on average most of the time, very engaging, far more interactive than what the PARCC assessment online items will be, and that’s, ’cause there are different design specifications for our tools as opposed to theirs. Its great preparations handle the assessments.

KED:    Okay, excellent. Thank you. Now, there’s another question about how this product, DreamBox or Intelligent Adaptive Learning fits into a regular school curriculum? I think you both touched on that, but, it kind of gets to the question about implementation. Tim, do you have, does DreamBox have guidelines for implementation or a minimum amount of use in a week that the students should be using the software, things like that?

TH:       Yeah, we have several case studies in implementation models for people to download from our website. There are tons of different implementation models. We recommend 60 minutes of DreamBox per week for an individual student. Somewhere between 60 and 90 is kind of ideal. So, we have a lot of schools maybe doing, you know, if they do five hours of Math a week, an hour of Math a day, maybe on two of those days, if they have access to computers in their classrooms, there’s, you know, if there’s two students for every one computer, maybe have the class is working on DreamBox for 30 minutes, the teacher’s working with the other half of the class, and then after 30 minutes they switch and the other half of the class gets on DreamBox. Some find DreamBox just for homework, some have it for an hour of lab time each week. It’s really about an individual school’s schedule, access to resources, but the key thing about an implementation with DreamBox is instead of a teacher having to go in like everyday or every week and look at each individual student’s progress and grade in their class and have to say, okay, these students, I need you to do this in DreamBox, this in DreamBox, this in DreamBox, that just takes a lot of data analysis and time that a lot of teachers just don’t have. So DreamBox is doing all of those formative assessment recommendations in real time so that it takes the burden away from teachers having to do a lot of that assignment work, DreamBox does it automatically.

KED:    Okay, excellent. Thank you. Cheryl, there’s a question here about when your research will be available, it’s available now and if so, where can people access it?

CL:       It can be accessed from the DreamBox site, and I think an email will be sent out to all of those registered for this session that will actually give you the specific URL where you can pick up the PDF, so it is available. I also noticed that someone asked about other subjects such as foreign language, and one of the things I suggest you do is perhaps just get online and search for Intelligent Adaptive Learning and that, and Tim might speak to the subject areas that are currently addressed in by DreamBox per se.

TH:       Oh, yeah, we currently only have mathematics and I’m not aware of any foreign language Intelligent Adaptive Learning products out there.

CL:       I’m not either, but it may be coming.

TH:       Yeah.

KED:    Okay. Okay, great. Well, unfortunately we have come to the end of our time. So, I’m going to have to conclude our Q&A session on behalf of District Administration. Of course, I’d like to thank our speakers, Cheryl and Tim for your very interesting presentation today. And thank you again, of course to our sponsor, DreamBox Learning for putting on today’s event. And to you our audience, thank you so much for joining us. I hope you found today’s web seminar informative. Producing events just like this one is just part of our mission to inform K–12 leaders like you about news and trends in school district management. You’ll find more coverage about issues such as the one we discussed here in the pages of our print magazine, as well as on our website and in our free daily e-newsletter, DA Daily. And for those in the audience who would like to share this seminar with your colleagues or review our speakers’ presentations at your own pace, you could access it from our website by going to, where it will be posted in the archive section in about 48 hours and you’ll get an email notice about that when it’s ready. If you’d like to download our speaker slides, again, you’ll find instructions and a link in a thank you email that you’ll receive later on today.

So, that’s it for today’s event. I’m Kurt Eisele-Dyrli for District Administration on behalf our producer, Kylie Lacey and my other colleagues. Goodbye, everyone and enjoy the rest of your day. Thank you so much for joining us.