Individualized?! That’s what they all say!

I was at conference recently discussing DreamBox with a pretty savvy industry insider when she asked “what makes your product so unique?” My response was “It’s incredibly effective, fun, and highly individualized to a child’s particular learning needs.”

That’s when she said it. “Individualized?! That’s what they all say!”

You can’t really blame her. Lots of companies (past and present) state that they offer individualized instruction through adaptations. But it’s not entirely clear that they are offering individualization in the most rigorous interpretation of the word. Put another way, the concept of individualization encompasses a lot of territory, from the slightest customization to a highly personalized experience. Add to this the fact that individualization is a hard thing to see and experience firsthand (especially with a single student playing), and you can understand her perspective.

Individualized Math Learning Paths

When most eLearning companies claim individualization, what they usually mean is that they have fixed lesson pathways, simple “right or wrong” scoring of answers, basic audio/visual hints, and preset questions. In the end, however, it’s all very linear and predetermined much like an interstate highway. When we talk about individualization, it means we dynamically adapt lessons, hints, difficulty, lesson pathways, pace, motivational elements, and more. That means our product is non-linear and flexible much like, say, space travel. Here’s another easy way to visualize all of this – it’s an example of how we handle lesson pathways (aka “sequencing”) for students:

Individualized Learning Lesson Pathways: DreamBox vs. Competition

Lesson Pathways: DreamBox vs. Competition

As you can see, most companies merely advance a student to the next lesson group whether they struggled or got there with ease. This generally means that all children advance the same way and usually in the same manner. DreamBox, on the other hand, provides many ways to traverse the lessons including offering the student a range of choices within a particular lesson group (i.e., you’ll note all the #3 choices).

Why are we able to do so much? The simple answer is that our team understands and is able to harness the full sophistication and complexity of the Web — such that we can analyze everything from clickstreams to session information. This means we have more tools at our disposal, which in turn allows us to take a more refined approach to adaptations. When you couple this with our intense desire to test what we’ve built again and again, it means that we can deliver a truly individualized experience, which is at the heart of effective learning.

  • Interesting. Given that you have so many lessons, how did you determine how they all are organized and how they flow into each other? I’m curious whether you came up with a bunch of lessons/games and then organized them, or whether you had a list of skills and criteria and figured out how they’d need to build on each other and then designed the games to fill in that framework.

    For games dealing with domains that are not predefined, like a typical physics-based game, it has been suggested that the developer should just come up with a ton of different levels, and only arrange them afterward in order of difficulty (the example refers to linear level progression, but I imagine it could work fairly well for a nonlinear structure as well). This is simply because it is almost impossible to predict the difficulty of a challenge in an unknown domain until you test it out.

    But for a game where you are trying to teach a certain set of skills and concepts, it seems like it might make more sense to come up with a skill chain progression and then design the games or levels around that. Are you familiar with Daniel Cook’s skill chain model of game design? He has some good advice on systematically turning non-game activities into games. His advice is not so much about game design process so much as guidelines and criteria, so I’m curious how you actually approached the process, how you balanced between games-first versus skills-first.

    This is something I am directly struggling with in my own work on Foldit, the protein-folding game. I’ve been trying to revamp the intro puzzles to more adequately equip players with the skills and knowledge necessary to fold proteins effectively and accurately. At the same time I have to make it more fun, so players will want to keep playing. Fortunately, those two goals coincide more often than not. 🙂

    At this point I have come up with what I think may be a more appropriate skill chain progression, but have not yet been able to begin redesigning the levels to match it. There are still a number of infrastructure changes to make to support a more expressive puzzle presentation and feedback. Any suggestions for me?

  • Axcho,

    Again, thank you for your post and questions.

    As you might imagine, the details of “how” we handle our adaptions are proprietary. So I apologize for the lack of advice.



  • “Our expert educators have created a deep, standards-based math curriculum with hundreds of interactive lessons. Our designers have wrapped those lessons in fun game-based adventures.”

    It looks like you’ve already revealed a few hints, on your Vision page. 😉 So, you started with the curriculum, the lesson plan, and then created games around those? Makes sense. Though I’d imagine there would probably be some back-and-forth between the games and the curriculum rather than a simple one-way flow.