Richard Culatta: Customization and Learning

Richard Culatta is an educational innovator at the CIA and is helping the Department of Education build the Federal Learning Registry.

Can you explain the value that data brings to customization of learning? Like, what fundamental changes could it bring to the way education works in this country?
We can get a sense of the value of using data to create a customized learning environment simply by looking at places where data is already being used to drive customization. Amazon and Netflix are two great examples – both are masterful at using data to tailor their site to meet our individual needs. When my wife visits Amazon she sees suggestions for violin sheet music and kitchen gadgets she can’t live without. When I go to Amazon I see educational software for Mac and DVDs of The West Wing. In both cases, we immediately feel as though the system knows us and has saved us time by presenting us with products that meet our needs – even if we didn’t know the products existed. Netflix does the same thing with movies. Their approach isn’t to just let you browse through a library of movies to rent, but to provide you with suggestions for next movies that are so compelling you can’t resist adding them to your queue (I have 135 movies in my queue as of today). Netflix cares so much about providing a customized user experience that they are willing to write a $1 million check each year for anyone who can improve their customization engine by as little as 1%.
Yet, the customization strategy that is central to the success of Netflix and Amazon is largely un-leveraged in education. Learning activities and curricula are shockingly indifferent to the student’s needs, past performance, or desired future application of the material. This is especially common in online learning where data-driven customization should be the easiest to accomplish. I’ve often wondered how much my math performance in high school would have improved if the algebraic concepts (something I didn’t care much about) had been related to flying planes (something I did care about).

Or how much more I would have learned if there had been an awareness of my history of deficiency in basic arithmetic, which often led to incorrect answers even though I correctly applied the steps for solving an equation?  I was recently explaining this concept to a friend of mine from the Department of Education and he said, “It sounds like what you’re suggesting here is creating an IEP for every student”.
Exactly! If we can tell more about a student from looking at their Amazon.com homepage than we can by looking at their math homework I would suggest that we need to be doing better.
Why is it that I can read any news story on the web and not only have many more stories delivered to me that are similar in nature or subject matter, but I can’t get this kind of subject delivery on a broad basis if I am in a high school and trying to learn about something? Why do things seem to be so locked into a system that takes much longer to process new information that is relevant than any other media that is already out there?
I think the answer to this question begins to touch on a much larger, systemic challenge in education: We’re evolving to the point where it is increasingly difficult for a single teacher, no matter how qualified, to provide the level of educational experience we are demanding and that our kids deserve.  Creating customizable learning experiences, including ensuring that a plethora of related and compelling materials are readily available to students, is NOT an unreasonable requirement.  But I can pretty much guarantee that we’ll never get there if every teacher is expected to figure out how to do it alone.  Let’s once again look at Amazon.com.  Can you imagine how ridiculous it would be for each new employee that went to work for Amazon to be expected to rewrite the code that drives Amazon’s website from scratch before they can add their unique contribution?  Yet this is exactly what I see in far too many classrooms.  I truly sympathize with teachers who say, ‘you want me to prepare good lessons, and do all of this assessment, and integrate technology, and involve parents, and customize the learning experience, and now on top of it all you want me to find a whole bunch of related media for students who want to read more?!’
The “stovepipe” educational model, where one teacher working alone is responsible for all aspects of the learning experience, is not sustainable as educational problems become more complex.  We need to do rethink our model in a way that allows teachers to work together to accomplish the hard tasks.  Why not divide and conquer in a way that permits some teachers to focus specifically on finding great media examples to support the learning while others focus on developing approaches to help students who are falling behind.  Some teachers could focus on developing activities for students who need additional challenges, while others might focused specifically on preparing meaningful assessments.

I’m optimistic about new models that are trying to do this. At the forefront are the Open High School of Utah, PA Cyber, and NYPS’s School of One, all examples where technology is being leveraged to change the role of the teacher – allowing them to focus on more specialized parts of the learning experience. Virtual schools hold huge potential, not because the technology makes for an inherently better learning experience but because it allows for good teachers to break out of an ineffective stovepiped model. John Chubb and Terry Moe’s book Liberating Learning it’s worth picking up if you’re interested in learning more about virtual schools.
Also, I can read a news story about North Korea, or about any political event, or about any natural disaster and with the right group of people, I can reasonably predict what will happen next and be able to prepare for it, but there are vast swathes of urban schoolchildren who are joining a chain of generations of people suffering an achievement gap? Why have we not been able to fix this? Is it a political problem, a technological problem, or a people problem?
So is it a political problem, a technological problem, or a people problem?  Yes.
Dedicated researches like Kati Haycock at Education Trust have focused their careers on dissecting the “people problem” part of this issue.  Kati points to the caliber of the teacher as the determining factor for closing the achievement gap: “Take two students who start at the same academic level, and in as few as three years, you could find them in far different places—based on how well they have been taught.”
The problem seems simple enough…  If we want to predict when the education gap in this country will be closed we just need to look at the caliber of our teachers.  If there are under-performing schools, we identify and redistribute high quality teachers.  If there aren’t enough high quality teachers then we need to recruit and educate more of them.  If there are under-performing teachers then we need to pull them out of the classroom for additional training or help them find new careers.  So why isn’t it this simple solution working?  In part because of the  “technological problem” side of the issue.
The technological side brings us back to what we’re doing with the data. The solutions proposed above can only happen after collecting, studying, and interpreting performance data for both teachers and their students. Problems get solved when we are open and transparent about the data. Whether you like or hate No Child Left Behind, you have to give it credit for forcing emphasis on reporting data that attempts to help us learn where we should focus our efforts.  But now we have to take the next step. Taking action requires more than just providing large unintelligible data sets. It also involves visualizing the data in proactive ways that help point to the genesis of the challenges.  Sites like data.ed.gov are just starting this process now, but are nowhere near where we need them to be.  In addition teacher performance data remains unavailable, it’s still too hard to compare data from one state to the next because there isn’t consensus on the data being collected, and it’s too confusing for the average parent to make sense out of the data in order to provide the needed level of accountability.  All of these data-issues obscure the process of identifying and addressing under-performing teachers.  Solving the data issues are relatively simple from a technology standpoint, but unfortunately sharing data about teacher performance becomes dicey from a… you guessed it: political standpoint.  And so in order to answer this question about closing the gap, we must address the three issues simultaneously.
What are some of the most interesting things you have seen out there that deliver social learning as a mechanism for education?
I think one of the most exciting things on the horizon for delivering social learning is educational games. For those of us who grew up in the age of solitary computing (i.e. pre-internet) the term ‘computer-games’ conjures up images of sitting in isolation playing games like Doom or SimCity that only allowed for interaction with a computer.  If you still think this is the current state of computer games, go find a teenager and have them show you how to play World of Warcraft.

What you’ll find is a very social experience where you can’t be successful if you do not effectively engage with a team and share knowledge in order to solve complex problems (see this article by John Seely Brown from Wired).
Imagine the power this type of social interaction could have if applied to an educational context. Projects like EVOKE begin to give a taste of the potential.
And while I’m waiting impatiently for more educational games to be developed, there are many other great examples where creative teachers are already leveraging social learning in the classroom.  Dr. Monica Rankin’s class at the University of Texas at Dallas uses Twitter to increase the amount and quality of interaction among students

Jay Swan, an AP science teacher at DC Public Schools, uses GoogleWave as a platform for student collaboration allowing students to work together in real-time from any location.  Jay can actually observe the increase in collaboration among students as they continue to work together from home at night. Here is a link to the Jay’s assignment in GoogleWave.
If you had the white board for education innovation, where would you start when drafting a systems approach to how education should work? I’m fascinated with how technological devices, for example, may completely revamp our concepts of time needed, subject matter that is relevant, and collaboration vs. individual study, to name a few.
The answer to this question could be the topic of another entire conversation in and of itself.  But to try to give a teaser, here are some of the essential elements that would would want to see in an ideal learning system:

1. Ability to leverage data to customize the learning experience based on student’s needs, past performance, or desired future application of the material.  This includes helping students make decisions in their personal learning profile (where they want to go to college, what they want to do when they grow up, etc)
2. Ability to encourage social interactions with other students and experts related to educational  topics (think adding a Facebook layer to all learning activities)
3. Ability to participate in learning experiences from mobile devices.  This is more than just formatting content to a smaller screen, but actually taking advantage of the affordances of the mobile device (see http://www.youtube.com/watch?v=1SZZgtJKgxk )
4. Ability for students to create and share their own educational media.  A great example of this is a tool called VideoNotes at Carleton University where students can take video feeds from all of their classes and re-mix and share them with each other as study guides (some are very creative)
5. Ability to access a wide repository of educational materials to support the learning.  This piece is already being developed by the Department of Ed (see Steve Midgley’s Learning Registry project on EdReformer)
6. Ability for teachers to grade and provide feedback on any artifact that the student may have created even if it is outside the system (such as a YouTube video or a conversation on Twitter).  The Agilix gradebook provides this functionality today.

These elements would need to be pieced together in a modular or “mash-up” format.  This is different from our current learning systems that try to stuff all possible functionality into a single software package.  The monolithic approach used today limits the variety in the learning experiences since all functionality is determined long before it ever reaches our classrooms.  As we try to squeeze our unique curriculum into the pre-determined mold of a monolithic system we lose some of the uniqueness and creativity that is a hallmark of our educational system.  With a modular approach, all of the tools, data sets, and media can be connected in an infinite number of ways to support a variety of learning experiences – think building with Legos.  All tools work independently, but create a uniquely customized experience when combined in creative ways (more on that here: Learning Customization ).