This post first appeared on Educating Modern Learners
“You keep using that word,” says Inigo Montoya. “I do not think it means what you think it means.”
I always think of this line from the 1987 classic movie The Princess Bride when I hear talk of “personalized learning.” On one hand, "personalized learning" sounds pretty good: a nod towards more student-centered learning perhaps, a move that honors the person learning not just the learning institution. But on the other hand, I do not think it means what you think it means. Often, what I see the term applied to gives me pause – “personalized learning” appears to be more focused on the scripting than on the student. Personalized learning isn't personal learning.
What do we mean by “personalization” in education and in ed-tech?
That’s the $1 million question, heck, the multimillion dollar question. We’re seeing hundreds of millions of dollars funneled into education/technology companies that promise tools to “personalize” and “customize” education for individual students, “adapting” to their individual skills and needs.
“Personalization” can be used to describe a variety of educational programs – different instructional approaches and different software and services. For many the term has become a rallying cry to challenge the “one size fits all” model of the traditional education system. Fair enough.
But “personalization” doesn’t (necessarily) seem to challenge the goals or the curriculum of traditional education. Instead, it promises to move students towards those goals more effectively, more efficiently.
The promise of “personalization” is closely tied to the promise of adaptive learning technology. Thanks to the data that students generate through their computer-mediated work, we’ll be able to better track each individual student’s progress. Moreover, with their data in aggregate, we’ll be able to design algorithms and from there software to deliver content aimed at each individual student’s appropriate skill level.
Thanks to data and to technology, we’ll finally be able to understand what each individual student knows, what each individual student lacks, what each individual students wants, and so on. Something – if you believe the marketing – we’ve never been able to do before.
And, ah, the marketing.
How “Personalization” Gets Sold to Schools
This is one of my favorite slides from a venture capitalist highlighting the transformations that technologies will break to the public school system:
It comes from NewSchool Venture Fund’s “The Future of K–12 Education” slideshow, and it features their investment portfolio company No Red Ink which promises personalized digital grammar worksheets.
Here the investment firm praises products that seek to “reimagine engagement through personalized content” with “problems based on personal interest and real-world applications.” But look at the example. How many kids in middle school today struggling with punctuation would feel so much closer to a “deep learning” of commas because of a reference to John Mellencamp, who had his breakout hit “Hurt So Good” in 1982, long before any kid in K–12 school today was born? How is this personalized?
Here’s another promise — one that’s less about “the past” and more about “the future”:
“If you learn concept No. 513 best in the morning between 8:20 and 9:35 with 80 percent text and 20 percent rich media and no more than 32 minutes at a time, well, then the odds are you’re going to learn every one of 12 highly correlated concepts best that same way.”
That’s a quote from Knewton CEO Jose Ferreira. Knewton, initially a test prep company, now is one of the leading promoters and producers of “adaptive learning” technology — the idea that data collection and data analysis at scale will lead to “personalization.”
Knewton has partnered with most of the major textbook publishers — Pearson (which is also an investor in the company), Cengage, Macmillan, Wiley, and so on — to deliver textbook content and assessments through its platform.
Knewton creates an inventory of all the concepts and skills in a particular textbook or course, then by analyzing a student’s clicks and running these through its proprietary algorithms, Knewton then recommends what the student should click on next — “learn” is the verb the company uses — to maximize her or his speed and success in moving through the course material.
"Depending on your learning style, it might be best to introduce linear equations through a visual, geometric approach, where you plot the lines and show the intersection.” That’s a quote from Knewton’s VP of research.
Knewton has raised over $105 million in venture capital. There’s no shortage of cash there to hire some of the leading minds in cognitive science research. So it’s a bit disconcerting to see the company’s VP of research point to “your learning style” here. While “learning styles” are a popular concept, there’s no scientific evidence supporting the idea.
Nevertheless, Knewton’s CEO recently wrote a blog post doubling down on this idea. “To me,” Ferreira insisted, ”it’s pretty obvious that different learning styles exist.”
What’s New About “Personalization”: Big Data
The phrase "To me it’s pretty obvious” does actually not reflect how science works, despite the fact that what Knewton and others promise is that “personalization” via adaptive technologies is based on science. What “personalization” is based upon is data to be sure. Knewton says it collects more than a million data points for each student over the course of a semester. These data points are used to build and refine its algorithms, and in turn this data is used to identify how well students fit into certain models, certain profiles.
The content remains standardized. The goals remain standardized. (These are goals of the educator or administrator typically — not the goals of the student.) The metrics remain standardized. “Personalization” is a simply a matter of how the technology optimizes a students path through the standardization.
Personal Learning versus Personalized Learning
That’s never what educators, specifically progressive educators, have meant when we have demanded that education be more personal, more student-centered. Personalization, framed in this regard, means empowering students to pursue their own interests, to follow their curiosity, to build their knowledge and carve their learning paths — to do so not through computer algorithms but through human agency.
Personalization, when it comes to something like “universal design for learning” for example — a concept that comes out of universities’ education departments and not out of corporate marketing departments, recognizes the need for flexibility when it comes to how we approach each student. UDL means expanding and accommodating students — recognizing differences, honoring differences. It doesn’t mean matching students to different profiles and funneling them through pre-ordained and standardized processes. It means seeing students as different people and supporting them as such.
What are the repercussions of radically “personalizing” education through computers? What do we gain? What do we lose?
There’s a very powerful strain of American individualism — and California exceptionalism — that permeates technology: an emphasis on personal responsibility, self-management, autonomy. All that sounds great when and if you frame new technologies in terms of self-directed learning.
But how do we reconcile that individualism with the social and political and community development that schools are also supposed to support? How do we address these strains of individualism and increasingly libertarianism as they permeate the classroom?
What do we do about the communal goals of education, for example — to produce good citizens, if nothing else — if we become maniacally focused on personal goals of education instead? What happens to meaningful moments to collaborate? What happens to discussion? What happens to debate? What happens to the idea that we must work through ideas together — not just in the classroom, but as part of our work and civic responsibilities?
And who gets the “personalized” education delivered through them via adaptive technology? And who gets the “personalization” that we hope a student-centered, progressive education would offer?
This image from a PBS documentary about Rocketship Education haunts me.
The chain of charter schools boasts personalization — "Rocketship uses the most adaptive and personalized programs available, and continues to push Silicon Valley vendors and others to create even more adaptive learning tools,” its website boasts.
So the problem with personalization via adaptive software isn’t simply that “it doesn’t work.” It’s that it might work — work to obliterate meaningful and powerful opportunities for civics, for connection, for community. Work to obliterate agency for students. And work not so much to accelerate learning, but to accelerate educational inequalities.
Image credits: Sarah