Sorry, adaptive software is not the same as personalized learning.
We all know that changes in public education move slowly, but there’s one specific educational dilemma we’ve been mired in for decades, with varying levels of rhetoric and hand-wringing: How can we maximize individual student achievement with group instruction?
This is what Education Secretary Arne Duncan was talking about in 2010 when he called for “transformational productivity reforms that can also boost student outcomes.” Over the last century, we’ve put a lot of effort into solving this problem with varying degrees of success.
Today we see it in the hyperbole around personalization, individualization, standards-based grading, differentiation, etc. It seems as if no conference schedule or edtech brochure is complete without some “New and Improved!” way to increase educational return on investment, the latest of which is “adaptive” learning.
This trend promises to deliver differentiated instruction, personalized to each student, at the optimal time, place and pace. The claims seem almost unbelievable. Not that there isn’t truth to the need for closer-to-individual instruction. On the contrary, this is, in fact, where our focus should lie.
However, the sheer volume of buzz overwhelms, so let’s remember where the personalization-revolution began.
Starting with Personalization Basics
Benjamin Bloom (of Bloom’s Taxonomy fame) had an elegant summary of the problem. He found that the most effective model of instruction is, by definition, individualized. Bloom examined conventional, mastery and 1:1 instruction and discovered that not only is 1:1 vastly superior in improving student achievement, but also that the best performance of students elicited by conventional classroom instruction is on par with the worst performance of students in the 1:1 model.
Notice where the Conventional and Tutorial curves intersect below:
Additionally, the average performance in the tutorial condition exceeds the highest performance in the conventional classroom. Bloom proposed that since most students are able to attain this high level of achievement, the mission of educators is to figure out how to provide 1:1-level results with group instruction. Hence the decades-long “how to scale” personalization dilemma.
(Next page: Where adaptive software plays into personalized learning)
Adaptive Software and Personalized Learning
Fortunately as problems go, scale is something we humans have experience solving. Typically, when we have to achieve more with fewer resources, we invent a tool. Want to move more goods but they’re really heavy? Invent the wheel. Need to disseminate information to a broader audience but everyone is far away? Here comes the newspaper. Want to elicit higher student achievement but you only have one teacher for 40 students? Thank you, adaptive software.
“Adaptive” in ed tech is, at its core, a way for schools to provide the right learning experience for each individual student at exactly the right time. Currently, the primary ways adaptive software adapt are with content, assessment, and sequence.
Imagine a program that delivers a pre-test or pinpoints a skill gap, then serves up learning content. While the student consumes content, the software is probing for understanding and remediating in real time.
Products with a legitimate claim to the “adaptive” label have one or more of those functions. It sounds almost magical—who needs teachers when you can buy devices, load up some software and let the machine take over, right?
Why Adaptive Doesn’t Mean Personalized
Here’s how I think about it. When Apple first launched Siri, it seemed as if she was magic. I could say “call Mom” or “get me directions home” or “I feel like Chinese tonight.” The possibilities were limitless. When you break it down, however, Siri is engaging in a very limited set of actions. She voice-enables functions on my phone (calling), interacts with other apps (directions) and finds things on the internet (local Chinese food). Not to downplay the really impressive voice recognition and semantics, but Siri wouldn’t be a replacement for a real personal assistant.
Adaptive software is essentially Siri for the classroom. It can assess, deliver content, and modify a student’s path through curriculum according to a limited set of programmed instructional heuristics. If that were all that happened in a classroom, we’d probably already be in the middle of an Orwellian shift to a computer-controlled Ministry of Education.
What happens when interacting with a good teacher is much richer. So many of the activities we reflexively engage in as teachers aren’t possible to program into software.
How do we get software to encourage a sense of community, model a love of learning, show compassion, teach metacognitive and critical thinking skills, instill respect for others, or foster passion? One of my favorite poets talks about how a teacher “can make a C+ feel like a Congressional Medal of Honor and an A- feel like a slap in the face.” That kind of experience, intuition and good teacher judgement can only happen when we have a personal relationship with our students—the kind of relationship that necessarily develops when a passionate teacher has one student instead of 40.
In today’s world of shrinking budgets, will adaptive software be the panacea the ed tech marketing departments would have us believe it is?
As one describer of the Prussian educational system upon which our public schools are based said, “the masters of our … schools must possess intelligence themselves, in order to be able to awaken it in their pupils.”
Until Siri can talk to my mom, drive me home, and pick up the Chinese food herself, I’m betting on our teachers to get better results than machines.
- In 2024, education will move to adopt AI—but slowly - December 8, 2023
- Mitigating data breaches with live patch management - December 8, 2023
- How video coaching helps us support teacher growth and retention - December 7, 2023