Don't let the Science of Learning be a Trap

Hi! This is Scholastic Alchemy, a twice-weekly blog where I write about education and related topics. Wednesday posts are typically a deep dive into an education topic of my choosing and Fridays usually see me posting a selection of education links and some commentary about each. If Scholastic Alchemy had a thesis, I suppose it would go a little like this: We keep trying to induce educational gold from lead and it keeps not working but we keep on trying. My goal here is to talk about curriculum, instruction, policy, public opinion, and other topics in order to explain why I think we keep failing to produce this magical educational gold. If you find that at all interesting, please consider a paid subscription here, or at the parallel publishing spot on Beehiiv. (Some folks hate the ‘stack, I get it.) That said, all posts are going to remain free for the foreseeable future. Thanks for reading!

Subscribe now

Share

Leave a comment

Constructivism

Today’s post is kind of a spiritual successor to something I wrote way back in 2013 about constructivism. The old blog is long-gone but the gist of that post was that constructivism could be construed as a way to enable some of the worst kinds of education reform. At the time, I was worried by the flatness of Common Core. We’d be required to write a full academic standard on the board and administrators would come by classrooms and ask kids to recite that standard. If they could, we were assumed to be teaching well and the students were assumed to be learning.

These standards are the source of so much that’s gone wrong in schools. They’re basically all skill-based, even with something like figurative language it wasn’t enough to have propositional knowledge but to demonstrate that they had the skill of locating examples of figurative language on a multiple-choice test. Because everything is tested as a single discrete skill, everything is taught as a daily separate discrete skill. Today we learn about figures of speech; tomorrow we learn about nuances in meaning. They are talking about teaching connotation, by the way, but never actually say it because the propositional knowledge is unimportant. Kids don’t need to be able to define connotation, they need to be able to look at a word or phrase and see what it might mean representationally, emotionally, or you know, figuratively.

That’s it. That’s the whole shebang right there. If the kid gets that one question wrong, then it’s evidence that the teacher didn’t teach that one thing effectively. The accountability system we initially designed under NCLB means that kids missing questions equals teachers not teaching. There was no other official, permissible, legal explanation. The challenge for administrators was in determining whether that was true, so a system of statistical observations came into being and offered a practical way to figure out if teachers were adequately teaching to the test. You, the teacher, would be observed at random throughout the school year by a district official. That official’s responsibility was to determine whether or not you were teaching the standards. At that time, there were three components to their checklist. First, teachers needed the exact standard for that day’s lesson written on the board. Second, students in the classroom needed to be able to tell the observer verbatim which standard they were working on that day. Third, because observers could not be bothered to stay and watch the class and see the lesson, we submitted fully written lesson plans for that day but sometimes for several days, and that needed to include the standards fully written out. This was entirely separate from the other observations my school admins conducted to evaluate my teaching. The district was only interested in those standards. Leadership justified this in a practical, pedagogical sense was, of all things, constructivism.

If you’re not super familiar with the major theories of learning, you probably haven’t thought much about constructivism but trust me when I say it looms large in the education world. And, trust me when I say it’s a really poor fit for this kind of test-based atomized instruction. Constructivism is all about how learners are active participants in the building of knowledge. Constructivism cares about the learning environment, social connections between students, teachers, and others in the school, and about how those interface with students’ prior learning. According to constructivism, learners have an understanding of the world, encounter new experiences and information that challenges this understanding, and therefore must accommodate their existing beliefs to the new experiences and information. It’s all a little bit Bayesian, in a sense. If you’ve “updated your priors” then you’re mentally working in a constructivist way. Constructivist views of learning lend themselves naturally to project-based learning, inquiries, and collaboration between peers.

Like I said, it’s a bad fit for standardized education. Like so many other theories, though, constructivism is often bent to serve the needs of reformers, policymakers, and their enforcement mechanisms. When our district laid out the plan to randomly observe classrooms for displays of the standards, kids’ knowledge of the standards themselves, and evidence of the standards explicitly stated in lesson plans, they did so in weirdly constructivist terms. They said that kids needed exposure to the standards — again, the literal word for word text of the standards — in their learning environment because kids construct learning from the environment. If they saw the standard and could remember the standard, then they had prior knowledge to help them learn. Moreover, if kids didn’t have access to the explicit language of the standards, then they would not know what they were learning. Now, I’m pretty sure we all understood that the real purpose here was to use random visits to enforce teaching to the standards and therefore to the test but the district never really said that part out loud.

Okay, so what does this all have to do with The Science of Learning?

Science of Learning

I’m very worried that the Science of Learning folks are walking into a trap that looks an awful lot like what I experienced with constructivism. Their advocacy may ultimately end up as the justification for years of educational malpractice as edTech, policymakers, and cash-strapped districts try to move forward with reforms to bring classrooms up to snuff on the NAEP and promote new accountability mechanisms. This is too bad because the actual insights provided by the Science of Learning are super useful on a day-to-day basis for teachers and as principles for designing and implementing curriculum. I’d hate to see the other major set of learning theories bastardized in the way that constructivism has been.

So, what is the Science of Learning? SoL, as it’s often abbreviated, has its origins in work done by psychologists in the 1960s and 70s about information processing. They were interested in understanding the mechanics of memory and learning and spent time designing experiments to better represent the dynamics of human minds. At the time this was often simplified as talking about the brain as a computer but that understates the usefulness of the basket of concepts that eventually became today’s SoL. We learned from the work of people like Miller, Atkinson, Shiffrin, Baddeley, and Hitch that there are important limits to what human minds can do, and those limits can help us structure our pedagogical efforts to better meet the nature of how we learn. We begin with our sensory processes which take in information. That information (often after some unconscious processing) enters an active and immediate kind of memory. Usually called sensory memory, this is the kind of memory that allows us to quickly access whatever pertinent sensory information is available. Any sensory data that is not pertinent is immediately forgotten. The attention process is powerful and what brings sensory data into working memory, the kind of memory that lets us structure information and make connections with older memories that may be related to whatever task is at hand.

Have you seen the selective memory test? I’ve embedded the video below. You’re going to see a scene of people passing a basketball and your job is to watch and count how many times the people wearing white shirts pass the basketball. Got it? Okay, it’s short. Like two minutes. Watch it.

What did you think? Did you see it? Be honest! Actually, if you’ve seen this before, it’s hard not to see the gorilla but I’ve shown this to students many times and usually nobody sees it. Why? Science of Leaning has taught us the power of attention. You’re so focused on counting and following just the white-shirted players that your brain’s attention economy is stressed. What does it do? It cuts back on the amount of extraneous sensory data that passes through to your working memory. The players in black shirts aren’t important to the task, so you brain quite literally edits them out of your working memory. Obviously they’re still visible but your attention is a powerful signal to the unconscious processes of your brain to ignore the black shirts. When a gorilla walks into frame, your brain lumps it in with the players wearing black shirts. When he beats his chest, your brain groups him with the passes made by the black shirts. (The reason we pretty much always see it every subsequent watch is because we remember being tricked so our brains make sure to create what’s called an episodic memory, another SoL concept!)

It’s not just our sensory systems that get paired down by the need to focus attention. When we’re working on tasks in working memory, there are limits to how much we can keep track of in working memory at once (this is, we think, what the selective attention filter evolved to do). Cognitive load theory tries to figure out how much our working memory can do before there are negative effects such as, slowed task completion, increased error rates, and frustration. When this happens, it’s also much harder for the stuff in your working memory to end up in your long-term memory. Ideas and skills that we learn need maintenance in the sense that we need to come back to those ideas and use those skills again and again. The more we use them, the more strongly they become encoded in our long-term memory and the easier it is for us to retrieve those ideas or skills for use in the working memory. Background knowledge or prior knowledge are functions of an effective long-term memory which is, in turn, a function of using that knowledge repeatedly. Under high cognitive load this whole process is less effective, hence efforts to design lessons to reduce cognitive load to just the important things being taught. It’s also a good reason to get rid of classroom tech and, especially, phones.

Anyway, just like with constructivism, there’s a lot more to SoL than I can fit here. I haven’t even gotten to executive functioning which is a critical component of the theory. The important thing to know for this post is that SoL is great and super useful for educators to know. If you are looking for some good guidance in this regard, I highly recommend Kirschner & Hendrick’s How Learning Happens: Seminal works in educational psychology and what they mean in practice. There’s also a Science of Learning Substack if that’s your kind of thing. You’ll notice that proponents of the Science of Learning are not usually recommending inquiry-based learning or projects or lots of collaborative learning. Instead, SoL is most commonly associated with instructional methods like explicit instruction, memorization, spaced retrieval drills, and other similar approaches. This is often represented as putting SoL at odds with constructivist approaches, I don’t think that’s really true in an absolute sense and they are probably more complimentary than either side of the debate wants to admit.

It’s A Trap?

But, that brings me to the trap. Because the Science of Learning is seen by laymen as the opposite of progressive educational practices, it’s being taken up by policy advocates as a way to replace and exclude practices based on other theories of learning. It is becoming politicized. I don’t think this is necessarily what SoL is teaching us about the mind or about learning. Centering students’ cognitive architectures doesn’t mean we can’t also build a collaborative learning environment. Spaced retrieval in order to build better connections between working memory and long-term memory would support a subsequent project-based assignment where students use what they’ve learned as background knowledge to solve a novel and challenging problem. We’re seeing this play out right now with reading comprehension. As the focus of reading instruction has moved toward phonics, we’ve embraced more practices associated with SoL to support the memory work being done with those foundational skills. At the same time, reading comprehension is languishing and we’re rediscovering the need for knowledge-building in order to support comprehension. This calls for a blend of approaches based on both constructivism and information processing but because constructivism is seen as an oppositional stance, it’s being sidelined.

Let’s say you want to have a Socratic seminar where students spend time discussing some set of topics relevant to your curriculum. SoL would remind us to lay the groundwork for supporting this discussion. Instead of merely assigning a reading and expecting kids to discuss intelligently, we should explicitly break down the reading and make sure students have a clear understanding of what they’ve read. Then they have the knowledge at hand for discussing and the discussion can also function as retrieval practice to help them keep the info in long-term memory. Anyway, it’s all complimentary! There’s no opposition here. So why am I worried? Let’s use what we’ve learned from SoL today to build a metaphor.

SoL is watching the ball being passed and they’re going to miss the gorilla when he enters the picture. That is, SoL is laser focused on improving teacher practices and they’re missing the ways in which their approaches are being hard coded into curriculum. As I’ve pointed out before, standardization and accountability have moved away from being purely test-based and outcome-based to being incorporated with curriculum.

A good example of this comes from the Houston Independent School District. HISD was taken over by the state of Texas a few years back and they appointed Mike Miles as superintendent. One of the big changes he’s made is what he calls his trademark New Education System, NES. NES requires scripted explicit instruction, daily timed high stakes quizzes, and a standardized scheduled curriculum across all schools. The curriculum is built to do something very familiar, teach to the test. This is all justified because Miles says he is adhering to the science of learning and applying it through the curriculum and to teacher practice. This level of control was never envisioned by SoL advocates who have long contended they are focused on improving teachers’ practices so they can better meet students’ learning needs.

As curriculum continues to be the focus of reformers, I worry we’ll see more and more standardization of the SoL and could, in turn, see the flexibility and innovation at the heart of SoL get lost in a sea of “fidelity” requirements. My experiences with last generation’s accountability reforms saw learning theories used as justification to force teachers to teach to the test. Those scores meant everything. This time around, rather than caring about standards, they’re going to come in and check if the teacher did enough spaced retrieval practice or if they attended to cognitive load by following the scripted lesson. Instead of asking about standards on the board, kids will be asked what prior knowledge they activated at the outset of the lesson. And a decade later, SoL advocates will be left trying to explain why these reforms weren’t really SoL and why everything was distorted. My hope, though, is that someone who advocates for the SoL reads this and sees the gorilla. Maybe we can save these invaluable practices from being flattened in the service of test scores and turned into highly standardized curricula where they’re used as teacher compliance measures instead of actual pedagogy.

Thanks for reading!