Self-Transcript for Road to Reading Podcast #23: Discussing Cognitive Science

[~0:00] What is the science of learning? [~7:00] Students learn better when they're actively solving problems and explicitly being told how to solve them. [~13:00] Students retain information longer when they space out their review with expanding intervals. [~19:00] Spaced repetition is so similar to weightlifting that you might as well call it "wait"-lifting. The wait creates the weight. [~22:00] Desirable difficulties: making the task harder in a way that overcoming the difficulty produces more learning -- but not all difficulties are desirable, and no difficulty is desirable if the student is unable to overcome it in a timely manner. Other desirable difficulties include interleaving (mixed practice) and the testing effect (retrieval practice). [~32:00] The testing effect (retrieval practice effect): students retain information longer when they're made to practice retrieving it from memory. Again, it's just like weightlifting. The way to build long-term memory is to use long-term memory. You're picking up a weight off of the ground of long-term memory and lifting it up into working memory. [~36:00] The power of automaticity, the ability to execute low-level actions without them exhausting your mental bandwidth. It's important to develop automaticity because we all have limited working memory capacity. Automaticity helps us overcome that limit. [~44:00] Automaticity is a critical component of creativity. It frees up space for creative thinking. [~48:00] The expertise reversal effect: the difficulty of the task needs to be calibrated to the ability of the learner. If expert-level tasks are given to non-experts (or vice versa), little learning will occur. [~55:00] Why it's important to transition from massed/blocked practice (repeating the same exercise consecutively) to interleaving (mixing/varying up the exercises). [~1:02:00] Effective learning strategies can feel counterintuitive / unnatural because the point is to increase effort, not to reduce effort. It's completely different from typical work or chores that you might do in batch. It's completely different from reading a fluent story from start to finish. It's about interrupting the flow of thought and coming back to it later. [~1:09:00] Deliberate practice: a high-level description of the most effective form of practice identified by the academic field of talent development. [~1:15:00] To what extent does the accumulated volume of deliberate practice predict whether someone is going to become an expert? Deliberate practice is the primary factor, but genetics is an important secondary factor. [~1:17:00] NON-examples of deliberate practice. Common pitfalls when people try and fail to do deliberate practice, and how to avoid them. [~1:23:00] How to learn more about the science of learning. [~1:29:00] The #1 takeaway: use interleaved spaced retrieval practice. You can use this in the classroom.

Cross-posted from here.

Want to get notified about new posts? Join the mailing list and follow on X/Twitter.

Link to Podcast

The transcript below is provided with the following caveats:

1. There may be occasional typos and light rephrasings. Typos can be introduced by process of converting audio to a raw word-for-word transcript, and light rephrasings can be introduced by the process of smoothing out natural speech patterns to be more readable via text.
2. The transcript has been filtered to include my responses only. I do not wish to infringe on another speaker's content or quote them with the possibility of occasional typos and light rephrasings.

Justin: Thank you, Elana. I’m excited to be here.

Justin: Sure. The science of learning is essentially the study of how to increase people’s performance on cognitive tasks that require them to use their brainpower to accomplish something.

The idea is to break it down very specifically—what mechanisms are at play inside students’ brains, how that affects their ability to accomplish a task, and how they retain that learning into the future.

At a nuts-and-bolts level, the goal is to understand how these processes work and use teaching strategies that align with them to optimize how quickly students acquire knowledge and how long they retain it. When done correctly, this can lead to massive gains in student learning and help them reach their potential.

Justin: Totally. Before I get into that, one thing I’d like to say is that it’s completely normal to come out of a teacher credentialing program without having learned these things. It’s kind of weird to think they’d be covered in a credentialing program, but I actually went through one myself.

How many times did I hear the phrases spaced review, spacing effect, interleaving, mixed practice, or even cognitive load? Not at all. This research has been around for decades, but it just hasn’t really made its way into the credentialing process.

If any listeners are feeling bad that they haven’t heard about these things before, there’s no reason to worry. Unfortunately, teachers usually discover the science on their own by going into the literature. If this is all new to any listeners, just know that this is step one in the journey to becoming acquainted with the science of learning.

Justin: Exactly.

Let’s go through that list. As you mentioned, I recently posted about a set of cognitive psychology findings that are well-established and can help students learn better.

Before I get into them, I should note that many findings in psychology haven’t held up. That’s a valid critique of the field, but it doesn’t apply to everything. Some findings have been reproduced over and over again for decades. I don’t want to throw the baby out with the bathwater. There are a number of these principles that are extremely solid.

Let’s start with the most obvious and move to the least obvious. First, something everyone already knows intuitively—students learn better when they are actively solving problems, retrieving information, or doing something rather than just watching a video or rereading notes.

This has been tested scientifically numerous times and is completely indisputable. There are several papers in academic literature showing that the standard approach in higher education—where a professor lectures for an hour or more while students passively listen—is ineffective. Students learn next to nothing during the lecture, and the real learning happens later during independent work.

It would be far better to combine those two things in the classroom: a minimal dose of instruction followed by problem-solving, then regrouping for another minimal dose of instruction.

This is one of the cognitive psychology findings that holds up—active learning beats passive learning. It has been reproduced in a countless number of papers.

That doesn’t mean students should never watch or listen to the teacher. It just means they should start actively solving problems as soon as possible after a minimal, effective dose of instruction.

Justin: That’s a great point. Unfortunately, it often gets tangled up in discussions about active vs. passive learning and direct vs. unguided instruction. These are two completely separate levers.

One lever is active vs. passive learning. Another is direct (explicit) vs. unguided instruction. These are two independent variables, but people often assume something about one when discussing the other, which isn’t necessarily true.

The big picture, as you said, is that the optimal mode of instruction is both active and direct. Students listen to the teacher for a demonstration or explanation, but as soon as that’s conveyed, they go straight into active practice.

What further complicates the issue is that many studies vary both of these variables at the same time. For instance, they might compare active unguided learning to passive guided learning. The conclusion often shows that students in the active group learned more.

Sometimes, people mistakenly take this to mean that not only is active learning better than passive learning, but also that unguided instruction is better than explicit instruction. That’s not true at all. The active vs. passive variable has a larger effect. If you take a large effect size and pair it with a smaller effect size, the larger effect still dominates.

If a proper study included active guided learning as another condition, that would come out ahead. Unfortunately, these nuances often get lost, especially when people are reading studies and drawing incorrect conclusions.

Do you want to move on to another finding?

Justin: Sure.

Another finding—again, starting with the obvious before getting more complex—is that if you don’t review and reinforce knowledge, you forget it. That’s completely obvious to anyone who has tried to remember vocabulary, facts, or anything else.

The key point is that forgetting can actually be modeled mathematically. There’s a forgetting curve that can be expressed as a mathematical function fitting the pattern of how much someone remembers over time.

A lot of research has been done on this, and forgetting behaves almost like a law of physics. The only real difference is that at this cognitive level, we deal with more noise, random perturbations, and variability. But you can still measure these effects statistically.

Now, active learning is important. If you don’t review material, that’s pretty obvious. But there are many other strategies that are less obvious and often not put into practice. They are just as well-supported and function like laws of physics, just like the principles we’ve discussed so far.

The first thing I’d highlight is the spacing effect. This concept goes back to the psychologist Hermann Ebbinghaus, who conducted experiments on it in the 19th century. It’s been around for over 100 years, yet I didn’t hear about it in my teacher credentialing program. I’m sure most people who have gone through a credentialing program didn’t hear about it either.

That’s strange because this is one of the most fundamental findings in the science of learning, and it’s not difficult to understand. The basic idea is that if you space out your practice instead of doing it all at once—rather than cramming—you will retain more, even if the total amount of practice time is the same.

For instance, instead of doing two hours of practice in one day, if you split that into six sessions of 20 minutes each, spread over several days, you will retain significantly more. One of my favorite researchers, Doug Rohrer, has conducted extensive studies on the spacing effect and other cognitive learning strategies.

To quote one of his papers: “The spacing effect is arguably one of the largest and most robust findings in learning research, and it appears to have few constraints.” I don’t know how much more of an endorsement a strategy could get.

The spacing effect has profound consequences. When you space out your review, retention increases. If you structure your study sessions—say, a 20-minute session today, another after a couple of days, another after a few more days—your memory stabilizes with each interval.

This means you don’t have to stack review sessions back-to-back. It’s actually more effective to space them out with expanding intervals. For example, you might review for 20 minutes today, 20 minutes tomorrow, then wait a few days, review again, wait a week, review again, then maybe wait three weeks before the next review.

This approach allows two hours of total review time to stretch over months or even a year, keeping the material in your brain far longer than if you crammed it all into one day and forgot most of it within a few days.

The idea might initially seem a little counterintuitive—why does it work that way?

What’s really going on under the hood in the brain is that it’s about the effort required to retrieve the memory. Whenever you retrieve a fuzzy memory, that strengthens it. You don’t get stronger from lifting weights that are easy for you to lift; you get stronger by challenging yourself with heavier weights.

In your brain, the analogy is that the difficulty of retrieval is like the weight being lifted. If a memory is easy to recall, that’s great, but it doesn’t extend retention very much. However, if retrieval is a little harder and you still manage to recall it, the memory becomes more ingrained and lasts longer. It’s like building muscle in the brain.

Justin: Sure.

Justin: This comes down to the question of what a teacher should do when a student is unable to retrieve information.

Think of it as the student trying to lift a weight. The teacher is the spotter, ensuring the student lifts the weight on their own but stepping in when absolutely necessary. The goal is for the student to retrieve the information successfully.

If a student is asked a question but can’t recall the answer, the teacher can provide small hints or reminders rather than giving away the answer. See if they can retrieve it with just a little help.

It’s like someone struggling under a bench press at the gym. The spotter doesn’t lift the weight for them but gives just enough of a push to help them complete the lift. That’s how a teacher should approach retrieval support—just enough assistance to help the student succeed without doing the work for them.

I’d like to zoom out to a more general level because this connects to something in the science of learning research called desirable difficulties.

This is one of those terms that has been around for many decades. Bjork and Bjork introduced it long ago. The idea is that you increase retention by introducing difficulties that make learning and recall more challenging for the student in a productive way—one that supports them and helps develop their mental “muscles” for recall.

A key example of desirable difficulties is spacing out retrieval—letting memory get a little fuzzy before recalling it. Another example, which we’ll discuss later, is mixing up problem types instead of doing the same type over and over. These strategies make tasks harder to complete but, if the student overcomes the struggle, they gain more from it. It increases the integration of knowledge into the student’s brain.

One thing that often gets overlooked about desirable difficulties is that if a student cannot successfully complete the task, the difficulty is not desirable. If you wait too long for memory to decay and the student can no longer recall the information, it’s not productive.

Back to the gym analogy—imagine someone trying to lift a 500-pound weight when they’ve only ever lifted 100 pounds before. If they can’t lift it at all, it’s useless. It’s like doing a workout where you just push against a wall for an hour—it doesn’t accomplish anything. The key is overcoming the difficulty.

This is often overlooked in implementations of productive struggle. If a student struggles but never actually overcomes the challenge, they don’t gain anything. Or, if it takes them so long to succeed that they only manage to solve one problem in an entire class, the learning is inefficient.

That would be like going to the gym, straining to lift a weight one time, putting it down, and calling it a day—not very effective.

Justin: Yes. These are all strategies introduced gradually into the learning process to ramp up the difficulty until the student can complete a task in the most challenging setting.

The testing effect is somewhat parallel to these strategies. The way to think about it is that these are different levers you can pull to increase the difficulty of a task.

You can introduce a delay before completing the task. You can mix up problem types so the student can’t easily predict what’s coming. You can require them to retrieve the information entirely from memory instead of referring to a worked example or notes.

Initially, all these levers are set at zero. When none are engaged, the task is much easier. This is necessary at the beginning of the learning process.

You start by explaining something, then have the student apply it immediately. If you wait too long after the explanation, they haven’t had enough practice yet and will forget, making the task unnecessarily difficult.

Similarly, at first, students should practice a specific problem type multiple times in a row rather than jumping between different types immediately. This helps them overcome initial struggles in execution.

This is also why you might provide a worked example or allow students to use notes when completing a task for the first time. It scaffolds their experience so they can overcome difficulties and successfully complete the task.

Over time, you start pushing those levers forward, making the learning experience progressively more challenging.

Now, this doesn’t mean you teach a lesson, assign homework, and the next day give students a timed pop quiz with no notes and completely interleaved problems. Or wait a month and then test them. That’s too extreme.

The goal is to periodically engage students in learning experiences that become slightly more difficult each time.

On the first day, you go through a short lesson, complete some learning tasks, and allow students to use notes. Maybe the next day, you set up a similar task but after a short delay so their memory is slightly fuzzy. You might tell them, Instead of looking back at your notes immediately, try to recall this from memory. If you can’t, that’s okay—check your notes if needed. The goal is to introduce a bit more challenge.

A few days later, instead of reviewing the same topic again immediately, you wait a week or so before reintroducing it. The student then tries to recall it again. You’re gradually ramping up the difficulty, but you don’t go from zero to 100.

Justin: Yes. We can go into more detail on the findings we’ve discussed so far.

We’ve taken a broad, high-level look at things like the testing effect and mixed review. Let’s go through each one independently.

The idea behind the testing effect or retrieval practice effect is that instead of looking back at your notes, you retrieve information from memory. The way to increase retention and build long-term memory is to use long-term memory.

There’s a specific way you need to load information into your working memory to improve retention. If you load information just by looking at notes, you’ll be able to use it temporarily, but you won’t remember it long-term. To retain it, you must pull it from long-term memory.

The testing effect is simply the idea that when you test your ability to recall information—retrieving it from your head instead of looking at a paper—that process strengthens retention.

I’d like to point out that this analogy to weightlifting runs deep into the mechanics of what’s happening in the brain.

To break it down, the retrieval process is about pulling information from long-term memory into working memory. That cycle—retrieving from long-term memory into working memory—is the key loop to think about when learning.

Initially, information is outside your brain. You read something on paper or hear an explanation from a teacher, bringing that information into working memory. At this stage, it’s not in long-term memory yet.

What you need to do next is practice using the information in working memory while continually trying to retrieve as much as possible from what little has been seeded into long-term memory. You keep pulling it from long-term memory into working memory repeatedly.

By doing this, you strengthen long-term memory like a muscle. The way I think about it is that retrieving information is like picking up a weight from the ground—lifting it from long-term memory into working memory.

When you look at notes or listen to a teacher’s explanation, it’s like having the weight handed to you. You haven’t actually lifted it yet. The act of lifting—retrieving the memory—is what makes you stronger.

Justin: That’s a great point about fluency. It’s an important topic for us to discuss.

Fluency and automaticity occur when something is so ingrained in long-term memory that the effort required to recall it is imperceptible. It’s like walking. When you’re first learning to walk, you have to focus entirely on it, or you’ll fall. But as you develop the skill, it becomes so ingrained that you can think about other things and perform other tasks simultaneously while walking.

In learning, you want to develop skills to a similar level to enjoy the same benefits.

In math, for instance, there’s a huge advantage to being fully automatic with multiplication tables. When learning more advanced topics—solving equations, factoring quadratics—you need to recall multiplication facts instantly while keeping your train of thought on the problem at hand. You want to be able to “chew gum and walk at the same time.” You don’t want to have to consciously think about moving each leg while talking to your friends.

The same applies to reading. If a student has to pause at every word to look up its definition, it interrupts the flow of thought. Without automaticity, the brain gets hijacked by low-level processes, making high-level thinking impossible.

To use another analogy, consider a basketball player dribbling. They’re moving, putting the ball between their legs, and evading defenders—all while strategizing and keeping track of their teammates’ positions. If they had to consciously focus on dribbling, they wouldn’t be able to engage in high-level strategic thinking. Learning works the same way.

Going deeper into cognitive science, the reason automaticity in foundational skills is so valuable is that everyone has a limited working memory capacity.

Think about what happens when you try to do too many things at once—having a conversation while texting and listening to music. You get overloaded and struggle to complete any of the tasks. It’s like a computer freezing when too many processes run at once. That’s the experience of working memory being overwhelmed.

Everyone has a limit to how much can be stored in working memory. At a physical level in the brain, this comes down to how long neurons can maintain activation. Every concept, word definition, times table, and mental representation corresponds to a specific neural activity pattern.

There’s a hard limit to how much neural activity can be sustained simultaneously when all of it requires conscious effort.

As you learn and information gets encoded into long-term memory, neural patterns become connected through synaptic pathways between brain cells. This makes it easier to activate these patterns with minimal effort.

It’s like toppling a row of dominoes. Instead of manually pushing each of 100 dominoes, you just tap the first one, and the whole stack falls. Once these patterns are established, retrieving information requires only a small effort to activate the entire sequence.

Initially, when information is not yet encoded in long-term memory, you must constantly reactivate all the neurons in the pattern, like juggling multiple objects. There is a limit to how much effort you can devote to that process.

Automaticity allows you to retrieve information with almost no effort because it’s already hardwired into long-term memory. This means you can pull information into working memory without it adding to the cognitive load of what you’re juggling.

This is why automaticity supports creativity. A common misconception is that practicing fundamental skills to the point of automaticity makes students robotic or unable to think critically. In reality, automaticity frees up mental processing power for higher-level strategizing and problem-solving.

By becoming fluent in foundational skills, students can focus their working memory capacity on complex thought processes instead of being bogged down by low-level tasks.

Justin: Exactly. It’s all about freeing up space for deeper processing.

If a student is overwhelmed by a task and isn’t thinking critically about what you want them to, it’s often not because they’re choosing not to engage. Instead, they’re using all their working memory capacity on low-level tasks they haven’t yet made automatic.

For teachers, if a student struggles to engage deeply with a concept, it often indicates they need more practice at a lower-level skill that should be supporting the higher-level task.

Justin: Absolutely. Go for it.

Justin: Sure. Let’s talk about deliberate practice last since it’s less directly related to what we’ve discussed so far.

The expertise reversal effect, on the other hand, flows naturally from our discussion.

As you mentioned, it’s tempting to ask students to do things that you as an expert would find to be a good learning experience. But that doesn’t always translate into a good learning experience for a student, especially if the task is too difficult at their current level.

The expertise reversal effect describes a general phenomenon where a learning technique or task is highly effective for experts—who have already built a strong foundation of knowledge and skills—but does little for beginners. If you give the same task to a novice, they often learn very little or nothing at all.

For example, if an expert reader is asked to read a passage from Shakespeare and interpret it, explaining how it fits with other Shakespearean works, that task is beneficial. The expert already knows how to read fluently, is familiar with Shakespeare’s themes, and understands how to analyze texts critically. This type of assignment helps them develop further because it integrates multiple skills and gives them practice in high-level synthesis.

This is similar to a gymnast performing a triple flip. If they’ve already mastered basic gymnastic prerequisites, this challenge can help them progress.

But if a beginner is struggling to recognize words, understand definitions, or decode challenging letter patterns, asking them to engage in a complex literary analysis is like asking someone who has never done a backflip to attempt a triple flip.

If they can’t even perform the fundamental movements yet, the high-level task does nothing for them. They won’t progress because they don’t have the foundational skills in place. That’s the essence of the expertise reversal effect.

This is just one direction of it. It also goes the other way. There are a lot of things you might ask a lower-level student to do that are not very effective practice for a higher-level student who has built more knowledge or are not at the appropriate level of difficulty.

For instance, you could imagine a student learning a high level of algebra or calculus. They’ve already solidly learned other fundamentals. They know how to do addition. It wouldn’t really make a whole lot of sense to have them work out addition problems on a number line and draw up their solutions or things like that.

The whole idea is that these tasks and types of learning that work for some students at one end of the knowledge spectrum do not necessarily work for students at the other end. Often, it’s the opposite—they just don’t work.

I guess this links back to what we’ve been talking about—desirable difficulties, scaffolding, and how it’s all part of a journey. It’s part of the process where you are gradually giving students more challenging things to work out but never overwhelming them.

A lot of people make the mistake of thinking, Oh, let’s give them the hardest exercise possible because that will produce the most learning. But it won’t produce the most learning because they can’t actually do it, so it doesn’t produce any learning.

Justin: I guess there are a couple more that we can go into.

Another type of cognitive learning strategy, a desirable difficulty that we’ve mentioned by name but haven’t really discussed in depth, is mixed practice or interleaving.

The easiest way to understand mixed practice or interleaving is by first considering its opposite—blocked practice. Blocked practice is when you repeat the same task over and over again.

In math, this might mean solving the same type of equation repeatedly. In an athletic setting, like basketball, it might mean taking three-point shots from the same spot on the court over and over. This kind of repetition allows students to settle into a robotic rhythm, where they apply a solution pattern without fully understanding or planning their approach.

At the beginning of the learning process, blocked practice can be useful because it makes the task artificially easy. Since tasks need to be calibrated to a student’s skill level, this artificial ease helps get them over the initial hurdle of completing the task. As students improve and gain confidence, you want to remove these artificial scaffolds and introduce real challenges.

One way to increase difficulty is to vary the practice conditions. Instead of having a student solve the same type of equation repeatedly, you might introduce different types of problems that require slightly different solution techniques. Instead of having a basketball player shoot from the same spot, you might have them move to different locations, forcing them to recalibrate their shot each time.

Interleaving exists on a spectrum.

At first, students might practice problems that are only slightly different, requiring small adjustments. As they succeed and become more comfortable, the difficulty increases. Eventually, instead of alternating between two slightly different quadratic equations, you might introduce an exponential equation or another problem type that wasn’t explicitly introduced beforehand.

Or the basketball player, you say, Shoot a three-pointer. Then go in for a layup. After that, come out the other side, and I’ll pass you the ball right away. You have to catch it and shoot another shot immediately.

You’re trying to interrupt their rhythm because they have reached a level where they should be able to recover from those interruptions. This is more like an actual game scenario. The reason for interrupting their train of thought with interleaving is to make the learning task as close as possible to a real-world scenario where they must recall information completely from scratch.

You want them in a situation where they haven’t spun up everything in their head. They don’t have all the context loaded at the front of their mind. Eventually, you want them to reach a level where they can be reading a passage, and if you suddenly ask them, What’s six times seven?, they immediately say 42, then return to reading without breaking their flow of thought.

You want them to develop automaticity. The ability to pull a skill from scratch is what interleaving helps scaffold.

Interleaving and mixed practice are synonyms.

Justin: No problem. That’s a great point. These are all great examples of how we might increase the challenge by having students perform a variety of tasks and retrieve each one from memory.

One interesting thing to build on is that this process can sometimes be counterintuitive because it differs from how good educational content is typically created.

What I mean is that when you’re designing a workbook or writing math problems, you usually create content in batches—writing problems or passages that exercise the same skill repeatedly. This minimizes the effort required to produce the material. However, if you want to present this content in an interleaved way, you actually have to mix up the tasks you’ve created.

A lot of effective learning strategies feel unnatural compared to how we usually complete tasks. When creating educational content, doing chores, or even washing dishes, we tend to batch similar tasks together to minimize effort. But for learners, we aim to do the opposite—we deliberately increase the effort required.

There are just a couple of other strategies you mentioned. I think deliberate practice was one of them. Deliberate practice operates at a higher level than the strategies we’ve discussed so far.

One last point about interleaved learning before we move on to strategies like deliberate practice—interleaving can be applied at multiple levels of scale and is fundamentally about spaced repetition.

By different levels of scale, I mean that on one hand, you might interleave problem types within a single practice session. But you can also interleave at a higher level, mixing new topics a student is learning. For example, in an arithmetic curriculum covering addition, multiplication, and division, you don’t have to complete all of addition before moving to multiplication and then all of multiplication before introducing division.

That would be almost like blocked practice, where you’re doing the same type of problem repeatedly—addition, addition, addition. Even if you’re practicing slightly different skills, such as adding negatives or fractions, it’s still all addition. To some extent, the knowledge of addition remains loaded in memory. It would actually be better to interleave these topics—do some addition, then some subtraction, then multiplication, then division.

Textbooks are often organized into units where all the problems focus on the same concept. There might be a whole unit on addition, a whole unit on subtraction, a whole unit on multiplication, and a whole unit on division. That doesn’t mean a student should progress linearly through each unit, completing one entirely before moving to the next. It’s actually better to interleave between these topics—do some addition, then some subtraction, then apply that knowledge to multiplication, then use that understanding for division.

This approach is often called a spiral curriculum. Not sure if you’ve heard the term before. A spiral curriculum is a loose approximation of spaced repetition. While it doesn’t necessarily follow expanding intervals of retention, it does leverage some spaced review.

This is an example of how interleaving applies at multiple levels—not just within a single task but across an entire syllabus. The daily schedule of a course can be structured to take advantage of spaced review. Interleaving pairs well with spaced repetition. If you introduce a topic now and revisit it after a few days for review, you need to fill the time with something else. Over the next few days, you work on different topics. When you switch from one topic to another, you’re not just spacing practice—you’re also interleaving.

It’s really hard to do spacing without interleaving or interleaving without spacing. In practice, they turn out to be very similar.

The last cognitive learning strategy we can talk about is deliberate practice. I wouldn’t really call it a cognitive learning strategy as much as a high-level description of what successful practice looks like.

Deliberate practice has been researched in the academic field of talent development for many decades. It addresses one of the most important questions in skill development: What is the optimal form of practice to learn something? To what extent does effective practice vary across domains like reading, math, or sports? Is there a universal principle behind expert performance across different fields?

Researchers in talent development have investigated these questions by studying expert performers—writers, athletes, musicians, and anyone who has reached a high level of excellence in a field. They analyzed the types of practice these individuals engaged in to reach their level of expertise.

It turns out that across all fields, the types of practice that lead to high-level performance follow common patterns.

Deliberate practice consists of individualized training activities specifically designed to improve targeted aspects of performance. These activities focus on refining skills that are right at the edge of a learner’s ability—things they can almost do but haven’t fully mastered.

For example, a gymnast who has just learned to land a backflip might technically get their feet on the ground, but their landing is unstable, sometimes causing them to fall. To improve, they might videotape their backflip, analyze their form, identify specific mistakes, and then practice repeatedly to correct those issues.

The same principle applies in math. A student might attempt problems that are just beyond their current ability. They understand the problem and have some relevant techniques, but there’s a new insight they haven’t yet grasped. They learn this new insight from a teacher, textbook, or another source and then practice applying it to different problems.

The key finding in talent development research is that among high-level performers, the biggest factor explaining differences in skill was the amount of accumulated deliberate practice. The most capable individuals had simply engaged in more deliberate practice over time.

For practice, it is not just any type of practice. It is very specifically individualized. Unfortunately, it is something that can only be approximated in a classroom setting. True deliberate practice requires a one-on-one coaching environment where the instructor provides specific feedback on what the student needs to refine. The student then repeatedly practices and adjusts based on that feedback.

This makes deliberate practice difficult to implement in a classroom, but it represents the ideal one-on-one instructional scenario. There are still some debates in talent development regarding deliberate practice. However, these debates are not about whether deliberate practice is the most effective learning method—that is a settled question. It is well established that deliberate practice is the most effective form of practice.

The ongoing discussions in talent development focus on the extent to which the volume of deliberate practice predicts whether someone will become an expert. Other factors also play a role. For example, in basketball, height is a major factor. It is extremely difficult to become a professional basketball player if you are short relative to the average population or even relative to other players. While other factors matter, they tend to be secondary to the amount of deliberate practice.

One important point about deliberate practice and how it relates to desirable difficulties and scaffolding is that many people believe they are engaging in deliberate practice when they are not.

Deliberate practice involves working at the optimal zone of challenge—doing the most difficult tasks you can successfully complete. You might need multiple attempts, but you should still be able to succeed.

Deliberate practice does not mean attempting something far beyond your current ability. It would not be trying to perform a triple backflip when you cannot do a single backflip. If you are practicing something so difficult that you cannot perform the skill at all, that is not deliberate practice.

If you are practicing the same thing over and over—like someone learning guitar who plays the same song repeatedly without breaking it down to improve specific aspects—that does not count as deliberate practice either.

One key finding in research on deliberate practice is that beginners often stay within their repertoire because it feels comfortable and creates a sense of fluency. You feel capable when practicing in your comfort zone, but deliberate practice is about pushing yourself beyond that, which is inherently uncomfortable. It is not always enjoyable. You don’t do it for fun—you do it to improve performance.

It is like a hard workout—difficult in the moment but satisfying at the end of the day.

Justin: I would say it comes down to whether they are increasing in their ability to perform the skill and whether they can execute it on their own by the end of the practice session.

To give an intentionally obvious example, you would not take a beginner reader who is struggling to sound out words and ask them to read a Shakespeare passage. Even if they show some minor improvement with coaching, the text is so complex that it is beyond their ability. You do not want to attempt deliberate practice at a level that is too far beyond their skill.

A common trap in project-based learning is assigning a project before students have learned the necessary underlying concepts and skills. For example, in an engineering or physics class, a student might be asked to complete an engineering project or physics lab before mastering the foundational knowledge. In such cases, the task becomes an ineffective learning experience.

Even if they receive coaching and make some progress, the additional complexity creates unnecessary cognitive load, which overwhelms them and slows down their learning. It is a balance. The task should be challenging, and struggling on the first attempt is okay if they can improve with coaching. However, the further the task is from their current ability, the more unnecessary cognitive load they experience, and the slower their learning will be.

Justin: I would suggest diving into some key books.

You mentioned How Learning Works—or was it Make It Stick?

Oh, Make It Stick, right. There are several really good books. Make It Stick, Visible Learning, and How Learning Works are some examples. I can’t remember the authors of each, but these books summarize a lot of the research literature.

I recommend starting with these books. One thing to look for is that they should have extensive references to research. Many books discuss learning but don’t include citations. They make claims that might sound reasonable at an intuitive level, but many of these claims are not actually supported by research.

As we’ve discussed, many effective learning strategies are counterintuitive at first. Understanding how they work requires looking deeper.

After reading these books, I recommend going further and exploring actual research papers. A lot of academic literature is behind paywalls in journals, but there is also freely available research online. If you go to Google Scholar and search for a paper by name, you can often find a PDF copy.

At the end of the day, everyone has to do their own homework. There are so many claims floating around in the education space that it’s hard to know who to trust. The only way to be sure is to fact-check and look up the original research.

This might seem overwhelming at first, especially for a teacher managing a full classroom and professional obligations with limited time. But learning the science of learning is like learning any other subject. You start with introductory material, then gradually go deeper over time.

As you read about these concepts, it’s a good idea to apply the principles of learning to your own study process. Instead of spending weeks reading only about spaced review, you might study that one day and then read about another strategy the next. You could interleave topics, create flashcards, and test yourself on learning principles—recalling key studies and findings. If you are learning about the science of learning, applying these principles as you study allows you to see them in action.

There are also a number of sites that summarize this information with direct explanations of how it applies to the classroom.

Justin: The one that comes to mind is Retrieval Practice. They create a variety of materials, including pamphlets and books. I’ve seen several pamphlets from them that summarize learning strategies. One of them, How to Use Retrieval Practice to Improve Learning, explains the concept in a straightforward way. Another, How to Use Spaced Retrieval Practice to Boost Learning, covers the spacing effect.

There’s also Interleaved Mathematics Practice: Giving Students a Chance to Learn What They Need to Know by Doug Rohrer, one of the researchers I mentioned earlier. He has a great quote about the spacing effect. These pamphlets provide clear and direct guidance on effective learning strategies.

To summarize some key points from these materials and from Doug Rohrer’s research, the most accessible and impactful strategies are the spacing effect, interleaving, and the testing effect. Combining these strategies creates an interleaved, spaced retrieval practice approach.

For example, if you are using a textbook, you wouldn’t go through it one unit at a time. Instead, you would move around, covering different sections while ensuring that prerequisites are always addressed before introducing new material. You might follow a learning path, take a break, switch to a different topic, and then return to the first topic the next day—interleaving the material while incorporating spaced review.

When assigning quizzes and homework, you don’t have to limit the content to just the material from the latest lesson. You can include review problems from earlier topics as well. This is something you should establish at the beginning of the year. If students become accustomed to homework only covering new material, they may push back if review problems suddenly start appearing later in the course.

I think the number one most powerful thing teachers can do is to use low-stakes, quick, frequent quizzes to leverage the testing effect. These quizzes should cover a wide range of previously learned material, providing spaced review and interleaving different topics and problem types. By including one question from many different categories across various units, you create a mix of problems that reinforces retention.

If these quizzes are frequent and low-stakes, they don’t have to be stressful for students. A lot of test-related stress comes from unfamiliarity. Research has shown that repeated exposure can normalize the experience, helping students adapt to frequent quizzes. These quizzes don’t have to take up an entire class period—just a quick 10- or 15-minute check-in.

Frequent quizzes allow teachers to cover a broad range of material and gather valuable feedback on what students know and don’t know. This helps tailor lessons more effectively. Ideally, students begin to see these quizzes as calibration tools for their learning, giving them opportunities to engage in retrieval practice that will ultimately prepare them for major exams, finals, and other high-stakes assessments.

I would say the number one strategy is frequent, intervally spaced quizzes that review a broad range of material and are treated as routine, low-pressure exercises. They not only reinforce learning but also provide critical information about what students have forgotten and where additional practice is needed.

However, it’s essential to take action based on quiz results. As we discussed earlier, a key aspect of desirable difficulty is the ability to overcome challenges. If a student misses a question on a quiz and nothing is done to address it, then the quiz serves little purpose. Simply continuing with the normal schedule won’t help the student. Teachers need to make adjustments and adapt instruction based on where students need extra deliberate practice.

Justin: It’s my pleasure. Thanks for the opportunity to spread the word about the science of learning. It’s long overdue for these ideas to become part of the mainstream conversation.

Prompt

The following prompt was used to generate this transcript.

You are a grammar cleaner. All you do is clean grammar, remove single filler words such as “yeah” and “like” and “so”, remove any phrases that are repeated consecutively verbatim, and make short paragraphs separated by empty lines. Do not change any word choice, or leave any information out. Do not summarize or change phrasing. Please clean the attached text. It should be almost exactly verbatim. Keep all the original phrasing. Do not censor.

I manually ran this on each segment of a couple thousand characters of text from the original transcript.

Want to get notified about new posts? Join the mailing list and follow on X/Twitter.