Does "learning by doing" work? What does the research say?
I often get asked about learning by doing. What is “learning by doing”? Does it work? When does it work?
In this column, I offer an overview of the process and explain how you can implement this yourself to harness your own potential as a learner. For instance, I really recommend approaches like “brain dumps.
What’s more, I flag when learning by doing does not work because many times the approach is not used in an effective way, and I would not encourage learning by “doing” if you’re at the start of the learning process.
What is learning by doing?
Learning by doing is the idea that we learn more when we actually “do” the activity.
For example, imagine you are a jazz musician looking to understand how chords relate to one another. Traditionally, you might play the chords over and over again alone in the studio. With the learning by doing approach, you would gain a basic understanding of the chords and then hop on stage to play the chords as an improvised piece with other musicians-- active engagement, not passive practice. Active engagement facilitates deep learning and encourages mistakes - i.e. those ‘wrong’ chords - and how to learn from those as well.
The American philosopher, John Dewey, first popularized learning by doing. For Dewey, this meant a heavy emphasis on student engagement. This approach upended the traditional notion that learning happens through lectures and rote memorization. Dewey became famous by arguing that we learn best when we deeply engage with the material. He believed that the best way to achieve that was to create a practical curriculum that had relevance to students’ lives and experiences. Dewey’s insight, already nearing a century old, is freshly becoming relevant again today as modern researchers empirically show the importance of learning doing (with some major caveats.)
The problem with short-term memory is that it’s so, well, short.
When does learning by doing work?
For learning by doing to work, you need to lay some initial groundwork. Recent research shows that learning by doing works when it comes at the right point in the learning process. What does this mean? First, it’s important to underscore that learning is a process. Learning builds on itself, and if learning by doing comes too early, people get overwhelmed. They don’t learn.
This makes intuitive sense in real life situations. Let’s go back to the jazz musician. Imagine hopping onstage with only the slightest knowledge of the new chords. Your fingers wouldn’t find the right spots on the guitar. The chords wouldn’t ring out and rather than enjoying your time improvising with other musicians, you would come off stage dejected and reeling from frustration. Before you can hop onstage, you need the basic ability to comfortably play the chords. The same principle holds in the classroom.
Part of the reason is short-term memory. It holds the key to making learning experiences productive. Researchers like John Sweller have shown that short-term memory is often where learning happens. If we want to successfully improvise with a set of new chords, then short-term memory has to process the chords. Only after short-term memory processes the chords can they arrive in long-term memory.
But the problem with short-term memory is that it’s so, well, short. Think about the jazz musician’s hands. Each chord involves moving up and down the guitar’s neck, placing each finger on the right string at the right time, allowing only the right strings to ring out -- a complex feat of memory and dexterity.
Learn too many chords too quickly and your hands won’t follow because your short-term memory has been overwhelmed. Your fingers fall out of place, each chord loses its shape and what should be a beautiful sound comes out a jumbled mess of dissonance. Your hands, just like your mind, need time to learn the basic techniques. Only after establishing this basic foundation will the musician’s hands find the way.
Understanding the ‘shortcomings’ of short-term memory helps explain why learning by doing can’t come too early. To learn, we need to break down knowledge and skills into digestible parts and concentrate on discrete bits of mastery. In other words, you have to learn a few chords at a time before you can master multiple chords in multiple key signatures.
If learning by doing comes too early, then we can’t learn. Hands-on learning doesn’t work in many schools and colleges precisely because our short term memory has yet to break down the material into bite sized chunks. The same principle holds with practice: too often, people try to develop their skills without knowing what exactly they’re developing, without any sort of goals or targets (the bite sized chunks) that lead up to and create the larger picture.
Why does learning by doing work?
Learning by doing works after you’ve already gained some familiarity with the content. It works because the technique calls you to actively engage with the material and generate the knowledge yourself, bit by bit, chord by chord.
One effective technique to facilitate learning by doing is what’s known as the “generation effect.” Also known as the testing effect or retrieval practice or even learning by doing, the “generation effect” underscores that students understand and remember material better when they have been asked to generate it themselves rather than reviewing an account generated by someone else (e.g., re-reading a section of a textbook, sight reading a musical piece).
Many teachers focus on imparting knowledge to students; they imagine themselves as “putting information into students’ minds.” But, the science of learning shows that students need to construct knowledge for themselves, and in many cases, effective learning would be better described as a process of “pulling information out of students’ minds.” Next time you read a new text, ask yourself these questions: What’s this text about? What point is the author trying to make? Is there anything here that seems confusing? These questions focus your attention on the substance of the material and guide you into the learning by doing technique.
All too often I’ve found myself reading scientific literature only to come up blank when someone poses the simple question, ‘what’s that about?’ I spent far too much time passively reading material. I learned that only after asking, ‘what’s the central question of this piece?’ ‘what’s the purpose of this article,’ do I fully learn the material.
We learn a lot more when we consistently ask these questions be it at the end of each paragraph—or even the end of a sentence. So while some textbooks might offer “reading comprehension” questions at the close of each chapter, you'll take away a lot more if you ask yourselves these sorts of queries more frequently.
Active engagement and techniques which force you to work harder to remember the material are the most effective ways to learn.
How to use learning by doing?
One way to incorporate the “learning by doing” technique into the classroom is to give students frequent low-stakes quizzes. By design, these quizzes do not really assess performance. Instead they prompt students to engage with the content and generate the learned information from their own minds.
Indeed, research shows that the very process of retrieving such information itself improves understanding, increases recall, and promotes the “transfer” of knowledge to new settings. In other words, it makes learning a matter of doing, an active, effortful process.
Psychologist Rich Mayer, an unlikely crusader for the learning by doing technique, has recently written a lot about learning as a type of mental doing. Mayer’s generally a pretty avuncular and soft-spoken midwesterner. In conversation, Mayer prefers “somewhat short of being exemplary” to the more crude (if not equally accurate) “someone screwed up” For someone like Mayer, people don’t have bad intentions, we just experience the consequences of bad decisions. Mayer’s favorite advice? “Don’t radiate negative energy.”
But on the issue of learning as a form of dedicated cognitive effort, Mayer has become something of a firebrand and his lab at Santa Barbara has repeatedly shown that we gain expertise by actively producing what we know. As he told me flatly: “Learning is a generative activity.”
Mayer gives a pretty good description of what we need to do in order for the technique to work. First, we select information and decide exactly what we’re going to learn—like maybe a bit of Soviet history or Buddhist philosophy. Then we integrate that information into what we know by creating some type of mental connection between current knowledge and the information we hope to learn.
So if someone is learning about the Soviet dictator Stalin, they need to link what they know (that Stalin was a dictator) to what they want to learn (that Stalin grew up in Georgia, killed millions, centralized power in Russia, and helped win World War II).
We see the power of mentally doing--of creating value in an area of expertise--in even the most basic of memory tasks. Want to remember the French word for home, “maison,” for instance? Simply delete the letter “o” when you read the word and you are far more likely to recall the word.
While deleting a letter to recall the word might seem counterintuitive, this makes sense when you consider what that deletion does to how we process the word. Adding the “o,” effectively completes the word. You’re finishing the thought—and in the most basic of ways, you’ve done some work to produce learning--and it’s that work that makes it more meaningful and more likely to stay in your long term memory.
This same basic idea of working harder to more effectively produce the learning extends to more difficult cognitive tasks like reading. If we push ourselves to dream up some sort of mental image of what we’re reading--say by imagining the text in our minds---we retain a lot more of the knowledge. By creating a type of “mind movie,” we're building more cognitive connections which makes learning more durable.
Recent research on learning as mental doing has shifted conventional wisdom around how people gain expertise. In a large and recent review of the research, Kent State’s John Dunlosky and colleagues evaluated various learning techniques and found that two of the most common learning techniques, highlighting and rereading, are also some of the most inefficient ways to learn. But the question remains, why?? The answer is that rereading and highlighting do not push people to build their knowledge.
So which approaches did Dunlovsky’s landmark analysis find to be most effective? Those which facilitated learning by doing. Over the phone, Dunlovskly told me that the most effective techniques were the more effortful activities like self-quizzing and self-explaining. “This is a fundamental feature of how our minds work,” and as he explained, when we learn, “we're not just copying the information. We are making sense out of facts.”
The most effective techniques were the more effortful activities like self-quizzing and self-explaining.
There’s lots of talk about learning by doing by organizations like the World Leadership Institute.
So, how can you follow these leaders and introduce learning by doing into the classroom? Rather than issuing a multiple choice quiz, ask your students to explain the concept or process in question. Use a concept known as elaboration and ask your students to make connections between different concepts and explain how they are connected.
Indeed, recent research led by Michelene T.H. Chi found that successful students tend to generate many self-explanations that refine and expand the conditions under which a concept may be applied. Conversely, less-achieving students rely more on isolated examples. Other research by Chi found that something as simple as saying explanations and elaborations aloud can lead to robust benefits for students’ learning.
Another simple to implement technique is what’s known as "repeat backs." The next time someone gives you a set of detailed instructions, take a moment and repeat the instructions back. Deliberate repetition forces you to generate that knowledge which means you’ll be far more likely to remember the instructions.
Brain dumps, or more formally “free-recall” are just that -- jotting down what you remember about a given topic. In spite of their simplicity, brain dumps have been shown to be a very effective learning technique. Because brain dumps force you to retrieve and roughly organize the material yourself (i.e. learning by doing), they can be really effective in fostering deep learning.
So what have we learned? Active engagement and techniques which force you to work harder to remember the material are the most effective ways to learn. Yet, you have to be mindful of when to implement these techniques. Jump in too fast without laying a basic groundwork for learning, and the benefits of learning by doing dissipate.
You have to give your short term memory bit sized chunks of information before you can expect to gain the expertise in question. It’s time to update the old adage from ‘practice makes perfect’ to the more accurate ‘actively engaged practice makes learning more effective.’
Interview with Pooja Agarwal
Ulrich Boser on TEDx Talk
Interview with Ken Koedinger
The Learning Curve publishes articles about how people learn. Please reach out with any ideas for articles on the research on learning