Make It Stick

You can use different strategies to learn something. But there is no learning without memory.

For decades or longer, experts regarded intelligence as innate and immutable. But recent research in behavioral science, psychology and neuroscience refutes this, along with much accepted wisdom about how people learn. 

“For the most part, we are going about learning in the wrong ways, and we are giving poor advice to those who are coming up behind us.”

Memories are formed through learning. But learning can occur in many ways and there are several strategies you can employ. Teachers and instructors often fail to consult peer-reviewed, controlled research findings and/or adjust their methods. Many instructors remain stuck on the outmoded notions that rereading, repeating specific elements of a wider skill or knowledge base, and cramming produce the best results.

Spaced out, repeated practice is more effective than cramming.

Instructors often pin a course’s grade on one or two exams per course or semester. As a consequence, cramming and rereading before an exam are popular practices among students. However, a series of shorter quizzes throughout a semester, each accounting for a small part of the overall grade, is more effective in the long term.

“Rereading text and massed practice of a skill or new knowledge are by far the preferred study strategies of learners of all stripes, but they’re also the least productive.”

At Columbia Middle School in Illinois, for example, teachers use frequent quizzes, flash cards, writing exercises and student presentations. Teachers space repetition, present it in different formats, and encourage students to attach the learning to prior knowledge while reproducing it in their own words. Teachers use quizzes and tests so they and students can gauge progress and identify areas for greater focus. 

At Washington University in St. Louis, Professor Andrew Sobel uses similar methods. For years, he graded students only on a midterm and final exam, but found students attended few of his lectures. He switched to surprise quizzes. Students hated them, and many dropped out of his course. Sobel then introduced a series of scheduled quizzes. Students knew these tests were coming and that the tests counted, which gave them the incentive to prepare. Attendance soared, and dropouts decreased. Students loved the new approach because it offered fast feedback and opportunities for targeted studying, and because one poor quiz performance didn’t jeopardize a course grade.

Frequent small tests and quizzes give students the spaced practice that improves memory retention and retrieval. Repetition reinforces the neuronal connections that are formed through learning.

“Interleaved and varied practice” counteracts forgetting.

More challenging learning can potentially yield greater benefits. One example is “interleaving” topics by studying all parts of a skill at once instead of in sequence.

You rapidly forget about 70% of what you read or hear. Spacing out retrieval practice – through quizzes or tests – interrupts the “forgetting curve.” Trying to answer questions or solve problems before you receive instruction (“generative learning”) appears to spark curiosity, priming your brain’s receptors to encode and embed learning.

“The more effort you have to expend to retrieve knowledge or skill, the more the practice of retrieval will entrench it.”

For decades, teachers and coaches believed in the massed practice approach. You isolate a movement, such as a mid-court forehand in tennis, or long, decimal division in math. This repetition yields fast but fleeting results. These results might translate into desired outcomes if tennis matches and math exams unfolded according to script. When baseball batters, for example, focus their practice only on curve balls or sliders – concentrating in later sessions on fastballs – they don’t hit nearly as well in games as players who practice hitting all pitches in the same session.

“Making mistakes and correcting them builds the bridges to advanced learning.”

Your performance improves over time if you introduce variety and simulate real-life situations. In experiments, students who break problems down to practice the parts discretely score significantly lower on exams than those who mix up problems in practice. Intuitively, it makes sense to teach people one part or stage of a job at a time, waiting until they master the easier parts before moving on. But you’ll gain better results by interleaving practice – moving between and within steps and levels. Your trainees will make more mistakes and perhaps believe they’re failing. But they will retain more.

Providing feedback to learners and encouraging trial and error emphasize more difficult, active learning over easier, passive learning. Varied practice might build better skills because it engages distributed networks of the brain.

“Practice like you play.”

A coach for the Los Angeles Kings hockey team drilled his players as most coaches did, by repeating the same plays over and over from the same place on the ice. When he moved to the Chicago Blackhawks, the coach drilled different types of passes from different parts of the ice, under circumstances that more closely resembled true game conditions. Perhaps not coincidentally, the Blackhawks went on to become world champions. This learning practice works whether identifying bird types, studying art history or mastering legal jurisprudence.

“It’s not just what you know, but how you practice what you know that determines how well the learning serves you later.”

A math textbook might address problem types in sequence, chapter by chapter. But on an exam, students see all types of problems, and usually not in any particular order. Studying should reflect this reality. You can introduce contrast by, say, tackling different types of math problems or studying different styles of paintings in the same session. That approach enables you to perceive differences that help you discern nuances and discriminate between types. Practice like you play, whether you fly jets or perform surgeries.

Practicing like you play is especially effective when your work might cost or save lives. It means on-the-job, experiential learning and/or simulations. Don’t rely on variety alone. Repeat experiential practice around specific skills. Train each skill separately so you don’t repeat one thing mindlessly. You must think and remember. When you master something, don’t set it aside. Revisit it occasionally to maintain memory. 

“Effortful retrieval” from memory reinforces the underlying neuronal correlates.

After memories have been encoded by your brain they may be consolidated to form a long-term memory. During memory retrieval, you recall what you have learned earlier. At this point, making an effort can be beneficial in many ways.

For example, as one form of effortful retrieval, spaced out practice reinforces the neuronal routes. As memories are reconsolidated, learning can be deepened.

Effortful practice over thousands of hours can also allow you to combine several related aspects of a subject or skill into a mental model. Mental models then enable you to respond fast and expertly to a trigger.

“The more we learn, the more possible connections we create for further learning.”

It’s therefore beneficial to design learning to make it reasonably difficult and challenging.

Employ mnemonics techniques to remember multiple items.

Attach difficult-to-remember things to memorable phrases or images, such as the mnemonic “I Value Xylophones Like Cows Dig Milk” for remembering ascending categories of Roman numbers from one to 1,000. “Memory palaces” involve associating new concepts with familiar places, such as your bedroom. For example, your desk may become the first step in solving an equation, your chair the next, and your bed the third. This helps you remember as you visualize your room from desk to chair to bed.

Heed what you don’t know, and beware your biases.

In most cases and for most daily decisions, automatism kicks in: This is what Daniel Kahneman in his book Thinking, Fast and Slow calls “System 1” subconscious thinking. System 1 thinking gets you through the day by not taxing your limited cognitive reserves to perform smaller, routine and repetitive tasks or mental chores. Sometimes, however, you can’t leave decisions to intuition. Pilots can suffer illusions under certain conditions, for example. In these cases they must deliberately fight System 1, which may tell them they’re climbing when they’re actually descending.

“By wading into the unknown first and puzzling through it, you are far more likely to learn and remember the solution than if somebody first sat you down to teach it to you.”

Your memories can change and suffer biases, particularly from the neat narratives you and others forge to describe the past and to make sense of life, or even from the power of suggestion. A wide range of biases can affect your thinking and memory, especially System 1 thought. To solve problems and make unfamiliar decisions, engage “System 2” – deliberate, conscious and slow thinking. This enables you to create increasingly complex mental models that gradually – through repeated use of System 2 – turn even complex routines into habits System 1 can manage.

This degree of mastery, however, may cause you to underestimate how long it will take others to learn things you know cold. Again, slow down and let learners build their mental models using System 2.

Learners who read about and believe they deeply understand a challenging concept tend to delude themselves. Until they have rewritten concepts in their own words, applied them to past knowledge and/or used them in practical application, they haven’t mastered anything. Help learners assess their performance and knowledge. For example, show them their test results in comparison with those of people with true mastery of the topic, and ask your learners to assess their gaps. This helps learners overcome biases and other obstacles. Pair learners with peers who have more experience, as airlines do with pilots or as interns pair with physicians.

Use frequent, spaced testing and retrieval to allow learners and instructors to gauge progress, calibrate feedback and learning, and to build lasting competence. Consider “dynamic testing” – using results from tests to tailor subsequent lessons or other learning exercises on the basis of the learner’s knowledge gaps. Avoid catering to individual styles, but help learners discover their idiosyncratic strengths and intelligences, and apply those to help students learn.

Reject the myth of matching instructional design with a learner’s best learning style.

People may have learning preferences, but scientists have recently questioned the usefulness of the popular practice of matching teaching styles to those preferences. As a teacher, consider your instructions in terms of how it best enables learning. Widen your approach as much as you can to engage learner’s “multiple forms of intelligence.”

Some learners utilize more successful habits than others. Those who recognize key elements of the learning and who build mental models create advantageous learning structures or frameworks. 

“Everyone has learning preferences, but we are not persuaded that you learn better when the manner of instruction fits those preferences.”

These learners build on concepts and can apply learning from one area more readily to another than those who memorize what they learn.

Adopt a can-do attitude.

Learning success largely comes down to believing you can do it. When kids adopt the attitude that they can learn anything, the chances that they will do so are higher.

“Our brains are capable of much greater feats than scientists would have thought possible even a few decades ago.”

Teachers and parents help kids more by praising their efforts than by telling them they’re smart. A person who believes he or she has natural talent may resist putting in the necessary thousands of hours of deliberate practice.