What works in every approach
Also called learning science, retrieval practice, spaced practice, the learning styles myth, cognitive load, evidence-based teaching.
Ages Every age; most studies are of school-age studentsGrades K–12Planning A plan you adaptScreens Screens fit wellShelf Whole-school philosophiesUpdated November 2026
Whatever approach you follow, a few habits make learning stick, and the research behind them is strong. Bring material back with quick quizzes instead of rereading. Space review out over weeks. Show worked examples before asking beginners to solve alone. Mix different kinds of problems once the basics are learned. Pair pictures with words, and ask "why?". And skip teaching to "learning styles": the evidence for it isn't there. This page explains each habit in plain words, with what to do this week, for parents and teachers in any approach.
1 Where it came from
These findings come from decades of research by cognitive psychologists on memory and learning. Two summaries pulled it together for teachers. In 2007 the US Institute of Education Sciences published Organizing Instruction and Study to Improve Student Learning, by Harold Pashler and six colleagues, with seven recommendations and a rating of how strong the evidence is for each. In 2013 John Dunlosky and colleagues rated ten common study techniques for how useful they are. Work on cognitive load, reviewed for teachers by the New South Wales education department in 2017, explains why beginners need more guidance than experts.
2 The big ideas
- Pulling it out beats putting it in again. Quizzing yourself (retrieval practice) makes memories last longer than rereading. The IES guide rates quizzing to re-expose key content as backed by strong evidence, and Dunlosky's review rates practice testing high.
- Space it out. Reviewing key material weeks or months after it was first taught helps it last (IES: moderate evidence; Dunlosky: high utility). Cramming feels faster and fades faster.
- Beginners need examples. Working memory holds only about four new "chunks" at once, so novices learn more from studying worked examples than from solving the same problems unaided. As skill grows, fade the examples and let them solve more alone.
- Mix it up once the basics are in. After a skill is learned, practice that mixes different problem types tends to beat a page of the same kind on later tests; Peabody's page on interleaved practice covers the classroom trials.
- Words and pictures together, abstract and concrete together. Combine diagrams with explanations, and link objects or drawings to symbols (IES: moderate evidence for both).
- Ask why. Deep explanatory questions ("Why does this happen? How do you know?") are one of the two IES recommendations with strong evidence.
3 What a lesson looks like
A 4th-grade math lesson built on these ideas might go like this. It opens with a three-question warm-up quiz, no grades: one question from yesterday, one from two weeks ago and one from last month. Students answer from memory, then check. The new skill is multiplying a two-digit number by a one-digit number. The teacher draws an area model (the picture) beside the written steps (the symbols) and talks through one full worked example. Students then study a second worked example and explain each step to a partner. Next come "completion problems", where the first steps are done and students finish them, and only then a few problems to solve alone. The practice set mixes the new skill with addition and division problems from earlier in the year. The lesson ends with a why question: "Why does the area model give the same answer as the steps?" A week later, two of these problems turn up again in the warm-up.
4 Reading and math
In math: short mixed reviews of earlier skills, worked examples for new procedures, drawings and objects linked to the written steps, and fact practice spread over many short sessions rather than one long one. In reading and subjects like science and history: after reading, close the book and tell back or answer a few questions from memory instead of rereading; come back to key vocabulary and ideas a week and a month later; and ask "why" and "how do you know" about what was read. Highlighting and rereading feel productive but were rated low in Dunlosky's review.
5 The adult's job
The adult's job is mostly to plan when things come back and how they are practiced: build quick, low-stakes quizzes into the routine, schedule review weeks later, explain new material clearly with examples before stepping back, and ask good questions. Equally important is keeping quizzes friendly. The point is practice remembering, not grading.
6 How progress is checked
Here the quiz is a learning tool as well as a check. Short, frequent, low-stakes quizzes show what still needs work and strengthen memory at the same time. The IES guide also suggests using tests to find what still needs learning and teaching students to judge their learning after a delay rather than right away (both rated low evidence, but cheap to try): a child who feels sure just after reading may not remember next week.
7 Strengths and criticisms
Strengths
- Works inside any approach: Charlotte Mason narration, classical memory work, Montessori practice, a school textbook or a game.
- Cheap and quick: a three-question quiz or a calendar note to review costs nothing.
- Some of the strongest, most repeated findings in education research.
- Gives parents and teachers a way to judge products and claims, such as "learning style" quizzes.
Criticisms and watch-outs
- Learning styles are not supported. Pashler, McDaniel, Rohrer and Bjork (2008) found no adequate evidence that matching teaching to a child's "visual" or "auditory" style improves learning; very few studies even tested it properly. The evidence is insufficient, not a proof the idea is false, but it is no basis for labeling a child.
- These habits feel harder. Quizzing, spacing and mixing feel slower than rereading and blocked practice, so children (and adults) may resist them. Explain why.
- Worked examples are for beginners. For students who already know a skill, extra examples can get in the way (the expertise reversal effect); move them on to solving.
- Mixing comes after learning. Mixed practice works once students have some grounding in each skill; teach the skill first.
- Some ratings are modest. Pre-questions and judgments of learning are rated low evidence by IES; treat them as worth trying, not guaranteed.
8 What the research says
The IES practice guide (2007) rated its recommendations from its own review: quizzing to re-expose key content and asking deep explanatory questions got strong evidence; spacing, alternating worked examples with problem solving, combining graphics with words, and linking concrete and abstract representations got moderate; pre-questions and the two recommendations on judging learning got low. Dunlosky and colleagues (2013) rated practice testing and distributed practice as the two high-utility techniques of ten, with summarizing, highlighting and rereading among the low. Universal Design for Learning (CAST) is a framework rather than a single finding: its guidelines (version 3.0) ask teachers to offer several ways to engage, to take in information and to show learning, so that every learner can work toward "learner agency". Offering variety for everyone is different from matching a child to one style.
9 How to try it
- Start every session with three quick questions. One from last time, one from last week, one from last month. No grades; just answer, then check. Peabody's guide to a retrieval practice quiz shows how.
- Put review on the calendar. When you finish a topic, note dates about a week, a month and a term later to bring it back with a few questions.
- Swap rereading for telling back. After a chapter, close the book: "Tell me what you remember" or "Write three things you learned." Then look back to check.
- Show, then share, then let go. For a new math skill, work one example aloud, do the next one together, then give a half-finished one, then a full one alone.
- Mix the practice. Once a skill is learned, put it in a set with earlier kinds of problems instead of a page of the same thing.
- Ask why, and drop the style labels. End a lesson with "why does that work?" and offer pictures, talk and hands-on work to every child, not just the ones labeled "visual" or "kinesthetic".
10 Using Peabody with it
- Retrieval practiceWhat it is, how to use it in class or at home, and the research, in more depth.
- Spaced practiceHow to schedule review so it lasts.
- Worked examplesHow to use and fade them in math.
- Learning styles, checked against the researchThe longer answer to whether children learn better in their own style.
- The science behind PeabodyHow the kids' site builds these in: known things come back after 1, 3, 7 and 21 days, and most of each round is things the child already knows.
Peabody is free and has no ads. Pages on peabody.kids are for children; pages here are for the adults teaching them.
Questions people ask
- What is retrieval practice?
- Practicing pulling information out of memory, by quizzing, flashcards, telling back or writing what you remember, instead of rereading it. The IES practice guide rates quizzing to re-expose key content as backed by strong evidence, and Dunlosky's 2013 review rates practice testing as one of only two high-utility study techniques.
- Are learning styles real?
- Children have preferences, but a 2008 review by Pashler, McDaniel, Rohrer and Bjork found no adequate evidence that teaching to a child's preferred style (visual, auditory and so on) helps them learn more. Very few studies were designed to test it properly. The better advice is to use pictures with words and hands-on work for everyone.
- How far apart should spaced practice be?
- The IES guide recommends reviewing key content weeks to months after first teaching it. A simple plan is a few questions a week later, again a month later and again near the end of the term or year.
- Is highlighting and rereading a bad way to study?
- Not harmful, but weak. Dunlosky's review rated highlighting, rereading and summarizing among the low-utility techniques. Closing the book and testing yourself works much better.
- What is Universal Design for Learning?
- A framework from CAST for planning lessons that work for more learners from the start. Its 3.0 guidelines have three principles: offer several ways to engage, several ways to take in information, and several ways to act on and show learning.
Sources
- Organizing Instruction and Study to Improve Student Learning (Pashler et al., NCER 2007-2004), Institute of Education Sciences (2007).
- Which study strategies make the grade? (Dunlosky et al. 2013), Association for Psychological Science (2013).
- Cognitive load theory: Research that teachers really need to understand, NSW Centre for Education Statistics and Evaluation (2017).
- Learning styles: concepts and evidence (Pashler, McDaniel, Rohrer and Bjork 2008), University of South Florida Digital Commons (2008).
- The UDL Guidelines (version 3.0), CAST.
Compare with
- Mastery-based learning: Children move on when they show they've learned a skill, not when the calendar says so, with help until they get there.
- The science of reading: A name for the research on how children learn to read: sounds and letters taught in order, plus the language and knowledge that make words mean something.
- Math fact fluency: Knowing addition and multiplication facts by heart, built from strategies first and then short practice, as one strand of math, not all of it.
- Classical education: Facts and memory work when young, logic in the middle years, argument and writing in the teens, with history taught in order.
- Charlotte Mason: Short lessons, real books instead of textbooks, telling back instead of worksheets, and time outdoors watching nature.
- Every approach in the curriculum guide
Written by Peabody Teach. No ads, no affiliate links, and no program pays to appear. Something wrong or out of date? Tell us.