Five ways to learn that hold up in the research
Most of the ways people study feel productive and fade within days. These five feel harder while you do them and last longer, and each has been tested in controlled studies.
1. Close the book and try to recall it
Rereading a chapter makes it feel familiar, and familiar is easy to mistake for known. Trying to pull the material out of your own memory is what makes it stick.
In a 2006 study by Henry Roediger and Jeffrey Karpicke, students read a short prose passage. One group read it four times. Another read it once and then spent the remaining sessions writing down everything they could remember. Five minutes later, the rereaders recalled more. A week later the result had flipped: the group that practiced recall remembered about 61 percent of the passage and the rereaders about 40 percent.
How to do it: after a page or a lecture, close it and write or say what you remember before looking again. Then check what you missed. Flashcards work the same way if you answer before you flip.
2. Spread the same hours over more days
Three one-hour sessions on three different days teach more than one three-hour session, even though the time is the same. A 2006 review by Nicholas Cepeda and colleagues pooled hundreds of experiments on this and found the advantage for spaced practice again and again. It also found that the longer you need to remember something, the wider the gaps between sessions should be.
How to do it: come back to yesterday’s material for ten minutes before starting anything new, and to last week’s once more a few days later. Cramming the night before a test can get you through the test. It does not get you to next month.
3. Mix up the kinds of problems you practice
Textbook problem sets usually come in blocks: twenty problems on one method, then twenty on the next. Inside a block you already know which method to use, so you never practice choosing it, and choosing is most of what a test asks.
Doug Rohrer and Kelli Taylor tested this with college students learning to find the volumes of four kinds of solids. Students who practiced in blocks did better during practice. On a test a week later, students who had practiced with the problem types shuffled together scored 63 percent and the blocked group 20 percent.
The effect is not limited to math. In a 2008 study by Nate Kornell and Robert Bjork, people learned to recognize the styles of different painters better when the paintings were mixed than when each artist’s work was shown together.
How to do it: once you have learned two or three methods, practice them in a shuffled set. Before solving each problem, name the method you are going to use.
4. When you are stuck, look one step down
A lot of being stuck is a missing piece underneath. Derivatives stop making sense when fractions are shaky. A chemistry unit on reaction rates falls apart if exponents are not solid. Working harder on the new topic does little when the problem is in the old one.
How to do it: when a topic will not click, find the first step in a worked example you cannot do on your own. That step usually belongs to an earlier skill. Practice that skill by itself until it is easy, then go back.
5. Explain each step in your own words
In a 1989 study of students learning physics from worked examples, Michelene Chi and colleagues found that the students who learned the most were the ones who stopped at each line and explained to themselves why it followed from the one before. The weaker students read the examples through.
How to do it: cover the next line of a worked example, say what it should be and why, then uncover it. If you cannot say why, that is the line to study.
What to stop doing
A 2013 review led by John Dunlosky rated ten common study techniques by how well they hold up across studies. Practice testing and spacing were the two rated high. Rereading and highlighting were rated low. They are not harmful, but they are slow, and they leave you feeling more ready than you are.
How Kolu uses these
Kolu is learning software built around these five habits. It starts with a short skill check to find the edge of what you know, asks you to answer rather than reread, brings skills back on a schedule before they fade, mixes old problem types into new sessions, and steps back to the missing skill when you get stuck. Starting is free.