The Protégé Effect: Why Teaching Others Helps You Learn

The protégé effect is the finding that you learn material better when you teach it, or even just expect to teach it, than when you study it for yourself. Students who prepare to teach organize the material and hunt for the main points; students who then actually teach it, without notes, have to pull it from memory and explain it, which strengthens it further.

The name comes from a 2009 Stanford study in which students worked harder to learn biology when they were teaching it to a computer character than when they were learning it for themselves. The broader idea, often called learning by teaching, is much older: the Roman philosopher Seneca wrote that people learn while they teach.

This guide covers what the research shows (including what it doesn't), why teaching works, a popular myth to ignore, and six practical ways to use the protégé effect even if you have no one to teach.

What is the protégé effect?

The protégé effect describes how a learner who takes on the role of teacher, the person with a protégé, learns the material more deeply. It covers three situations:

1. Expecting to teach. You study knowing you'll have to explain it to someone later.
2. Teaching. You actually explain it, to a classmate, a camera, or a study group.
3. Being responsible for someone else's learning. You care whether your "student" gets it, so you put in more effort.

It's related to two other techniques, but it isn't the same. Self-explanation is explaining things to yourself while you study, with no audience. The Feynman technique is a method for testing your understanding by explaining simply to an imagined beginner. The protégé effect is about what happens when another learner, real or expected, is part of the picture.

What does the research show?

Expecting to teach changes how you study

In 2014, John Nestojko and colleagues at Washington University in St. Louis gave students a passage to study. Half were told they'd be tested on it. The other half were told they'd teach it to another student, who would then be tested. In reality, nobody taught; everyone took the same test.

The students who expected to teach recalled more of the passage, and their recall was better organized. They also answered more questions correctly, especially questions about the main points. The researchers' explanation: when you expect to teach, you naturally use better study strategies, such as sorting out what matters most and how the ideas fit together.

Actually teaching helps your memory last

Logan Fiorella and Richard Mayer had college students study a lesson on the Doppler effect. Some studied expecting a test, some expected to teach, and some prepared and then taught the lesson in a short recorded video.

Preparing to teach gave a boost on a test right after studying, but that boost had faded a week later. Students who actually taught the lesson still outperformed the others on the one-week test. Expecting to teach makes you study better; explaining it out loud makes the learning stick.

The pooled evidence

In 2019, Keiichi Kobayashi combined 28 studies comparing students who studied to teach, or studied and then taught, against students who simply studied.

Both helped with deep understanding as well as basic recall, and both effects held after a delay. The biggest gains came when students expected to teach interactively, face to face with someone who could ask questions, and then did so. Recording a one-way video or writing an explanation helped less.

Why teaching works: mostly retrieval

A 2018 study by Aloysius Koh, Sze Chi Lee, and Stephen Lim tested one explanation directly. After studying a text, some students taught it on camera without notes, so they had to recall it while explaining. Others taught it on camera by reading from a prepared script. A third group simply practiced recalling the text, and a fourth did an unrelated task.

A week later, teaching without notes and plain recall practice produced similar comprehension scores, and both beat teaching from a script and the unrelated task. The authors' conclusion: much of the benefit of teaching comes from retrieval practice, the effort of pulling information out of memory.

That's the most useful finding for students. If you teach from your notes, you give up much of the advantage.

Tutors learn too

This isn't only a lab effect. A 1982 meta-analysis of 65 school tutoring programs found that tutoring helped the students who were tutored, and that the tutors themselves also gained a better understanding of the subject they taught.

And the effect that gave the protégé effect its name: in Catherine Chase and colleagues' 2009 study, eighth graders using a program called Betty's Brain either taught a computer character named Betty or learned the same biology for themselves. Students who taught Betty spent more time on learning activities and learned more, with the biggest benefit for lower-achieving students.

In a second study, recordings of fifth graders thinking aloud as they taught suggested why: they felt responsible for Betty, and when Betty got something wrong, the mistake felt like hers rather than theirs, which made it easier to try again.

Why does teaching others help you learn?

Four mechanisms show up across these studies:

1. You organize instead of collecting. A teacher has to decide what comes first, what matters most, and what can be skipped. That's how the expecting-to-teach group recalled more main points.
2. You retrieve. Explaining without notes is a recall test you set for yourself. See what is active recall for why that matters.
3. Questions expose gaps. A listener asks "but why?" at exactly the point where your understanding is thin. That's why interactive teaching beats a one-way lecture.
4. You care more. Feeling responsible for someone else's understanding makes you put in effort you might not put in for yourself.

Is it true you remember 90% of what you teach?

No. The "learning pyramid" or "cone of learning" image, which claims people remember 10% of what they read, 50% of what they discuss, and 90% of what they teach, has no research behind it. A 2013 review in Medical Teacher traced its use in medical education and found the percentages came from misleading sources, with authors citing work that didn't contain them, citing each other in circles, or citing nothing at all. The reviewer's recommendation was to stop referring to it.

Teaching does help learning. The real effects are moderate, they depend on how you teach, and nobody has measured a fixed percentage. It's the same kind of tidy, unsupported claim as the learning styles myth.

Limits of learning by teaching

• You can confidently teach errors. Fluent explanations can be wrong. Check claims against the textbook, especially before you teach someone else who will believe you.
• It's slower than flashcards. Teaching suits concepts and processes. For vocabulary or drug names, flashcards with spacing are more efficient.
• Preparing alone fades. Expecting to teach helped right away but not a week later in the Fiorella and Mayer study. Follow through and actually explain it.
• You need something to teach. It works after a first pass through the material, not as a replacement for learning it.

6 ways to use the protégé effect

You don't need a classroom. Pick the option that fits your week.

1. Study as if you'll teach it

Before you open the chapter, tell yourself you'll be giving a 5-minute lesson on it tomorrow. While you read, collect what a teacher needs: the three main points, one clear example, and the question a confused student would ask. This is the Nestojko effect, and it costs nothing.

2. Teach from memory, never from notes

Put your notes away before you start explaining. If you get stuck, stop, note the gap, and keep going. Check the gaps afterwards. Teaching from a script is the version that showed little benefit.

3. Find a listener who asks questions

A roommate, a sibling, or a classmate who missed the lecture all work. Ask them to interrupt whenever they're lost. Their "wait, why?" is worth more than a polite nod. Interactive teaching produced the largest effects in the meta-analysis.

4. Run a teach-back study group

Split a chapter into sections. Each person prepares one section and teaches it to the group without notes, then takes questions. Everyone gets the expecting-to-teach boost for their section and a retrieval-and-questions session at the meeting. Our guide to study groups that actually work covers how to keep them on task.

5. Record a 5-minute lesson

No listener? Explain the topic to your phone camera as if to a classmate, from memory. It's less powerful than a live audience but better than silent rereading, and the playback shows you where you hesitated.

6. Help someone one step behind you

Answer questions in your class group chat, help a younger student, or tutor a course you finished last year. Explaining fundamentals to a beginner is where gaps in your own foundations show up.

A 30-minute teach-to-learn session

Use this for one topic you need to understand, not just memorize.

Then schedule a short re-teach two or three days later. It's the second retrieval, after a delay, that makes it stick.