The Protégé Effect — or why you discover how well you know something only when you try to explain it

From the "Our Everyday Mind" series

---

You're explaining something to a friend, out loud, something you were sure you had a solid grip on. And halfway through your own explanation, you stop, suddenly unsure. A connection that felt clear in your mind a moment ago turns out to be fuzzy the second it has to be put into words for someone else.

A minute ago, you were convinced you knew it. Your explanation just showed you, more clearly than any test you could give yourself, how well you actually knew it.

Betty, the virtual girl who needed teachers

Researchers Catherine Chase, Doris Chin, Marily Oppezzo, and Daniel Schwartz studied, in 2009, a digital environment called Betty’s Brain, used to study exactly this pattern. In the program, students weren't learning biology for themselves — they were learning in order to teach the material to a virtual character, Betty, who "learned" whatever was built, step by step, into her knowledge map.

All the students were working with the same content — a text about how the body responds to fever. The only difference was the framing: some believed they were learning the material for their own test, others believed they were teaching it to Betty, who would be tested. The students in the "teaching" group spent more time reading and reviewing the material and, in the end, learned more — with the difference being largest among lower-performing students.

A second study, done with younger students, showed part of why: children in the "teaching" group treated Betty like a real person, attributing intentions and responsibility to her, and when Betty got something wrong, they reacted with a visible trace of frustration or disappointment — and kept correcting, revising, trying again. It was as if the mistake wasn't only theirs anymore: it became something they had to fix for Betty.

From Betty to the little brother you help with homework

You don't need a piece of software to recognize the pattern.

An older sibling explaining math homework to a younger one often discovers, mid-explanation, an old misunderstanding of their own, hidden until then under a correct answer memorized by rote. An experienced employee, asked to onboard a new colleague, ends up clarifying, in their own words, procedures they'd been repeating on autopilot, unable to describe from scratch anymore. A programmer talks out loud to a rubber duck on their desk, explaining, line by line, what a piece of code does — and often finds, on their own, the bug that silent searching wasn't turning up.

It happens to me even as I write these articles: only when I have to explain clearly, to someone who hasn't read about it before, a psychological pattern I thought I understood well, do I notice the corners I hadn't actually fully understood.

Why being responsible for someone else changes everything

Part of the explanation has to do with responsibility. When you believe you're about to answer for someone else's knowledge, not just your own, a mistake no longer feels like a direct personal failure — it feels like a problem to solve together. This seems to free up more energy for correcting and revising, not less — exactly what Chase and her colleagues observed in the children who were "protecting" their virtual student.

Another part has to do with simply preparing to explain, regardless of whether the explanation ever actually happens. Long before Betty’s Brainpsychologists John Bargh and Yaacov Schul showed, in 1980, in a study titled „On the cognitive benefits of teaching”, that simply expecting to teach material to someone else — not just be tested on it — changes how you study it from the start. Anticipating the explanation leaves its mark on learning before any explanation ever takes place.

Recognizing something isn't the same as being able to explain it

Explaining doesn't just require recognizing information, the way you do on a multiple-choice test. It requires retrieving it from memory with no prompt at all, choosing what's essential and what can be left out, building the causal links between ideas yourself, noticing the gaps in your own understanding, and anticipating where the listener might get lost.

These are two very different levels of knowing something: "I recognize it when I see it" and "I can produce it myself, from scratch, for someone else." Trying to teach — whether real or only imagined — pushes us toward that second level, one that ordinary tests don't always demand.

You don't always need a real student

Here's a nuance that partly contradicts the initial intuition: it doesn't seem to matter decisively whether the audience for the explanation is real. The Feynman Technique, inspired by physicist Richard Feynman's practice of breaking down complicated concepts into simple explanations, calls for writing an explanation clear enough that a child could understand it — without any child actually being in the room. The exact four-step formula circulating today under his name isn't a method Feynman himself ever published under that title. Programmers who talk to a rubber duck on their desk are, intuitively, using a similar mechanism: externalizing their reasoning forces them to walk, line by line, through what they'd otherwise do automatically — without the duck understanding, obviously, a single word.

An important part of the benefit seems to appear even before the actual teaching begins, in the moment the mind starts organizing what it knows into a coherent explanation for someone else. Interacting with a real student, like Betty, can add other mechanisms on top of that — questions, feedback, responsibility, the need to adapt on the spot to someone else's confusion.

Who could you explain this to right now

Next time you want to know how well you actually understand something, stop asking yourself. Find someone — or, if needed, imagine someone — and try explaining it to them, out loud, from scratch.

The exact spot where you get stuck, where you search for a word that won't come, where you feel the need to stop and rephrase — that's the spot where you didn't actually know it, no matter what you believed a moment before.

Maybe we don't really learn something the first time we understand it, but the moment we're forced to find, for someone else, the right words to explain it.

---

If you feel you want to explore more of what is happening in your emotional life, there is a space for that.