Mentor Mindset Practice

Nineteen techniques from David Yeager's 10 to 25.

17% → 71%
Revision rates, one sentence of wise feedback
0% → 82%
Same, among students lowest in trust
70%
Felt threatened when the officer said nothing first

The practice

Why it works this way

01 — Recall
You type first
Reading a model answer and nodding is recognition, and recognition feels like learning while mostly not being it. Producing one from memory is what you can still do in a corridor six weeks later.
02 — Interleaved
Never two in a row
You never get two situations from the same technique consecutively. Working out which technique a moment calls for is the hard part, and blocking them hands it to you for free.
03 — Spaced
Bad ones come back
Score badly and it returns in about three scenarios. Nail it and it returns in about thirty. The queue aims to keep you succeeding around 85% of the time.

You then grade yourself against a rubric, which is self-explanation — about the strongest single technique in the learning science literature, at g ≈ 0.55 across Bisra, Liu, Nesbit, Salimi and Winne's 2018 meta-analysis of 69 effect sizes.

There is no multiple choice, no timer, no score, no streak and no badge. The book's own argument is that external rewards move the reward from inside to outside, and a tool that taught the mentor mindset while bribing you to return would be arguing with itself.

Every scenario shows you the protector version alongside the enforcer version once you have answered. The wrong answer that feels warm is the one worth being able to spot in your own mouth.

Everything you type stays in this browser. Progress is kept in localStorage — no account, no sign-in, nothing sent anywhere, nothing I can see.

The nineteen techniques

Setting the standard

1
Calibrate to the zone of proximal development
Set the task just past what they can do alone. If they finish it easily, raise it a step rather than repeating the level.
2
Hold the standard, don't dismiss the fear
Both at once. "That sounds real, and we're still doing it. What would make it possible?" Never "you'll be fine", which dismisses, or "okay, skip it", which lowers.
3
Inclusive excellence
More opportunity to get good, never a lower bar. Keivan Stassun's Fisk–Vanderbilt bridge became the largest US producer of Black and Latino physics and astronomy PhDs without lowering admissions. Uri Treisman's Berkeley calculus groups were harder than the regular course.

Saying why

4
Transparency statement
"My intention here is ___." In Kyle Dobson's policing study, officers who approached with no explanation left 70% of civilians feeling threatened. Opening with one sentence of intent doubled the length of the conversation. Same officers.
5
Wise feedback
"I have high standards, and I'm giving you this feedback because I know you can meet them." Revision rates went 17% to 71% among Black seventh-graders; among those with low trust in teachers, 0% to 82%.

The form of a correction

6
Correct without characterizing
Never say anything negative about the person. Address the work, the situation or the feeling. Not "that shot was bad" but "how did that feel?" The standard stays; the insult goes.
7
Ask, don't tell
Three tests: authentic (you don't already know the answer), uptake (your follow-up uses their words), collaborative ("so we can fix the problem"). Two moments where it matters most — when correcting, and when angry.

Handing over control

8
Presume agency, then hand over the thing
Three levels: respect autonomy, presume agency, give responsibility. The strongest is that they run it, not attend it.
9
Assure long-term commitment
If they fail: "We're going to keep working on this. I'm not going anywhere."

Status

10
Honor their status, never invoke yours
If you're tempted to say "because I'm the adult", say what you actually need and why.
11
Recruit their values instead of forbidding
Prohibition raises the status of the thing prohibited. Align the behaviour with autonomy, status and fairness instead. Bryan, Yeager, Hinojosa et al. 2016, PNAS — 536 eighth-graders read an exposé of food-industry marketing and chose healthier snacks a day later, in an unrelated setting.
12
Validate and explain, don't diminish
"That makes sense to be frustrated about. Here's what's behind it."

Meaning

13
Purpose now, and make the argument broad
Future payoffs don't move adolescents; present meaning does. And narrow relevance excludes most of the room. Not "you'll use this in a finance job" but "this makes your brain stronger so you can improve the world".
14
Self-generated writing
The student writes the argument. If the adult delivers it the mechanism is gone — they can discount you, they can't discount themselves. Four-part story: normalize, change is possible, action steps, snowball.
15
People can change
Target the belief that character is fixed. Yeager's work here is on adolescent aggression and bullying.

Stress

16
Stress reappraisal
The racing heart is the body delivering oxygen — the system working, not a sign of failure. Demands exceed resources; the mentor raises the resources rather than lowering the demand.
17
Synergistic mindsets
Growth mindset and stress reappraisal delivered together. Teach that ability can grow but not that the strain of growing it is useful, and the difficulty reads as proof they can't.

Repair

18
The do-over
If you broke the standard, go back, apologise for not upholding it, and ask curious, authentic questions. Yeager calls this the most realistic tool in the book, because implementation failure is guaranteed.
19
The six questions
Their meaning: what does this mean to you, how does that make you feel, is that serving your goals. Then an alternative meaning and the same two questions again. A truce ends fighting; a treaty establishes terms.

The four principles underneath

Status and respect drive behaviour from 10 to 25. Puberty makes the brain crave pride, admiration and respect, and become intensely averse to humiliation and shame. Social status was what mattered most for survival in ancestral group life, so young minds focus hard on fitting in.

They're built to change, not to be reckless. The neurobiological incompetence model — teen brains are simply underdeveloped — was argued by the MacArthur network to spare juveniles the death penalty. A legal strategy became a scientific consensus. The alternative account, from Ron Dahl and the National Academies' The Promise of Adolescence (2019), is that they are extraordinarily responsive to social opportunity, which looks like impulsivity from the outside.

Anything left unsaid between an authority and a non-authority gets guessed at. And the guess runs negative. Silence is not neutral.

Beliefs gate behaviour. If someone doesn't believe prejudice can be reduced, they avoid cross-group conversations. Change the belief and the behaviour follows.

Sources

Every technique, every name and every piece of evidence on this page is David Yeager's, from 10 to 25: The Science of Motivating Young People. The tool is mine; the content is his. If any of this is useful, buy the book — it is considerably better than a practice app.

  • Wise feedback — Cohen, Steele & Ross 1999; Yeager et al. 2014, Journal of Experimental Psychology: General. Three double-blind randomized field experiments with seventh-graders.
  • Transparency statement — Kyle Dobson's policing field study.
  • Recruiting values — Bryan, Yeager, Hinojosa et al. 2016, PNAS 113(39). Double-blind, randomized, placebo-controlled, 536 eighth-graders. Follow-up in Nature Human Behaviour, 2019.
  • Belonging — Walton & Cohen 2011, Science. N = 92; the Black–white GPA gap cut 52% over three years, from one hour.
  • Purpose — Yeager, Henderson, Paunesku, Walton, D'Mello, Spitzer & Duckworth 2014, JPSP 107(4). d = 0.21 for low performers; 0.06 and not significant for high performers.
  • Stress reappraisal — Jamieson, Peters, Greenwood & Altose 2016, SPPS. 93 community college developmental maths students across five semesters. The 2024 meta-analysis across 44 effect sizes puts reappraisal at d = 0.23.
  • Synergistic mindsets — Yeager, Bryan, Gross et al. 2022, Nature 607(7919), 512–520.
  • The do-over and the six questions — Lorena Seidel.
  • Self-explanation, which is why this tool makes you write first — Bisra, Liu, Nesbit, Salimi & Winne 2018, meta-analysis of 69 effect sizes, g ≈ 0.55.