The Ninth Problem

Act one · The room

You are ten years old and someone has come to get you

It is a Tuesday morning in 1978. A woman you don't know puts her head around the classroom door, says your name to the teacher, and you follow her down a corridor that smells of floor polish and radiators.

The room she takes you to is small and bright. One window, low morning sun, a stripe of it lying across the table. She has a clipboard, a stopwatch, and a deck of cards face down.

She turns over the first card. Two shapes on it, side by side. One is a red square with a dot in it. The other is a blue triangle with a star in it. She explains the game, and the game is beautifully simple: one of the six things you can see — red, blue, square, triangle, dot, star — is the right one. The same right one, for the whole deck. Point to the side you think has it. She'll tell you if you're right.

You point. She says yes, or she says no, and either way you have learned something. Four cards later you know it's star. The next problem starts and the right thing is something else, and you find that too. You get the first one. You get the second. You get the eighth.

Not hard. Not tricky. Impossible — arranged in advance so that no answer you can give will be consistently right. Eight problems you solved, then four you cannot, and nobody in the room is going to tell you which kind you are on.

The cards were never the experiment. The experiment was the next thirty seconds.

Instrument one

The deck

Here is the real task, rebuilt. Every card shows two figures. Each figure has a colour, a shape, and a symbol inside it. One of the six values is the winner for the whole problem; find it by pointing and watching what happens.

Nothing here is scored and nothing here is a test of you. You can stop at any card, and a shorter version is one click away.

Instrument two

Before we look at the record

The page kept count. You did too, somewhere. Let's see whether the two agree — answer from memory, not arithmetic, and don't scroll back.

Act two · The clipboard

What the woman with the stopwatch was actually writing down

Carol Diener and Carol Dweck ran two studies: seventy fifth-graders in the first, sixty in the second, half boys and half girls. Before any child saw a card, they had all filled in a questionnaire about why school things go well or badly — the Intellectual Achievement Responsibility scale. Children at one end of it were called helpless; children at the other end, mastery-oriented. Nobody in the middle was invited.

Then the cards. And the clipboard recorded two things that sound dull and turn out not to be.

One: the quality of the strategy. There are good ways to play this game — hold a hypothesis, check it, drop it when the feedback kills it. And there are ways that cannot possibly work: always point left; go left-right-left-right; always pick the card with the star on it whatever else has happened. Psychologists had already catalogued these. The useless ones are what four- and five-year-olds do before they learn better.

Two: what came out of the child's mouth. From problem six onward each child was asked to say aloud whatever they were thinking and feeling. Raters sorted every sentence into five bins: useful strategy talk, useless strategy talk, attributions, feelings, and things that had nothing to do with the task at all.

Then the ninth problem, and the floor opened.

Among the children who had been labelled helpless, 52% blamed their own ability — their brains, their memory, their not-being-smart. Among the mastery-oriented children the same figure was zero. Not low. None of them said it.

More than two thirds of the helpless children's strategies fell apart under failure; over 60% dropped into the kinds of play a preschooler uses. Meanwhile 80% of the mastery-oriented children held their strategy steady or made it better while failing. Some of them got more sophisticated on problems that could not be solved.

Here is the part that gets skipped in the retelling. The mastery-oriented children mostly did not explain their failures at all. They did not think they were failing. Instead of causes they produced instructions — to themselves, out loud, in the present tense. Try the border. No. Then it's the colour. Nearly two thirds of them offered some statement that it was going to come out all right.

Two children, same cards, same impossible ninth problem. One is writing a biography. The other is writing a to-do list.

Act three · The half nobody retells

It is not that failure hurts them more. It is that success does not arrive.

Two years later the same pair published the other half, and it is stranger than the first. This time they looked at what the children did with winning.

Same task, same two groups, and crucially the same performance — everyone solved the solvable ones. Afterwards the children were asked simply: how many did you get right? How many did you get wrong?

The mastery-oriented children remembered about 7.6 successes. The helpless children, who had done the same work, remembered 5.5. Asked how many they'd got wrong, the mastery group said 3.7; the helpless group said 6.1. The same afternoon, counted two different ways.

Asked how many of fifteen more they'd expect to get, the mastery children said about ninety per cent. The helpless children said fifty. And asked why the good run had happened, the helpless children mostly credited the problems for being easy rather than themselves for being able.

Then a single failure was introduced at the end — and 35% of the helpless children said they could no longer redo the very problems they had already solved. Of the mastery-oriented children, every single one said they could.

Which is why, a moment ago, the page asked you to guess before it showed you the record.

Act four · Two sentences

A cause, or an instruction

The 1978 raters found that the two groups didn't just say different things. They said different kinds of things. One group produced explanations — sentences with a cause in them, usually a cause shaped like a permanent fact about the speaker. The other produced instructions — sentences with a next move in them.

You can hear the difference in three words. “I'm no good at this” is a cause. “Try the colour” is an instruction. Both are true-ish, both are about you, and only one of them tells you what to do at nine o'clock tomorrow.

So here is the instrument, run on you rather than on a fifth-grader, using something real and small.

Game · cause or next move?

Cause or Next Move?

Act five · The other side of the table

Twenty years later, the same lab found a way to plant it in six minutes

In 1998, Claudia Mueller and Carol Dweck ran six studies with fifth-graders. Every child did a first set of puzzles and every child was told they had done well. Then came one sentence, differing by a handful of words.

Some children heard a sentence about them: you must be smart at these. Some heard a sentence about what they did: you must have worked hard at these. A third group heard nothing extra.

Afterwards, offered a choice between an easy set they'd do well on and a hard set they'd learn from, the children praised for being smart mostly took the easy one. Given a genuinely difficult round and then a third easy round, their scores fell. And when asked to write a note to children at another school about the task — anonymously — some of the praised-for-smart children inflated their scores.

Act six · What happened to the idea

The poster on the wall

In 2006 Dweck published Mindset, and the two patterns from the 1978 card room got two names — fixed and growth — and then they got a poster, a lanyard, an assembly, a laminated bookmark, and a line in a million school improvement plans. And then, as ideas that get that big always do, it met the people whose job is to count.

Here is the count, plainly, because you deserve it plainly.

2018. Victoria Sisk and colleagues ran two meta-analyses: 273 studies with 365,915 people for the link between mindset and achievement, and 43 studies with 57,155 people for whether teaching growth mindset moves grades. Their summary: “overall effects were weak for both.” The exception they found worth flagging: students who were poor or academically at risk might genuinely benefit.

2019. David Yeager and a large team published the best-designed trial of the lot in Nature: 12,490 ninth-graders in 65 randomly-sampled American schools, pre-registered, independently analysed, with two online sessions of about 25 minutes. The effect on grades for lower-achieving students: 0.10 grade points. Real, replicable, and roughly one-tenth of what the assembly promised. It also only showed up where other students backed the idea of trying hard things — a belief, it turns out, needs a room to live in.

2023. Brooke Macnamara and Alexander Burgoyne went further: 63 studies, 97,672 students, average effect d = 0.05, and when they kept only the six highest-quality studies, d = 0.02 — statistically indistinguishable from nothing. They also reported that researchers with a financial interest in the result published larger effects than researchers without one.

Also. Yue Li and Timothy Bates tested 624 Chinese children aged 9–13 and found no link between mindset and resilience to failure, and none with grades.

Dweck herself, in 2015, wrote the sentence that the posters never printed: we are all a mixture of fixed and growth mindsets, and probably always will be. She also warned about praising effort when the effort isn't working, which she called a false growth mindset — and pointed out that trying harder at a strategy that has failed nine times is not growth, it is stubbornness in a nicer jumper.

Instrument six

The thing you'd try if failing at it were allowed

The 1978 children who kept going weren't braver. They were, for whatever reason, playing a game where being stuck was information rather than a verdict. You can't install that by wanting it. But you can notice what you'd do if it were true.

The graded part

Eight questions — and these ones do have right answers

Everything above this line was yours and unscored. Everything below it is about the 1978 room, the people in it, and the arguments since. Only this part has a mark.

Take it with you

Your page

Everything you wrote, in one block, in your own words. Nothing is stored in this browser and nothing leaves this page unless you take it.

Where the facts come from

Sources

1. Diener, C. I. & Dweck, C. S. (1978). An analysis of learned helplessness: Continuous changes in performance, strategy, and achievement cognitions following failure. Journal of Personality and Social Psychology 36(5), 451–462 · doi.org
2. Diener, C. I. & Dweck, C. S. (1980). An analysis of learned helplessness: II. The processing of success. Journal of Personality and Social Psychology 39(5), 940–952 · doi.org
3. Dweck, C. S. & Leggett, E. L. (1988). A social-cognitive approach to motivation and personality. Psychological Review 95(2), 256–273 · doi.org
4. Dweck, C. S. (1975). The role of expectations and attributions in the alleviation of learned helplessness. Journal of Personality and Social Psychology 31(4), 674–685 · doi.org
5. Dweck, C. S. & Reppucci, N. D. (1973). Learned helplessness and reinforcement responsibility in children. Journal of Personality and Social Psychology 25(1), 109–116 · doi.org
6. Seligman, M. E. P. & Maier, S. F. (1967). Failure to escape traumatic shock. Journal of Experimental Psychology 74(1), 1–9 · doi.org
7. Mueller, C. M. & Dweck, C. S. (1998). Praise for intelligence can undermine children's motivation and performance. Journal of Personality and Social Psychology 75(1), 33–52 · www.columbia.edu
8. Krakovsky, M. (March 2007). The Effort Effect. Stanford Magazine (the “I love a challenge” boy) · stanfordmag.org
9. Dweck, C. S. (November 2014). The power of believing that you can improve. TEDxNorrköping (the “Not Yet” grade) · www.ted.com
10. Dweck, C. S. (September 2015). Carol Dweck Revisits the “Growth Mindset.” Education Week · lfonetwork.uoregon.edu
11. Sisk, V. F., Burgoyne, A. P., Sun, J., Butler, J. L. & Macnamara, B. N. (2018). To what extent and under which circumstances are growth mind-sets important to academic achievement? Two meta-analyses. Psychological Science 29(4), 549–571 · doi.org
12. Yeager, D. S. et al. (2019). A national experiment reveals where a growth mindset improves achievement. Nature 573, 364–369 · doi.org
13. University of Texas at Austin (2019). A growth mindset intervention can change students' grades if school culture is supportive (12,000+ students, 65 schools, two 25-minute sessions, 0.10 grade points) · stat.utexas.edu
14. Macnamara, B. N. & Burgoyne, A. P. (2023). Do growth mindset interventions impact students' academic achievement? A systematic review and meta-analysis with recommendations for best practices. Psychological Bulletin 149(3–4), 133–173 · pubmed.ncbi.nlm.nih.gov
15. Burnette, J. L. et al. (2023). A systematic review and meta-analysis of growth mindset interventions: For whom, how, and why might such interventions work? Psychological Bulletin 149(3–4), 174–205 · doi.org
16. Barshay, J. (5 December 2022). Proof Points: Does growth mindset matter? The debate heats up with dueling meta-analyses. The Hechinger Report · hechingerreport.org
17. Li, Y. & Bates, T. C. (2019). You can't change your basic ability, but you work at things, and that's how we get hard things done. Journal of Experimental Psychology: General 148(9), 1640–1655; summary: University of Edinburgh, “Pupils do not benefit from so-called growth mindset, study shows” · www.build.edinburghneuroscience.ed.ac.uk