Chess and school success: what the studies really say

Parents and teachers often read that chess would improve marks, concentration or even IQ. These claims sometimes rest on real studies, but simplify them a great deal. To form a fair view, you need to look at what each study measured, on how many pupils, for how long, and with what method.

Transfer of learning

The idea that a skill acquired in one area (here, playing chess) improves performance in another (mathematics, reading, general reasoning). It is precisely this transfer that studies on chess in schools try to measure.

Chess and school success: table of the main studies

The Danish study “Your Move” (2017)

Michael Rosholm, Mai Bjørnskov Mikkelsen and Kamilla Gumede followed 482 children at 5 schools in Aarhus, in 34 classes from 1^st to 3^rd grade. In 323 of them (the treatment group), one of the four weekly 45-minute maths lessons was replaced, from spring to mid-October 2013, by a chess-based maths lesson, around 30 lessons in all. The other 159 pupils followed their usual lessons.

Result: an estimated gain of 0.10 to 0.18 standard deviations on the maths tests, mainly on pattern-recognition exercises. The effect is concentrated among children who said they were unhappy or bored at school; it is zero among those who are never bored.

Limits acknowledged by the authors: the classes were chosen by head teachers rather than drawn by lot; the sample was too small to detect with certainty an effect below 0.23 standard deviations; a single teacher, a club player, gave the lessons, sometimes in the presence of the usual teacher; the significance disappears in some models. The authors describe their work as a pilot study and call for a randomised trial.

The Sala and Gobet meta-analysis (2016)

Giovanni Sala and Fernand Gobet pooled 24 studies (40 effect sizes) that compared pupils who had taken chess lessons with control pupils, a total of 2,788 young people in the “chess” groups and 2,433 in the control groups. They obtain a moderate overall effect (g = 0.338), a little stronger on mathematics (0.382) than on reading (0.248), and a clearer effect when the teaching lasts a long time; they put the useful threshold at around 25 to 30 hours.

The decisive limit: almost none of the studies analysed compared the chess-playing pupils with an “active” control group, that is, one doing another new and stimulating activity. It is therefore impossible to rule out that the effect comes from novelty, from the enthusiasm of the instructor or from the attention paid to the pupils (a placebo effect), rather than from chess itself.

The English randomised trial by the EEF

This is the most robust study to date. Funded by the Education Endowment Foundation and carried out by a team from UCL Institute of Education (John Jerrim, Lindsey Macmillan, John Micklewright, Mary Sawtell and Meg Wiggins), it randomly selected 100 English primary schools. In the schools of the “chess” group, all Year 5 pupils (aged 9–10) followed a standardised 30-hour programme, led by instructors from the charity Chess in Schools and Communities, during the 2013–2014 school year.

The main outcome was the mark in the national end-of-primary maths exam (Key Stage 2), marked by external examiners and taken a year after the lessons ended. Of 4,009 pupils, results are available for 3,865 (97%). Result: an effect close to zero, with a confidence interval running from about −0.15 to +0.15 standard deviations, and no effect either in reading or in science. The authors conclude that rigorous methods are needed before benefits are claimed.

What this study does not say: it measures a lasting effect on exams, a year later. It says nothing about possible immediate effects, nor about social benefits or the pleasure of learning, which the authors themselves mention as possible avenues.

The Italian study by Trinchero and Sala (2016)

Roberto Trinchero and Giovanni Sala compared 931 primary school pupils split into three groups: two groups taking chess lessons and one control group. The group whose instructors explicitly taught chess-specific problem-solving strategies got better results in mathematical problem solving than the other two. For the authors, it is the explicit teaching of these strategies, more than the game alone, that favours transfer.

What can reasonably be concluded

1. No miracle effect. The most rigorous study finds no lasting effect on exams.
2. Possible but modest effects, mainly in mathematics, mainly in the short term and above all when the game is used to do mathematics.
3. How it is taught matters: explicit strategies, a link with the subject, sufficient duration.
4. Some pupils benefit more, such as those who are bored at school in the Danish study.
5. Chess has other merits than marks: enjoyment, socialising, respect for rules. It deserves to be judged on these criteria too.

Does chess make you more intelligent?

Good players are often seen as very intelligent, but correlation is not causation: children drawn to chess may already have natural strengths. Controlled studies do not show that learning chess raises general intelligence. Le Palais des Échecs reviews this received idea in an article entitled does chess really make you more intelligent?, and also asks the opposite question: do you need to be intelligent to play chess?

My child is struggling at school: can chess help?

It can be part of the answer, not replace it. A child who is bored or discouraged may rediscover the taste for searching, thinking and succeeding through the game, and the Danish study points that way. But an established difficulty in reading or mathematics first calls for a conversation with the teacher and, if need be, suitable support. Chess is a motivating complement. On the school side, see chess in schools in France.

For the related question of non-academic benefits (concentration, patience, self-confidence), see the benefits of chess for children, and for the precise link with mathematics, go back to the Danish study and other leads on our chess and maths page.