Archery and executive function: Liu 2023 study

In 72 children in Shanghai, a 45-minute session of traditional Chinese archery led to immediately better results in three tests of executive function. The children, in their fourth year of primary school and aged 9 to 11, were split into 36 in the archery group and 36 in a control group doing the usual after-school activity: 10 minutes of walking on the track followed by homework in the classroom.

Straight after the session the archery group responded faster and more accurately in three computer tasks: one of inhibition (a fish Flanker task), one of working memory (N-back) and one of cognitive flexibility (DCCS). Comparing the groups after adjusting for baseline, the differences favoured archery on 11 of 14 measures, with partial eta squared between about 0.06 and 0.22, which by the usual benchmarks means medium to large effects. The largest concerned accuracy on the hardest memory task, the 2-back: it rose from 70.7% to 75.7% in the archery group and fell from 67.4% to 62.4% in the controls. Speed on the 1-back and accuracy on the two simpler flexibility tasks showed no difference.

The study is quasi-experimental: the children were allocated to the groups by intention rather than at random, and the groups were unbalanced by sex (23 boys and 13 girls in the archery group, 15 and 21 in the controls) and differed in height. The archery group had already had five lessons and knew the basic technique and etiquette. The session lasted 45 minutes outdoors, led by an experienced instructor: 5 minutes of stretching, 3 of mental preparation, about 34 of shooting, 2 of meditation and 1 of deep breathing, with the instruction to focus on the movement rather than on hitting the centre.

The tasks were run on a computer (E-Prime) in a quiet room, once before the session and again within 10 minutes of the end, after a few minutes' rest. For each task the authors recorded the reaction time of correct responses and the percentage of correct responses, then compared the groups with an analysis of covariance that takes the starting result into account.

The authors themselves note that the design is not randomised, that exercise intensity was not measured and that a real-world setting brings other uncontrolled factors. The comparison also does not isolate archery: one group had an outdoor session with an instructor, meditation and breathing, the other walking and homework, which at the end of the school day can be tiring. It is not known whether the person running the tasks knew which group each child was in, and the effect was measured only minutes later, with no follow-up in the days after.

Some of the statistics do not add up: the degrees of freedom reported (1 and 36) do not match 72 children, and for reaction time on the 2-back the table and the text give two different effect sizes (0.01 and 0.06), and for accuracy on the colour and shape tasks the text swaps the table's values. The outcomes are cognitive: the study measures nothing about shooting, neither accuracy nor score.

The study looks at archery as a school activity for children who are learning it; it says nothing about how archers perform. It shows that a lesson of traditional archery, done calmly and with attention to the movement, is followed by better results on tasks of attention, working memory and flexibility than an afternoon of walking and homework. The effect is short-term and was obtained without random allocation, so it does not allow anyone to say that archery improves children's attention over time, or that it does better than other activities with similar features.

Anyone teaching young children can take one thing from it: a lesson of archery did not dull their concentration, and for a few minutes afterwards they responded faster and more accurately. In the registry the study counts as evidence on beginners who practise archery as an activity, which is the population the claim about attention refers to.