Working memory and kyudo: Kitahara 2025 study

This is a one-page presentation at the 2025 congress of the Japanese Psychological Association. In 36 members of a university kyudo club (21 men and 15 women, mean age 19.42), the authors measured working memory capacity and hit rate over 20 shots, to see whether the two were related, along with years of practice.

They found no relationship. The correlation between hit rate and working memory was -0.05 (p = 0.76), and with years of practice 0.18 (p = 0.31). In a regression with both variables neither predicted hits: for working memory β = -0.04 (p = 0.85), for years of practice β = 2.03 (p = 0.32).

Working memory was measured with the Japanese version of the operation span test, a task that alternates simple sums with items to remember, in sets of 3 to 8 presented in random order, 99 problems in all: a test that demands divided attention and resistance to interference. Shooting was assessed with 20 shots at the 36 cm kyudo target, counting any arrow inside the target as a hit, as is normal in the discipline; the distance is not stated and testing took place in the usual training setting. Sex, age and years of practice were collected with an online form.

This is a poster summary, not a peer-reviewed article. Descriptive data are missing: the participant table mentioned in the text is not reproduced, and neither the mean hit rate nor mean years of practice are given. Twenty shots are few for estimating a shooter's precision, and scoring only whether the arrow is in or out of the target, with no information on how close to the centre it landed, discards a lot of information. With 36 people, only fairly strong relationships would have shown up.

The authors note further limitations: a single day of testing, in training and therefore under little pressure. They suggest working memory might matter more in competition, under pressure, but that is a question for future studies. The coefficient of 2.03 for years of practice is almost certainly unstandardised, meaning percentage points of hits per year, and alongside a confidence interval from -2.12 to 6.17 it should not be read as a large effect.

For anyone who shoots, the study makes a simple point: in a group of university kyudo archers, better working memory on a computer test did not go with more hits in training, and years of practice made no appreciable difference either. This fits the idea that in a highly repeated movement performance depends on other things, from technique to body stability, but the study does not measure them.

It says nothing about competition, where attention and handling pressure might weigh differently, or about archers in other disciplines. It is an early, small null result, presented by the authors themselves as exploratory.