Attention in young archers: Korobeinikova 2017

Sixty-five Russian archers aged 13 to 19 completed four computer tasks on attention and memory, on competition days and after shooting. There were 34 boys and 31 girls, with a mean age of 16.3 years and between 1 and 8 years of practice (4.5 on average); 36 shot recurve and 29 compound, and 39 of the 65 held the Russian Master of Sport title. The study measured nothing about shooting: no scores, no arrows.

The finding the authors put at the centre is the comparison between the 29 selected for the national youth team and the 36 who competed only at national or regional level. On the third and last series of the red-black tables, the task that asks you to find numbers while alternating between two colour sets, the squad members made fewer mistakes (2.38 against 3.50) and were quicker on the reaction index (4.26 against 4.79). These are the only two differences that emerged between the groups, and only in that last series.

Everything else is flat, or goes where the authors did not expect. No difference emerged between recurve and compound, nor between the groups split by years of practice. Between boys and girls the difference lies in the Landolt rings: the boys finished the task in 493 seconds against 564, at a pace of 117 symbols per minute against 101 and with a higher concentration index (104.6 against 89.1), which amounts to a minute less work on average. Two measures get worse with age: working memory productivity falls from 0.91 at 13 to 0.69 at 19, and the drop in performance when the squares to be checked go from five to twenty is sharper in the older archers (3.40 against 2.85 in the younger ones). Sporting rank, by contrast, explains nothing, which the authors put down to the Russian rank system not tracking an archer's actual level.

The four tasks were run on a computer, on a 12-inch monitor about 50 centimetres from the eyes, using the Praktika-MSU software from Moscow State University; responses were given with the mouse, except in the change blindness task, which used the keyboard. The order was always the same: working memory first (geometric figures), then change blindness, then the red-black tables, and the Landolt rings last. Each archer took about thirty minutes in all.

The red-black tables measure the distribution and switching of attention; the Landolt rings measure productivity, meaning how many symbols are checked and how accurately; the change blindness task uses the flicker paradigm, in which two images alternate and the participant has to spot what changes, and by varying the number of squares from five to twenty it also gives a measure of how much performance falls as the load rises.

Testing took place on competition days, after the archers had shot: the squad members and some of the others during the Russian youth championships, the rest during regional competitions. Access was arranged through the regional and personal coaches, and participation was voluntary. Group comparisons use the Mann-Whitney test for sex, age, bow type and level, Kruskal-Wallis for years of practice, and a one-way analysis of variance for the effects of age, practice and sporting rank. The study comes out of a contract with the Ministry of Sport of the Russian Federation.

There is no comparison group from outside archery. The authors' conclusion, that archers have very stable attention, rests on the fact that the task placed last was also the one done best (0.89 on a 0 to 1 scale): that is an argument, not a comparison with anyone else.

The level of professional success is membership of the national youth team, not a score. Nobody shot an arrow inside the study, so the link between attention and performance falls outside the design. The differences found are two out of many measures compared, concentrated in the last series of a single task, and the authors do not correct for the number of comparisons: with 65 people split into groups of 29 and 36, p values around 0.02 should be read cautiously. Testing happened immediately after competing, a moment when fatigue and the outcome of the competition can both change performance, and it was not repeated at rest for comparison.

Two things in the paper do not add up. In Table 1 the Russian heading puts the girls first and the boys second, while the English heading does the opposite: the numbers are the same, the attribution is reversed, and this record follows the Russian text, which states explicitly that it is the boys who have the higher pace. In the discussion, the authors then note that the Landolt rings have their gaps at four and eleven o'clock, the same coordinates archers use to describe where an arrow landed, and suggest this familiarity helped them on that task: it is a conjecture, not something they tested.

For someone who shoots, the study says little directly, and one interesting thing in the background. It measures no scores, so it does not show that better attention produces more points, nor that training attention on a computer does anything on the shooting line. The only difference tied to level concerns the ability to hold several things at once, and it shows up only in the most loaded part of the task: the youth squad archers coped better with the last series.

The absence of any difference between recurve and compound is worth knowing, because it takes weight off the idea that the two disciplines call for different kinds of attention, at least in the tasks used here. More curious is the picture around age and years of practice, because it runs opposite to what anyone expecting practice to sharpen attention would predict: with age, working memory productivity falls, and performance drops further as the number of things to monitor rises. The authors read this as a narrowing of attention produced by years spent concentrating on few stimuli, and on that basis they recommend training the distribution of attention as part of mental preparation. That is their reading of a cross-sectional association: measuring archers of different ages once each leaves a training effect, a generational difference and a selection effect indistinguishable.