Psychology and physio for archers: Dhillon 2016

Thirty Indian archers at national and international level, 17 women and 13 men, were split into two groups by the toss of a coin. One followed a six-week package combining mental training and strengthening of the shoulder muscles, the other carried on with the coach's usual programme. Measurements were taken at the start, on day 21 and on day 42: three questionnaires (movement imagery, mental skills, competitive anxiety), a shooting test at 50 metres and the strength of the shoulder muscles.

The authors report significant improvements on every measure in the treated group; of the controls they say only that anxiety did not change. The figures in the paper, however, cannot be read as scores. For the shooting test the text gives a "mean value" of 56.969 in the treated group and 3.670 in the controls, for the mental skills questionnaire 50.841 and 0.932, and for anxiety 50.041. The anxiety questionnaire runs from 10 to 30 and the mental skills one from 20 to 100, so these values cannot be means: they look more like test statistics. Strength is reported in the same way, with values between 7.258 and 54.330 in the treated group and between 0.715 and 3.418 in the controls. By how many points the shooting changed, the paper does not say.

Archers were eligible if they were aged 18 to 25 (mean age 21.93), shot recurve or compound, had been with the team for at least a year and had no upper limb injury. The treated group worked for an hour a day, five days a week. Mornings were given to mental training: imagining the scene before a competition, a modified version of Jacobson's progressive relaxation, rational emotive behaviour therapy sessions on pre-competition anxiety, and attention control exercises with thought stopping and positive self-talk. Evenings began with 20 minutes of warm-up (running, walking, stretching) followed by dumbbell exercises: lateral, front and rear raises, military press, upright row, shrugs, single-arm rows and side-lying internal and external rotation.

The shooting test is a version of the AAHPER test (spelt AAPHER in the paper): 4 practice arrows and then 12 arrows in two ends of six, all at 50 metres including the compound archers, with rings worth 9, 7, 5, 3 and 1 points, for a maximum of 108. Strength was measured with a digital strain gauge. The sample size calculation called for 30 archers per group; 30 were recruited in all, and the text does not say how many ended up in each group.

The main problem is that the paper does not report its results in a usable form: the means and standard deviations of each measure at the three time points are missing, and the values called means are incompatible with the scales used. For movement imagery, too, the treated group has two values (12.93 and 13.47) for three time points, and it is unclear which they belong to. Charts of the shooting scores are included, but without figures.

Even with clear results the design would have its limits. The treated group received five extra hours a week of attention and work and the control group nothing, so the difference may come from the time spent rather than the content. Mental training and strengthening were given together, so there is no telling which of the two mattered. The questionnaires were filled in by the archers, who knew they were in the treated group, and the paper does not say whether the assessors knew the groups. Recurve and compound archers shot the same test, at 50 metres with five scoring rings, and the results are not separated by bow type. The sample is half of what the sample size calculation required. The journal has questionable editorial standards.

For an archer the study points in a direction without giving a measure. According to the authors, a six-week package of imagery, relaxation, anxiety management, attention control and shoulder strengthening was followed by better scores than usual training alone. By how many points cannot be said, and there is no knowing which part of the package counted.

That mental skills and shoulder strength belong in an archer's preparation is a widely held view, supported by other work; this paper adds no evidence that can be weighed. In the registry it counts as a pointer, not as evidence on scores.