Static strength training in archers: Sharma 2015 study

Thirty male archers aged 17 to 24 from a coaching centre in Amravati, India, were split into two groups of fifteen, matched on the mean of their baseline results. One group followed six weeks of strength exercises, the other did no specific training. The two measures are pull-ups, counted in repetitions, and back strength on a back and leg dynamometer.

At the end the trained group managed 13.0 pull-ups against the control group's 10.47, and scored 135.17 on the dynamometer against 110.87: two differences that the between-group comparison reports as significant. The control group stayed flat, from 10.27 to 10.47 pull-ups and from 111.57 to 110.77 on the dynamometer.

Within the trained group, however, neither change comes out as significant: pull-ups go from 12.0 to 13.0 and back strength from 110.87 to 135.17, and for both the authors report a value below the threshold. Putting the two together, the finding of the study is that at the end of the six weeks the two groups differed, while the trained group's own improvement did not clear the statistical test. The abstract, by contrast, asserts a significant improvement in arm and back strength.

No measure concerns shooting: no arrows, no score, no accuracy. The paper does not describe the exercises used over the six weeks.

All thirty archers come from the Shree Hanuman Vyayam Prasarak Mandal archery coaching centre in Amravati. The paper says they were selected at random, but the two groups were formed by matching the means of the pre-test results, so the allocation is not a draw.

Arm strength is measured with pull-ups: the bar is raised so that the tallest participant's feet do not touch the ground, the grip is forward, the chin must clear the bar and the arms must return to full extension; anyone who does not extend, kicks or uses a kipping motion is credited half a repetition, and at most four half counts are allowed. Back strength is measured with the back and leg dynamometer following the Rogers test: standing, trunk flexed ten to fifteen degrees forward at the hip, one hand above and one below the bar, the bar set just below the fingertips with the arms hanging in front of the thighs, hands shoulder width apart, feet about six inches apart, knees and back kept straight, a steady upward pull with no jerking and no leaning back on the heels; two or three attempts are allowed and the best is recorded.

The experimental group followed six weeks of exercises; the control group did no specific training. After the six weeks both tests were repeated. The analysis is a t-test between the pre and post means of each group and between the two groups.

What the exercises were, at what load, how often a week and for how long per session, the paper never says: it refers only to "selected exercises" for arm and back strength. Nor is there any record of what bows the archers shot, what level they were at, or what they did in their normal sessions during the six weeks.

Back strength is reported with no unit. The method says the dynamometer score is recorded "in pounds or kilograms", and nowhere is it specified which: the values between 110 and 135 are therefore numbers without a magnitude. The inventory record gave them in kilograms, but the text does not.

The numbers in the tables do not fit together. The control group's post-test for back strength appears as 110.77 in one table and as 110.87 in the final comparison; 110.87, however, is the experimental group's pre-test value, so the difference of 24.30 on which the main result rests may be calculated against the mismatched figure. In the between-group comparisons the authors use a critical value with fourteen degrees of freedom, when two groups of fifteen people give twenty-eight degrees of freedom and a lower threshold. One table reports a value of 1.6 and the text immediately below it reports a different one, 1.012. Above all, the tables for the experimental group are marked as not significant, and the discussion two paragraphs later speaks of significant improvement in arm and back strength.

There is also an imbalance at baseline: on pull-ups the experimental group started at 12.0 and the control group at 10.27, a difference very close to the significance threshold before any training. The final result on pull-ups, 13.0 against 10.47, is largely that initial gap.

The exercises are missing, so the study cannot be repeated: there is no way to know what produced the effect, if there was one. It is not stated whether the assessors knew which group each archer belonged to, there is no follow-up, there are no standard deviations on the gains, and the sample is thirty young men from a single centre. Accuracy was not measured: the study allows nothing to be said about what strength does to a score. The journal has a small readership and peer review is not stated; the bibliography of nine items consists entirely of general coaching manuals and references to football, without a single piece of work on archery.

For someone who shoots this study moves nothing. It says that six weeks of some strength exercises can leave a group of young archers stronger than a group that did nothing, which was foreseeable, and it says so with figures that do not survive close inspection. It does not say which exercises, so nobody can repeat them; it does not say whether the strength gained changed how they shot, because no arrow was shot in the measurements.

In the registry it counts as an example of a before-and-after study with a control group that stopped halfway: the difference between the two groups at the end is there, the direction is the one you would expect, and everything else (what training, at what loads, with what effect on the shot) stays outside the paper. Anyone who finds this work cited in support of the idea that back strength adds points is reading more into it than was measured.