Video feedback for young archers: Arisman 2024

Sixty archers under the age of eighteen were split into two groups of thirty. The first alternated sessions in the field with online sessions reviewing their own video with the instructor, the second carried on with its usual training. On the 40 metre test the video group went from 18.30 to 22.53 points on average (standard deviations 2.63 and 2.85), a mean gain of 4.23 points with a t of -15.94 on 29 degrees of freedom, p below 0.001. The control group went from 19.10 to 19.37, a difference the test does not separate from chance (p = 0.133). At the end the gap between the two groups was 3.17 points, with a 95% confidence interval from 1.68 to 4.66.

What those points are, however, the paper does not say. It gives neither the number of arrows in the 40 metre test nor the maximum score, so a gain of four points stays a figure without a scale: there is no telling whether it is worth one arrow or a tenth of an end. There is also an inconsistency about the design, which the text calls a single-group pretest-posttest while the results present two groups and a comparison between independent samples.

The participants were sixty athletes of no more than seventeen years of age preparing for an under-18 competition, taken as a whole from a single setting and divided into two groups of thirty. The programme ran from February to April 2022, with fourteen training meetings plus the two testing sessions, seven in the field and seven online according to the count given in the abstract. At each meeting the athletes filmed their own shooting from several angles, the instructor analysed the videos and gave feedback, and the first recording served as the reference for later comparisons. The control group followed what the paper calls conventional training, about which it says nothing.

The two tests, before and after, were shot at 40 metres with the same procedure, since that is the distance of the national elimination rounds the authors take as their reference. The analysis, run in SPSS 26, checks the normality of the data with Shapiro-Wilk and Kolmogorov-Smirnov and the homogeneity of variances with Levene, then compares before and after within each group with a paired t test and the two final scores with an independent samples t test. It does not state how the sixty were assigned to the groups, the sex and years of practice of the athletes, the discipline and the equipment, the number of arrows and the maximum of the test, who recorded the scores, or the wind conditions.

The more serious point is that the comparison does not isolate the video. On one side were fourteen structured meetings, with an instructor watching the footage and giving feedback one point at a time, on the other the usual training, which the paper does not describe: the difference in score may come from the extra hours, the extra attention or the structure of the programme as much as from the archers watching themselves. The allocation to groups is not explained and no randomisation is claimed; the two groups already differed at the start, with the controls 0.8 points higher; and it is not known whether the people recording the scores knew which group each athlete was in.

The figures carry a few printing slips: in the final comparison table the Levene test appears as 114 and 737 instead of 0.114 and 0.737, and the t for unequal variances as 0.252 instead of 4.252. The discussion, for its part, never returns to the data: it repeats general statements about blended training and video without tying them to the results obtained, and uses very broad language for a gain of four points on an undeclared scale. Nothing about technique is measured, although the text speaks of improved technique: only the score changed. There is no follow-up after the two months, so whether the difference holds is unknown. The journal is flagged among those with questionable editorial standards.

The study is about boys and girls under eighteen preparing for a competition, not about formed archers, and its measure is the score at 40 metres. What it says is that two months of work in which half the sessions were spent reviewing their own footage ended with more points than two months of the usual training, while the group that carried on as before did not improve in any way distinguishable from chance.

By how much cannot be carried anywhere else: without the number of arrows and the maximum of the test, the 3.17 point difference does not translate into anything comparable. Nor can the result be put down to the video, since the two groups were not doing the same amount of training with a camera added. For anyone coaching, what remains is an observation of its moment, from the period when training had moved online: a plan that alternated field sessions with video analysis worked at least as well as the previous one, in a group of young archers, over a single season and with no follow-up.