Push-ups and pull-ups for archers: Yachsie 2024

Twenty archers aged 16 to 18, selected from the thirty-eight in a setting in Banyumas Regency, were split into four groups of five and spent eighteen sessions doing either push-ups only or pull-ups only. The split is a double one: first by arm endurance measured at the start, taking the top 27% and the bottom 27%, then by type of exercise.

The outcome is a measure of arm strength on the bow, the Holding Bow Digitec Test, expressed in kilograms. In the table, push-ups with high endurance go from 18.00 to 24.59, pull-ups with high endurance from 17.00 to 19.87, push-ups with low endurance from 16.00 to 17.75 and pull-ups with low endurance from 15.00 to 16.93. The two-way analysis of variance gives an effect of method (F 7.738), an effect of starting endurance (F 51.203) and an interaction between the two (F 51.203 as well), all with p = 0.000 in the table.

The authors' conclusion is the interaction: push-ups would suit those who start with high endurance and pull-ups those who start with low endurance. The Tukey post hoc test finds push-ups with high endurance different from the other three cells, by 2.20, 2.60 and 2.20 kg, and gives a difference of exactly zero between pull-ups with high endurance and pull-ups with low endurance.

The design is a two by two factorial: training method, push-ups or pull-ups, by starting endurance level, high or low. The thirty-eight archers in the setting were assessed on an endurance test, from which the top 27% and the bottom 27% were taken, ten athletes each; each half was then split into two groups of five by ordinal pairing, that is by distributing the ranked results in pairs. The stated inclusion criteria are elite archery athletes aged 16 to 18.

The training ran for eighteen sessions, with intensity from 70 to 80% of maximum, eight to ten repetitions, three to four sets and three minutes between sets. The measuring instrument is the Holding Bow Digitec Test, which according to the authors gives arm strength in kilograms and arm endurance in seconds; for the instrument they quote a validity of 0.961 for strength and 0.895 for endurance against a table value of 0.344, and a Cronbach's alpha of 0.978 and 0.944. The procedure has three steps: an endurance test to divide high from low, a baseline test of strength and endurance, and a final test after the eighteen sessions.

The analysis is a two-way analysis of variance, with preliminary tests of normality (Kolmogorov-Smirnov) and homogeneity (Levene) and, where an interaction appears, Tukey comparisons, in SPSS 20, with a 5% threshold. The push-up and pull-up programmes are shown as two figures, which the text version of the article does not reproduce.

The first problem is that the numbers contradict each other inside the article. The text commenting on the table describes means going from 16.00 to 21.67, from 16.00 to 17.67, from 14.00 to 15.83 and from 14.00 to 17.83, while the table prints 18.00 to 24.59, 17.00 to 19.87, 16.00 to 17.75 and 15.00 to 16.93. The Tukey differences, 2.20 to 2.60 kg, do not match the gaps in the table, which run from 4.7 to 7.7, and one of the pairs is given as a difference of exactly zero between groups that sit almost three kilograms apart in the table. The interaction has the same sum of squares and the same F as the main effect of endurance, which two distinct terms in a model cannot have. The English abstract gives p = 0.027 for the effect of method where the table gives 0.000, and the Ukrainian and Russian abstracts give F 5.671 and 41.107 where the English gives 7.738 and 51.203.

The column for endurance in seconds has a heading but no values: the variable used to split the groups is never reported. That leaves five athletes per cell, no control group, since everyone trained, and a split made on baseline values, which on its own pulls the extreme groups back towards the mean even without any training.

The study measures nothing about shooting: no score, no accuracy, no sway. The word elite in the selection criteria sits awkwardly with the average of 8.3 that the introduction reports for these athletes on a hanging test, described as below standard. The text, finally, is a machine translation that is corrupted in places: the conclusion speaks of arms and legs and of resistance and stretching exercises where the content is about push-ups and pull-ups.

For an archer the content of the study is this: eighteen sessions of push-ups only or pull-ups only, in boys of sixteen to eighteen, followed by a stronger hold on the bow as measured by a holding test. There is no data at all on where the arrows ended up, so the source says nothing about scores.

Its most interesting idea, that the exercise should be chosen according to where each athlete starts, is exactly the one the published numbers cannot support: the post hoc differences do not match the means in the table, and the variable used to split the athletes is not reported. In the registry the source counts as a marker of a question, not as evidence of a differentiated effect.