Self-confidence and 50 m archery scores: Yachsie 2021

Twelve high-school archers from the Banyumas district of Indonesia, shooting in the standard bow division, filled in a twenty-item self-confidence questionnaire and then took a scoring test at 50 metres. The group's mean score is 205.33 with a standard deviation of 75.48: an enormous spread, which shows how different from one another they were.

From the questionnaire the authors derive four confidence categories and place the twelve archers in them: four come out very high (33.3%), two high (16.7%), three moderate (25%) and three low (25%). The simple linear regression between confidence category and score is significant, with an F of 73.120 on 1 and 10 degrees of freedom and a significance reported as 0.000. From the table's sums of squares, 55,123.869 for the regression and 62,662.667 in total, confidence works out as explaining about 88 per cent of the variation in scores.

The questionnaire is built on two families of items: the first concerns the moment of drawing the bow, from executing the technique with conviction to not rushing once you are already sure, through to the bow arm that must not drop before the arrow is in the target; the second gathers the problems the archers meet most often, namely hesitation at the moment of release, wind that makes the arm sway, doubt about the technique they have learnt, and the inability to stay still as the string goes. Internal consistency, measured with Cronbach's alpha, comes to 0.896.

The study is correlational, on a single group at a single moment. The twelve archers were chosen purposively, against four stated criteria: still being at secondary school, training regularly with competition in view, not being injured or under medical care, and agreeing to take part.

The sequence is simple: first the twenty-item questionnaire, then the scoring test at 50 metres. The validity of the items was checked with product-moment correlation and internal consistency with Cronbach's alpha, which comes to 0.896. Before the analysis came a Kolmogorov-Smirnov normality test (0.167 for confidence, 0.200 for the score) and a Levene homogeneity test (0.133, on 3 and 8 degrees of freedom). The hypothesis was then tested with a simple linear regression.

The choice of 50 metres is explained in the text. In the Indonesian federation's standard bow format archers shoot at 50, 40 and 30 metres and the three scores are added together; from the data of a regional POPDA competition the authors observe that at 40 and 30 metres their archers' scores sat between 300 and 312, while at 50 metres they fell to between 180 and 200. It is the distance where wind counts most and where, on their reading, the total is decided.

What the method does not state is how many arrows the scoring test comprised, in how many ends, or on which target face and in what conditions it was shot.

Twelve archers are few for a regression, and 88 per cent of variance explained across twelve people is a figure to treat with great caution: with a sample this small the explained share is inflated, and the paper reports neither the slope, nor the intercept, nor an adjusted R squared, nor any confidence interval.

There is also a question about what went into the regression. The descriptive table gives confidence a mean of 2.42 with a standard deviation of 1.240, and the homogeneity test has 3 and 8 degrees of freedom: these are the figures of the four-level category, not of the raw questionnaire scores, which are never shown. Confidence, then, appears to enter the analysis as an ordinal category from 1 to 4, and the link described is the one between four groups and their mean scores.

The design is cross-sectional: questionnaire and test on the same day, no intervention, no follow-up. There is no separating the confidence that makes someone shoot well from the confidence that comes of shooting well, and the archers knew their score was about to be recorded. A standard deviation of 75 points on a mean of 205 indicates a group spanning very different levels, and in such a group any measure connected with ability will correlate with the score.

Missing are the details of the scoring test and the provenance of the questionnaire, which the authors built themselves and which is not a published instrument: its items mix confidence with technique (the bow arm that must not drop) and with external conditions (the wind). In the introduction the test is once called an anxiety test, and the discussion contains a passage about the volleyball smash that has nothing to do with these data. The text is in Indonesian, with an English abstract.

Read for what it is, this source describes a group of twelve youngsters in which those who reported more self-confidence scored more points at 50 metres on the same day. It does not say that training confidence raises the score, because nobody trained anything, and the figure of 88 per cent of variance explained should not be carried outside this group.

The part worth reading is the questionnaire, not the statistics. The four items on recurring problems describe well how uncertainty shows up on the shooting line: you hesitate at the moment of letting the string go, you doubt the technique you have learnt, you cannot stay still as the arrow leaves, and wind magnifies all of it. These are archers' phrasings rather than psychologists', and they say that when confidence is missing it shows in the mechanics of the shot, not only in the thinking.

Also worth remembering is why the authors picked 50 metres: in their competition data the long distance of the standard bow round was losing around a hundred points compared with the two nearer ones. That is an observation on one district and one competition rather than a rule, but it frames why they looked there.