Psychological questionnaires

What a questionnaire measures

A psychological questionnaire does not measure an archer's anxiety or confidence: it measures what the archer says about them, at that moment and in the words the form puts in front of them. Almost the whole of the registry's section on the mind rests on this method, and it is also why so many entries in that section stop at "reported": under the registry's rules, an association between answers collected at a single point in time goes no further.

Trait questionnaires ask how a person usually is: personality, habitual anxiety, mental skills. State questionnaires ask how they feel right now. The first large study, of 188 American recurve archers, relied mainly on trait scales: anxiety and attentional style did not move with the archers' average yearly scores, while confidence did, with r = 0.32.

The scales that keep coming back

For competition anxiety the most common instrument is the CSAI-2, together with its revised version, the CSAI-2R. It gives three scores, cognitive anxiety, somatic anxiety and self-confidence, and is filled in shortly before shooting. It already appears in a Greek study from 1991 on visualisation, and since then in more than ten papers, from Greece to Iran. The Sport Anxiety Scale-2 splits anxiety into worry, concentration disruption and somatic anxiety instead, and it is the instrument behind the two Malaysian samples, 44 and 207 archers, on which the entry on worry before competition rests.

For mental skills the usual choice is the ACSI-28, seven scales of four items each, and less often the OMSAT-3 and the PSIS; for mood, the POMS and its short form, the BRUMS; for personality, the Big Five scales and the sensation-seeking scale, used in comparisons between archers, drag racers and golfers. Alongside these sits a long tail of questionnaires built for the occasion, often in regional journals, without their items or any reliability measure.

When it is filled in

Timing matters as much as the instrument. The entry on anxiety and score is "contested", and much of the split follows the moment of measurement: the link appears when anxiety is asked about shortly before a competition, with correlations between -0.45 and -0.7 in Malaysia, Turkey and Brazil and -0.55 in the largest sample, and it disappears with trait scales or with mood on training days.

A Turkish study of eighteen national team archers shows why: reported cognitive anxiety was 19.22 before a training session and 19.27 before a national competition, and a measure that does not change cannot show up in the score. Some questionnaires are even completed afterwards: the twelve Hong Kong archers in the mindfulness programme completed the CSAI-2 once the competition was over, and among the Korean university archers, in the lower-scoring group a lower score went with more anxiety even half an hour after the end, r = -0.72 in nine people. A state questionnaire can therefore only be read once you know when it was completed relative to the arrows.

When the scale is too narrow

The entry on short questionnaires grew out of two Indonesian studies. In the first, thirty-two young archers answer twenty-four yes-or-no questions: thirty-one land in the top band, all of them reach the maximum on adaptation, with a standard deviation of zero, and nobody comes out as lacking in anything. In the second, 269 archers at competitions complete a mental toughness questionnaire and every score falls between 32 and 40, with standard deviations between 1.28 and 1.93; the paper does not state the maximum possible score, and a difference of barely one point between distances becomes significant only because of the number of people.

Narrow does not mean blind, though. Among seventy elite Greek archers the ACSI-28 separates those with six or more years of competition from those with one or two: concentration 5.4 against 4.7, confidence 4.7 against 3.9. Its scales have an internal consistency between 0.60 and 0.76, which makes it a noisy measure: it separates groups somewhat, individuals much less. The entry on years of competition rests on this single study.

A link does not give a direction

The entry on confidence shows the method's typical limit. Confidence goes with score in almost every study that places the two side by side, with coefficients between 0.24 and 0.70, but always in the same design: the two measures are taken once, at most a week apart, and correlated across different archers. No one has recorded confidence first and score afterwards across several competitions by the same archer, and no one has tried to raise confidence to see whether the points follow. In the Indian comparison of twenty medallists and twenty non-medallists, where the scale is reversed and fewer points mean more toughness (4.45 against 14.20 out of 24), the questionnaire was completed after the medals had been awarded, and winning may make people more confident rather than the other way round. This is why the entry on medallists' mental toughness stays at "reported".

The same applies to path models. Among the 115 Italian archers of a study on choking, the direction of the arrows was set by the researchers, and no competition score was collected. And in the comparisons between sports behind the entry on personality and mental profile, nobody asks whether character comes before archery or after it.

Figures that do not add up

Some studies report results that their own figures cannot support. In the Indonesian mental toughness study of 269 archers, the effect on accuracy comes from a t-test on the difference between the questionnaire score and the arrow score, two different scales. Another study by the same group publishes standardised coefficients of -1.14, -2.04 and -2.21, which in a regression with a single predictor cannot exceed 1 in absolute value. Among eighty Indian juniors a correlation of -0.156 is declared significant, which with eighty people it cannot be. In the study at the 2011 Asian Championships on women archers and anxiety, the reported effect size, d = 1.84, is about twice what the paper's own table gives. The registry keeps their direction and drops their size, and a questionable venue never decides a status.

There is also a design flaw that recurs in the Malaysian studies on coping skills: the shooting score goes into the clustering that is then compared, so the "high" group scores better on the questionnaire partly because that is how it was formed.

What to check

• which scale, whether it is established and validated in the respondents' language, and how consistent it is internally;
• whether it measures a trait or a state, and when it was completed relative to the arrows;
• how widely the scores spread: minimum, maximum, standard deviation and the highest possible score;
• whether cognitive and somatic anxiety are read as separate measures, when in archers they move together with correlations as high as 0.868;
• whether the competition score was also used to form the groups being compared;
• whether the conclusion claims a direction that a cross-sectional design cannot provide.

How design and sample are weighed is explained in How we read a study; the body measures that often accompany these questionnaires, heart rate and its variability, in How archery studies measure heart rate and HRV. How mental skills are trained is the subject of the article on mental skills training.

@misc{campagna2026_enresearchmethodspsychologicalquestionna,
  author = {Campagna, Matteo},
  title = {Psychological questionnaires},
  year = {2026},
  howpublished = {Research, matteocampagna.com},
  url = {https://matteocampagna.com/en/research/methods/psychological-questionnaires},
  note = {accessed 2026-10-16}
}