Imagery in air pistol shooting: Lee 2023 study

First thing to know: this is not a study on archers. The participants are five young Korean air pistol shooters, three boys and two girls aged 17 to 18, with three or four years of practice and no structured psychological skills programme behind them. It enters the register as a transfer from another aiming sport.

The five followed eight weeks of psychological skills training: one 50-minute group session a week plus 30 minutes of individual counselling, covering orientation, goal-setting, self-understanding, imagery across two sessions, relaxation, cognitive restructuring and routine. Before and after, an electrocardiogram recorded for six minutes in three conditions: quiet breathing, recall of a critical moment experienced in competition, and imagery of a perfect performance. Plus four questionnaires.

Of the many values measured, four change. Among the questionnaires, perceived competence goes from 2.52 ± 0.95 to 3.36 ± 0.73 (p = 0.049) and trait sport confidence from 4.94 ± 1.17 to 6.60 ± 0.65 (p = 0.049). Among the cardiac indices, the low-to-high frequency ratio falls during recall of the critical moment, from 3.4 ± 2.3 to 1.014 ± 0.71 (p = 0.038), and during imagery of successful performance, from 1.933 ± 0.917 to 0.988 ± 0.572 (p = 0.046).

Everything else stays above threshold, including things that move a good deal: the standard deviation of inter-beat intervals during relaxation goes from 69.14 to 89.78 thousandths of a second at p = 0.068, low-frequency power in the same state from 1,673 to 4,111 at p = 0.081, high-frequency power during the critical moment from 505.6 to 871.4 at p = 0.168. The authors report Cohen's d values between 0.91 and 0.94 for these, that is large effects which with five people do not reach significance.

In the relaxation condition, by contrast, the frequency ratio rises rather than falls, from 2.298 to 2.95. The ordering between conditions inverts: before the intervention, recall of the critical moment gave a higher ratio than rest (3.4 against 2.298), after it a lower one (1.014 against 2.95). No shooting scores were collected.

The five participants signed informed consent for themselves and for their legal representatives, with a confidentiality agreement and in accordance with the Declaration of Helsinki.

The programme was built after a literature review and consultations with athletes and coaches, by a group of three experts: a professor of sport psychology, a sport psychology counsellor and a shooting coach. The eight sessions cover, in order, orientation, goal-setting, self-understanding, imagery across two sessions, relaxation through breathing, cognitive restructuring and routine. Each group session was followed within 48 hours by a 30-minute individual session, with a worksheet to complete and feedback on the previous week's worksheet.

The electrocardiogram was collected with a PolyG-I, electrodes on the palm side of both wrists and the right ankle, in a sound-attenuated room with controlled temperature and humidity, at 512 Hz, with a 60 Hz filter to remove mains interference, and analysed with Kubios HRV 2.2. The three conditions lasted six minutes each, 18 minutes in all, with a one-minute interval between them to remove the preceding condition's effect and the order assigned at random; on request participants could rest before moving to the next. The indices considered are, in the time domain, the standard deviation of the intervals and the root mean square of successive differences; in the frequency domain, low-frequency power between 0.04 and 0.15 Hz, high-frequency power between 0.15 and 0.4 Hz and their ratio, with the powers expressed in milliseconds squared.

The questionnaires are four, in their Korean versions: the psychological skills inventory for archery and shooting, 25 items on a seven-point scale with five sub-factors (imagery control, attention and anxiety, achievement motivation, arousal regulation, confidence); the intrinsic motivation inventory, 16 items on seven points with four subscales (interest and enjoyment, perceived competence, effort and importance, pressure and tension); the sports anxiety scale, 21 items on four points with three subscales (somatic anxiety, worry, concentration disruption); and the trait sports confidence inventory, 13 items on nine points. The reliability coefficients reported come from previous studies, not from this sample.

The analysis is by paired t-tests, separately for each condition and each index, in SPSS 21, with the threshold at 0.05, two-tailed tests and Cohen's d as the effect size. The individual counselling sessions were transcribed and used as qualitative material to interpret the results, without any method of analysis being described.

Five participants and no control group. A paired t-test on five people has four degrees of freedom, and the four significant results all sit between p = 0.038 and p = 0.049, that is barely inside the threshold: moving one participant takes them out. Over the eight weeks these 17 and 18 year olds went on training, grew, and repeated the same procedure a second time, including imagery on command with electrodes attached. The design does not separate the programme's effect from any of this.

There is an irony worth noting, because the authors write it themselves. In their introduction they cite the meta-analysis documenting publication bias in psychological skills training studies and the fact that overly positive effects can come from participants' expectations, and offer this as a reason to seek objective indicators. They then use a single-group design, that is the one in which expectation operates unchecked, and as the objective indicator a cardiac index collected in a condition that requires a voluntary effort of imagination.

The comparisons number about 28 across cardiac indices and questionnaire subscales, with no correction. Four significant is more than chance would give, but with those margins and that sample size the picture stays fragile. The reverse holds too: three values have a Cohen's d between 0.91 and 0.94, that is large effects, and fail to reach the threshold because the sample is too small. Reading only the four that made it, and not the three equally large ones that did not, is a selection.

The cardiac result is presented in a way that is hard to follow. The low-to-high frequency ratio falls during the two imagery conditions, and the abstract describes this as a «strengthened responsiveness to stress» of the autonomic nervous system; usually a lower ratio during a stressful stimulus reads the opposite way, that is as less sympathetic predominance. The harder figure to reconcile is another: in the relaxation condition that ratio rises, and after the intervention recall of the critical moment gives a lower ratio (1.014) than simply breathing quietly (2.95). An ordering like that, with the stressful stimulus below rest, is not easily explained as a better stress response. Add that the frequency ratio is a contested index as a measure of sympatho-vagal balance, and that here it has standard deviations wide relative to the means, 3.4 ± 2.3 before the intervention.

There is no measure of performance. No scores, no competition: the conclusion that the programme helps athletes cope better with difficult situations in competition was tested neither in competition nor on scores. What was measured is how the heart and the questionnaires respond to imagery performed sitting in a room.

Two things remain. The questionnaires' reliability coefficients come from previous studies, as is inevitable with five people, and one of the sub-factors of the inventory specific to archery and shooting has low reliability in those studies, 0.63 for imagery control; one of the reported values is printed as «84» instead of 0.84. The counselling sessions are used as interpretive material without any method of qualitative analysis being described. And, above all for this register: the sport is air pistol shooting, not archery. The action, the length of the shot cycle and the physical load are different, so any use of this study for archers is a transfer to be declared.

What this study offers an archer is not a result but a programme, described in enough detail to be reconstructed. Eight weeks, one 50-minute group session a week plus half an hour of individual counselling within two days, with one worksheet a week and feedback on the previous one. The topics, in order: why mental strategies matter, short and long-term goals distinguished between process, performance and outcome, knowledge of one's own characteristics, imagery across two sessions (first the mechanism, then the practice), breathing to regulate arousal, cognitive restructuring with self-talk built for specific situations, and routine. It is a reasonable and transferable structure, built with a professor, a counsellor and a coach.

The measurement side is the interesting idea. Rather than asking only how the athletes feel, the authors record the heart in three conditions, six minutes each: breathing quietly, recalling a critical moment experienced in competition, and imagining a perfect performance. Asking an archer to recall the end that went badly and watching what the heart does is a concrete way of seeing whether mental work has changed anything, and it needs no laboratory.

The results, by contrast, cannot be taken away. Five shooters, no control group, four significant results all between p = 0.038 and p = 0.049, and three equally large effects that fail to reach the threshold. Nothing was measured on the score. And the cardiac result is hard to interpret: after the eight weeks the frequency ratio was lower while recalling a critical moment than while breathing quietly, which does not resemble a better response to stress.

One last point about how to read this literature: the authors themselves recall, at the outset, that psychological skills training studies suffer from publication bias and that effects which look too good can come from participants' expectations. That is a good yardstick to keep on this study too, and on the many similar ones found with a single group and few participants.