Competitive Anxiety and Self-Confidence in Archery: A Narrative Review of the Empirical Literature

Introduction

There is a moment in every archery shot when nothing external is happening. The archer is at full draw holding roughly twenty kilograms of limb tension steady. The arrow will leave only when the archer decides to let it. No clock runs down in any meaningful way. What fills that gap is entirely internal. It is the reason archery has been described as a mental sport by the people who compete in it. For the same reason, it is treated as useful research setting by the people who study the sport (Haywood, 2005).

The structural features that produce this are shared with a small family of sports. Archery is closed skill, meaning the environment does not change in ways the performer must react to. It is self-paced, meaning execution is initiated by the athlete rather than by an event (Rampp et al., 2022). Rampp and colleagues (2022) used archery for exactly this reason when they wanted a motor learning task in which the decisive movement is small and internally regulated. The reason why it was chosen was different, but it amounted to the same thing namely that scoring is continuous, granular, and objective. The archer has one dominant strategy which never varies (Bucciol & Castagnetti, 2020).

The earliest work went about it through the body. The state of cardiovascular and cortical activity during self-guided marksmanship was noted by Hatfield et al. (1987). They observed a distinctive attentional state before the archer takes the shot. During the process of an archer taking the aim, there was a greater level of alpha change in the left hemisphere. Salazar et al., (1990) extended the approach to 28 elite archers and reported the same pattern. Landers et al., (1991) converted the observational approach into an intervention by feeding corticol activity back to 24 pre-elite archers. The developmental baseline for adolescent archers for autonomic measures was set by Carrillo et al. (2011). Gonzalez et al. (2017) showcased that expert archers normally tend to hold a longer fixation before the act of release. The do so particularly when visual conditions are made challenging.

There is a parallel tradition of self-reporting that focuses on how the archer feels rather than on what the archer’s body is doing. The primary instrument for this is the Competitive State Anxiety Inventory-2 (Martens et al., 1990) and revision of the said inventory (Cox et al., 2003). The inventory comprises three subscales that are administered just before the competition. These are self-confidence, cognitive anxiety and somatic anxiety. This field has been shaped largely by this instrument, more so than by any theory. It is also why the two constructs are treated together here rather than separately. Confidence is not measured in archery because researchers set out to study confidence. It is measured because it arrives bundled with anxiety, operationalised as anxiety inverted rather than as the sport specific construct described by Vealey (1986).

These traditions have produced a body of work that is larger than it looks. Interventions have been tested using relaxation, most recently by Rizal et al. (2019) and by Verawati & Valianto (2020). Imagery protocols have been trialed by Indahwati & Ristanto (2016) and by Isar et al. (2022). Mindfulness has been trialed by Wu et al. (2021). Meditation has been trialed by Billy Yachsie et al. (2023). Biofeedback has been trialed by Paul et al. (2012). A yoga module has been built and validated specifically for archers by Baishya et al. (2025). Machine learning has been applied to psychological coping profiles in archers by Taha, Musa, Abdul Majeed, et al. (2018); Taha, Musa, Abdullah, et al. (2018). Qualitative work has examined the sport specific breakdown known as target panic (Prior & Coates, 2020). It has also examined the conditions under which archers choke (Diotaiuti et al., 2021). A Korean meta-analysis has aggregated seventeen psychological skill training trials (Kim et al., 2021).

What has been missing is a synthesis asking whether the combined findings support the confidence with which conclusions about archery psychology are drawn. Coaching practice treats anxiety management as foundational. Adjacent constructs have received closer attention. Attentional networks in elite precision sport athletes have been examined by Lu et al. (2021) and P. Li et al. (2021). Mood and burnout have been addressed by Xu et al. (2024)Xu et al. (2024). The review addresses three questions. What does the evidence show about the relationship between these constructs and archery performance. How sound are the instruments and designs that produced it. Where does the field claim more than it can support.

Method

The corpus includes case observation of 5 athletes (Miyamoto, 1994). It includes cross sectional surveys, quasi experiments with and without controls, structural equation models (Wang et al., 2022). It includes support vector machine classification (SVM) (Taha, Musa, Abdul Majeed, et al., 2018). It also includes Poisson modelling of a decade of archived competition results (Li & Zhao, 2024).

The review is reported in accordance with the Scale for the Assessment of Narrative Review Articles (Baethge et al., 2019). PubMed and Google Scholar were searched on 2 February 2026. PubMed was searched on the population term alone, using archer[tiab] OR archers[tiab] OR archery[tiab] OR recurve[tiab] OR compound bow[tiab]. Google Scholar was searched with four title restricted strings entered using the allintitle operator, so that a record is returned only where every term appears in the title. These were allintitle archery anxiety, allintitle archery confidence, allintitle archery self-confidence, and allintitle archer anxiety. Together they returned 65 records.

No instrument name was searched, since instruments are rarely named in titles. A broader string pairing archery with the quoted phrase competitive anxiety, without the title restriction, returned approximately 1,890 records and was not used. Further studies were identified from the reference lists of the articles retrieved. No date limit and no design filter were applied. Every study discussed was read in full text.

Studies were included where archers were the sole or primary sample and at least one of competitive anxiety or self-confidence was measured with a named instrument. Studies of rifle and pistol shooting were retained separately where they address a question the archery evidence cannot, namely the trials by Dana et al. (2022), Lee et al. (2023) and Zhao et al. (2024). Instrument development and validation sources were retained for the section on measurement. Only English language full texts were assessed. It is a substantial restriction in a field whose most productive programmes publish in Indonesian, Malay, Turkish, and Korean. One Persian language study of skilled archers was identified and could not be read. The 17 trials aggregated by Kim et al. (2021) are not individually available in English.

studies measured competitive anxiety or self-confidence. Other work is cited where it bears on the argument but is not counted among the 23. Studies of archers that measured no anxiety or confidence outcome, including the psychophysiological, gaze, attention and archived competition literature, are cited as background. Studies of shooting, fencing, golf and darts are named as such wherever they appear and are never offered as evidence about archers. Sources describing an instrument or a theory are cited as framing.

Included studies are reported in three categories, identified in the final column of Table 1. Competitive archery evidence comprises the 15 studies of archers competing at provincial, national or international level. Non-competitive archery evidence comprises the four studies of novices, university classes and club participants. Related and mixed sample evidence comprises the four studies in which archers were not the only participants. The conclusions rest primarily on the first category, and where a claim depends on the second or third this is stated where the claim is made.

Studies of archers measuring autonomic control, concentration, trait emotional intelligence or self-efficacy are treated as background, since these are constructs distinct from competitive anxiety and sport confidence.

Four questions guided that reading. Whether the sample was large enough for the estimate reported, since nine of the 23 analysed fewer than 30 participants. Whether performance was measured from recorded scores or not measured at all, since six studies report no performance outcome. Whether any condition was controlled or withheld, since only five studies compare conditions rather than naturally occurring groups and only one is randomised. And whether the instrument was named with its reliability reported in that sample. The Table 1 records the principal limitation of each study in these terms. The 23 met these criteria. In table.1, 19 of archers alone are listed first, followed by four in which archery data are identifiable but archers were not the only participants. The principal limitation column records the feature that most constrains what each study can support, and is intended to be read alongside the finding rather than as a score. Every value was read from the full text of the source.

Defining Features of the Selected Studies

Three features of this literature constrain the conclusions that can be drawn from it.

Small Samples

All 23 studies report a sample size. Nine analysed fewer than 30 participants. Eight analysed between 30 and 60. Three analysed between 61 and 100. 3 of the total exceeded 100. The smallest was the 5 kyudo archers followed by Miyamoto (1994). Among the core studies the largest was the 115 competitive archers of Diotaiuti et al. (2021). Every one of the 23 reports a single unambiguous total, which was verified against the full text in each case.

Figure 1. Number of studies by participant sample size, across all 23 studies. The vertical axis reports the number of studies.

Geographically Concentrated

of the 23 studies come from two countries. Malaysia contributes seven and Indonesia contributes three. Four that are all are from overlapping author groups centered on one researcher. Korea contributes one study to this corpus despite a national prominence in the sport. This has been described as unbroken since archery entered Korean schools in the 1960s (Lim, 2018). The substantial Korean language literature aggregated is not individually accessible in English (Kim et al., 2021). Japan likewise contributes 1, the earliest study in the corpus (Miyamoto, 1994). India contributes 3 studies to the corpus (Tomor & Shah, 2015;Aggarwala & Dhingra, 2017; Baishya et al., 2025).

What remains untested is the transferability of the findings that are generated in Southeast Asian university and sports school settings. This question has practical implications because the training hours, sports science support that is embedded and coaching density in such settings differ from decentralized pathways. In Table 2, the evidence type column identifies how each source is used. Core and boundary sources are the 23 studies on which this review draws directly and which appear individually in Table 1. Contextual sources are studies of archers that did not administer an anxiety or confidence measure. Related sport sources are studies of shooting. Synthesis sources are meta-analyses rather than primary studies. Framing sources establish the properties of an instrument or the theory behind a construct. Only core and boundary sources are counted among the 23.

Figure 2 Number of studies by country of the participant sample, across all 23 studies. The values are counts of studies, not counts of participants, and they sum to 23.

Measurement of the Constructs

Seven of the 23 studies used the Competitive State Anxiety Inventory-2 or its revision. Seventeen of the 23 measured anxiety with a named inventory, so the CSAI family accounts for seven of those seventeen. No other instrument appears more than once. Original instrument CSAI-2 was used on 30 Indian archers by Tomor & Shah (2015). Revised form of CSAI-2 has been used in the studies by Rizal et al. (2019), Wan Ahmad et al. (2023), Cahyo et al. (2025) and Baishya et al. (2025). CSAI-2 has been utilized even for shooters by Dana et al. (2022). Three studies have either made use of the Athletic Coping Skills Inventory-28(ACSI-28) (Smith et al., 1995) or a coping skills profile which is actually derived from it, namely Bebetsos (2015) and the two machine learning papers by Taha, Musa, Abdullah, et al. (2018); Taha, Musa, Abdul Majeed, et al. (2018).

The dominance of one instrument would matter less if that instrument were secure but it is not. The inventory was submitted to confirmatory factor analysis on 1213 competitors, testing 2 independent subsamples. Lane et al. (1999) then tested the combined set and found that the proposed three factor structure fitted poorly every time. They traced part of the problem to item wording. It was noted that the cognitive anxiety items ask whether the athlete is concerned rather than worried, which are not the same state. Meta-analysis of 29 studies was done using the instrument. It was found that in closed skill sports, the family archery belongs to only the self-confidence subscale which bore a relationship to performance that differed reliably from 0 (Craft et al., 2003). A further complication was added by showing that the intensity of an anxiety score and the athlete interpretation of that intensity as helpful or harmful behave as separate variables. It is a distinction almost no archery study has recorded (Kais and Raudsepp, 2005).

Figure 3. Instrument family across the 23 studies. Where a study administered more than one measure, the primary anxiety or confidence instrument is counted.

When the construct is measured through a single instrument, findings about the construct cannot easily be separated from properties of that instrument. This is the reasoning that was advanced when recommending that the inventory be revised or replaced by Lane et al. (1999). It was reported that lower-level archers show higher anxiety and lower confidence (Lim, 2018) and (Wan Ahmad et al., 2023). Some part of that pattern may reflect a questionnaire built on the assumption that confidence is anxiety inverted (Martens et al., 1990).

The only archery study in this corpus to use a different anxiety instrument is (Shaari et al., 2023). Before a national competition, 44 Malaysian archers who were in the age bracket of 18 to 24 were given the 15 item Sport Anxiety Scale-2 (Smith et al., 2006). Their performances were then recorded across 36 arrows at 70 meters. The dominant subscale came out to be worry. It had a mean of 12.62 and a standard deviation (SD) of 3.54. The second was concentration disruption. It had a mean of 10.47 and SD 2.93. The close third subscale was somatic anxiety with a mean of 10.33 and SD 2.91r = -.45, p = .002, N = 44. The source reports this coefficient as r(44). The article states an analysed sample of 44 archers in three places, so the value in parentheses appears to be the sample size rather than the degrees of freedom. For a Pearson correlation on 44 cases, it would conventionally be 42. The sample size is therefore reported explicitly here to avoid reproducing an ambiguous notation.

There are two other significant departures that are worth recording. Bebetsos (2015) used the Greek version of the ACSI-28 with 70 elite archers at a national championship, which asks about dispositional tendencies rather than about state in the minutes before shooting. It produced findings about experience that the state inventories structurally cannot. Lim (2018) abandoned questionnaires almost entirely, asking 18 Korean athletes to rate anxiety on a single 10 point scale while collecting salivary alpha amylase and cortisol. The crudeness of that measure is obvious.

Anxiety and Performance

Where an association has been tested and reported with a coefficient, it has been negative. Shaari et al. (2023) reported r = -0.45 across 44 national level archers. Lim (2018) found that in her high performing group, competition score correlated with anxiety measured 3 minutes before the start at r = -0.86 and that group outscored the low performers by 339.5 against 323.3 across 36 arrows at fifty metres. Trait anxiety related to success was found across 407 individual sport athletes, of whom 70 were archers analysed separately (Sanioglu et al., 2017).

Consistent direction across independent samples is worth something. It is not worth as much as it looks. These are correlations from small samples with wide intervals. They cannot establish causal direction. An archer shooting has an obvious reason to feel anxious that owes nothing to anxiety having caused the shooting. Evidence from outside the questionnaire tradition is compatible without settling the matter. Archived competition data were modelled and performance decrements under pressure were found (Bucciol & Castagnetti, 2020). A comparable pattern across a decade of tour results was reported by C. Li & Zhao (2024). But neither study measured anxiety in the athletes concerned.

The most stimulating result in this section came from the study with the crudest instrument. Anxiety 30 minutes before competition, 3 minutes before, and 30 minutes after was measured by Lim (2018). In her high performing group, the association with score peaked at the 3 minute point and had dissipated by the third measurement. In her low performing group, the pattern was different in a way no single measurement could have detected. Their scores correlated with anxiety recorded thirty minutes after the competition had ended, r = -0.72. The athletes who shot poorly were still anxious well after there was anything left to be anxious about. It was reported that in post competition interviews they expressed regret, self-blame, and difficulty accepting their results (Lim, 2018).

Every other archery study located for this review measured anxiety once. Rizal et al. (2019), Wan Ahmad et al. (2023), Shaari et al. (2023), and Cahyo et al. (2025) all administered their inventory in the period immediately preceding performance. The variable separating the two performance groups was not pre competition anxiety but the failure of anxiety to resolve afterwards, which is a different phenomenon requiring a different schedule (Lim, 2018).

Anxiety and physiological changes

A cluster of studies has paired self-reported anxiety with physiological measurement. Together they undermine the assumption that a somatic anxiety score describes what the body is doing. Competitive anxiety alongside heart rate variability and respiration was examined in archers and it was concluded that the correspondence is conditional on personality traits rather than direct (Dal et al., 2018). It was found that heart rate and somatic anxiety moved differently across preparatory practice, formal practice, and matches in 5 kyudo archers (Miyamoto, 1994). Autonomic control between novice and experienced archers was compared at the Sports Authority of India (Aggarwala & Dhingra, 2017). It was reported that trait emotional intelligence predicted the speed of autonomic recovery after a shooting session in 87 novice archers, a construct that appears nowhere else in this literature (Dal, 2019).

Read together these suggest the somatic subscale should be understood as a self-report of perceived bodily state rather than as a proxy for measured arousal. The two are related but they are not the same variable. This point was made explicitly by Dal et al. (2018). It was observed as implicit in the dissociation (Miyamoto, 1994). The point is reinforced from the other direction, finding across one hundred and eighty-four professional matches that neither heart rate nor vagally mediated heart rate variability distinguished high scoring from low scoring performances (Yuan & Chang, 2025).

Anxiety phenomenon in archery

The target panic, a performance block in which the archer loses volitional control of the release has been examined by Prior & Coates (2020). It was done through interpretative phenomenological analysis of 7 affected archers. Their account describes a sequence in which excessive focus on shot outcome drives an attempt to seize conscious control. This produces loss of control over execution which is followed by remedial strategies to recover automaticity. The related territory was approached quantitatively across 115 archers, testing anxiety, coping style, and decentering against self-reported choking frequency. It was found that the ruminative component of worry is the factor directly implicated (Diotaiuti et al., 2021). Neither phenomenon appears in the subscales of the state inventory. This means the instrument that dominates this literature cannot detect the two forms of anxiety most specific to the sport.

Cahyo et al. (2025) compared anxiety across the 40, 50 and 70 meter distances in 63 Indonesian archers and found no significant difference. F = 0.478, p = .622, although the 70 meter mean was marginally higher. But this negative finding is highly useful in this context. Coaches assume that the technical difficulty of the shot drives competitive anxiety. These data do not support that. There is one caution that is attached to it. Although the paper has named the three subscales of the revised inventory separately, at the end it reports a combined anxiety score based on all three subscales. If confidence was summed into that total, the resulting number is hard to interpret, since higher confidence would inflate an apparent anxiety score.

Self-Confidence and Performance

The survey of 61 athletes from Chinese national archery team spanning international, national, and first class grades modelled the relationships between cognitive reappraisal, sport-confidence, attention, and performance using partial least squares structural equation modelling (D. Wang et al., 2022). The important methodological choice was not the modelling. It was that they did not extract confidence from an anxiety scale. They used the State Sport-Confidence Inventory in its Chinese version together with the Athlete Sport-Confidence Inventory. Both are built to measure confidence as a sport construct in the tradition established by Vealey (1986). Performance came from an actual competition under Olympic rules, combining the qualification ranking score across 72 arrows with the match play set score.

Sport-confidence returned a total effect of 0.574 on performance, ahead of attention at 0.344, in a model explaining 33.6 percent of variance. Cognitive reappraisal showed a negative direct effect of -0.268. The total effect of cognitive reappraisal was statistically indistinguishable from zero in the survey conducted by D. Wang et al. (2022).

Within the studies reviewed here it is the largest reported effect on archery performance. It rests on the largest elite archery sample located and it was validated against competition scores rather than a laboratory task. It also concerns a construct that most of the field records as a subscale and then declines to discuss. This pattern is visible in Rizal et al. (2019) Wan Ahmad et al. (2023) and Cahyo et al. (2025) as each of these collected confidence data and reported it without analysis.

A far simpler route was used with Indonesian archers, using a purpose built 20 item confidence questionnaire and testing ability at 50 meters by Yachsie et al. (2021). Like Wang and colleagues, they measured confidence directly rather than by subtraction, and like them they found a relationship with performance. The study is methodologically modest and the instrument is not validated beyond that sample. It is nonetheless a second independent instance of the same pattern.

Antecedents of self-confidence

The five factor personality traits and narcissism were tested as predictors of anxiety and self-confidence in 52 college student athletes before an archery class final examination by Dal (2018). The design is unusual for this literature in treating the 2 constructs as outcomes rather than as predictors of performance. It is the only source located here that looks upstream of them. Its setting limits how far the findings travel since an examination is not a competition and the participants were students rather than competitive archers. But the question it asks is one the field has otherwise left alone.

The most useful evidence on its origins is offered by Bebetsos (2015). Across 70 Greek elite archers, differences emerged by competitive experience on freedom from worry, concentration, and confidence. More experienced archers were higher on all three. Age differences appeared on freedom from worry and on peaking under pressure. Older athletes reported better emotional control with less distress. Confidence in archery therefore looks at least partly accumulated rather than dispositional. That is encouraging for coaching. It also means cross sectional comparisons between experienced and inexperienced archers cannot separate confidence from everything else experience brings.

Interventions reported to improve performance

A PETTLEP imagery protocol was applied to archers preparing for a provincial competition in East Java and effects on competitive anxiety and concentration were reported Indahwati & Ristanto (2016). Progressive muscle relaxation was applied at a university club in Medan and improved concentration was reported Verawati & Valianto (2020). Imagery was combined with meditation and better accuracy was reported at 40 meters (Billy Yachsie et al., 2023). Imagery, self-talk, and both together across 18 sessions were compared and it was found that the combination outperformed either alone on state anxiety, confidence, and performance (Isar et al., 2022). Baishya et al. (2025) built a yoga module specifically for archers, validated its 32 practices against the ratings of 40 experts, and then delivered 6 supervised sessions over 2 weeks to 14 elite archers, reporting reduced competitive anxiety and improved self-confidence alongside gains in reaction time, accuracy, and balance. Effects from sensorimotor rhythm neurofeedback have been reported by Paul et al. (2012). A mindfulness programme was conducted with 23 archers for 4 weeks. It reported that there is an improvement in shooting performance as well as cognitive function (Wu et al., 2021). Significant gains in flow and mindfulness were reported when it delivered a mindfulness programme to not just archers but to a combined group of archers and golfers (Kaufman et al., 2009). Medium to high effects on mood, burnout and confidence were reported when it used a single case withdrawal design with a total of 16 archers of national team (Xu et al., 2024).

The single aggregated estimate has been provided here by Korean meta-analysis. 17 psychological skills training trials with archers were pooled by Kim et al. (2021). It reported a Hedges g of 0.469 which is a small effect that is moderated significantly by the level of athlete and duration of training. Adults and longer programmes did better.

Pointing opposite ways: The two controlled trials

archers from the novice category of archers were randomised by Rizal et al. (2019). They were gender matched. Progressive muscular relaxation or a control condition of light stretching was performed. The intervention ran nine sessions across three weeks. Anxiety was measured with the revised inventory. Heart rate was measured by pulse oximeter. Performance was measured as 20 arrows at 10 meters. The result was null. Time effects appeared for cognitive anxiety, p = .037, and performance, p = .019, but in both groups. Between group and interaction effects were absent, and nothing emerged for somatic anxiety, self-confidence or arousal.

A comparable design in shooting was run by Dana et al. (2022). 45 male shooters aged 19 to 30 were randomly assigned to mindfulness training, mental skills training, or control. They were divided into 15 shooters per group, across 6 sessions of 60 minutes over 6 weeks. Both interventions reduced competitive state anxiety significantly and neither outperformed the other.

Rizal et al. (2019) worked with novices introduced to archery a week before the pre test, while Dana et al. (2022) worked with experienced shooters. The Rizal protocol ran 3 weeks against 6. The Rizal performance task was 20 arrows at 10 meters. This distance was chosen because participants could not reliably hit the target further away, which is a considerable departure from competition.

Level was found to be a significant moderator (Kim et al., 2021). If interventions work better in trained athletes, then the only randomised controlled trial in archery was run in precisely the population where it was least likely to detect anything. That is bad luck rather than bad design. It means the null should not be read as evidence that relaxation does not help archers.

It also leaves something less comfortable. The uncontrolled positive findings come from trained archers and the controlled null comes from novices. So, design and population are confounded across the whole intervention literature. Broader syntheses in athletic populations support mindfulness efficacy under controlled conditions, as Y. Wang et al. (2023) established across randomised trials and as Noetel et al. (2019) argued from a wider review. A randomised trial in dart throwing found effects on beginner’s skill acquisition (Zhang et al., 2016).

Competitive level and Performance

The most frequently reported finding here, though it rests entirely on cross sectional comparison, is that higher level archers report less anxiety and more confidence than lower level archers. It was found between high and low performing groups at a national trial by Lim (2018). It was found among 32 recurve archers during enforced training isolation that lower level archers were around 5% higher on both anxiety subscales and top archers were substantially higher on confidence (Wan Ahmad et al., 2023). Intermediate and elite compound archers were compared on psychophysiological and shooting measures (Mat Salleh et al., 2023). In the shooting analogue, 45 adolescent shooters were compared by competitive grade and significant differences were found in both cognitive and physical state anxiety (Zhao et al., 2024).

The interpretive problem is the familiar one. Cross sectional designs cannot say whether low anxiety produces good performance, good performance produces low anxiety, or both follow from the years of training separating an elite archer from a developmental one. Bebetsos (2015) comes closest to disentangling it, since his experience groupings sit within a single elite sample where all participants compete nationally.

Sex and performance

No significant differences between men and women on any of seven factors across 70 Greek archers was found by Bebetsos (2015). Differences on all three subscales were found across 30 Indian archers. Women were higher on cognitive anxiety at 21.20 against 16.00, higher on somatic anxiety at 19.07 against 15.47, and markedly lower on confidence at 22.80 against 30.33 (Tomor & Shah, 2015). The question was approached through archived competition data across a decade and it was reported that female archers showed lower last arrow scores following adversity (Li & Zhao, 2024). It was examined through gut microbiota composition in response to anxiety in shooting and archery athletes (Wei et al., 2025).

Discussion

The field measures one construct well and the more important one badly

Competitive anxiety receives the instruments, the interventions, and the attention. Self-confidence is collected nearly universally and treated as a byproduct. In the one study measuring confidence with purpose-built instruments against real competition scores, sport-confidence carried a total effect of 0.574 while every other predictor trailed behind it. Attention returned a total effect of 0.344. Cognitive reappraisal, which the authors had expected to assist performance, showed a negative direct effect of -0.268 and a total effect that was not significantly different from zero (D. Wang et al., 2022). A compatible conclusion was reached by a different route, finding that in closed skill sports the confidence subscale was the only one of the three with a reliable relationship to performance (Craft et al., 2003). The inventory of Martens et al. (1990) arrived before most of the archery work did and bundled confidence in as a third scale. The research question appears to have followed the instrument rather than the other way round.

Instrument monoculture is the central weakness

Seven of the 23 studies used the Competitive State Anxiety Inventory-2 or its revision, which is seven of the seventeen that measured anxiety with a named inventory. No other instrument appears more than once. The factor structure of that inventory was rejected on over twelve hundred athletes (LANE et al., 1999) whose subscales relate inconsistently to performance (Craft et al., 2003). It does not record the intensity and direction distinction that Kais & Raudsepp, (2005) showed to matter. The single archery study using a different instrument produced the strongest correlation in the corpus (Shaari et al., 2023). One comparison cannot settle whether the instrument is the reason. It does identify the study that should be run.

A related problem concerns what the somatic subscale indexes. The physiological work indicates that self-reported bodily anxiety and measured autonomic activity relate conditionally rather than directly (Dal et al., 2018). It can dissociate across conditions (Miyamoto, 1994). Autonomic measures themselves failed to separate good from poor competitive performances across 184 matches (Yuan & Chang, 2025). A subscale named somatic anxiety invites the assumption that it indexes the body. The evidence suggests it indexes the athlete perception of the body, which is a different and arguably more interesting variable.

Four studies that would help

A randomised controlled trial of an anxiety intervention in competitive archers, at competition distance, with an active control and a dose of at least eight weeks, would resolve more than another 20 surveys. An instrument triangulation study giving one archer sample the revised state inventory, the Sport Anxiety Scale-2 and a dedicated confidence inventory together would answer directly whether the findings survive a change of measure. A repeated measures design sampling anxiety after competition as well as before, following the schedule Lim, (2018) used, a finding that has not been followed up. Any of these conducted outside Southeast Asia, particularly in a large archery nation with a decentralised development pathway, would test the transferability the field currently assumes.

Limitations

This is a narrative review and makes no claim to completeness. Screening and appraisal were carried out by one person without duplicate assessment. One substantial restriction in this field is that only full texts in English language were used. This restriction is not just a mere formality. One Persian language study related to skilled archers was identified but could not be read. Accessibility was largely limited to sources that are open access. This may bias the included set towards journals that publish openly.

The studies were weighed by one reviewer without a second reader, so the judgements about what limits each study are open to disagreement. Reporting in this literature is uneven. One included study states in its abstract that no significant differences were found while its results table reports three. A study excluded from the review reports a total sample that cannot be reconciled with its own breakdown by sex.

One source identified during screening was removed at the drafting stage. It was an extended conference abstract reporting on a large Malaysian sample. A further limitation is the papers that were read in full text. Another limitation concerns what could be synthesized. The evidence mixes a single case observation with survey, quasi experimental and modelling designs. The evidence base is small and unevenly distributed. 23 archery studies is not a large base from which to draw conclusions about an entire field. Ten of the 23 studies come from two countries. Nine analysed fewer than 30 participants and three exceeded 100.

Conclusion

This review set out to ask what the evidence shows about competitive anxiety, self-confidence and archery performance, how sound the instruments and designs that produced it are and where the field claims more than it can support. Where an association between anxiety and performance has been tested and reported with a coefficient it has been negative.

Three features of the 23 studies constrain every conclusion. Fewer than 30 participants and only three exceeded 100. And all but one were published from 2015 onward, so the field is younger in substance than its earliest study suggests. This is an instrument whose proposed three factor structure was tested by confirmatory factor analysis on 1213 competitors and did not hold. Its subscales were shown by meta-analysis of 29 studies to relate to performance inconsistently. It reported that the strongest correlation of the subscales of the dominant inventory cannot detect target panic or choking, the two forms of anxiety most specific to this sport. A somatic anxiety score is better understood as a self-report of perceived bodily state than as a proxy for measured arousal.

The largest effect on archery performance reported study measuring confidence directly rather than by subtraction found a relationship in the same direction. The meta-analytic finding that in closed skill sports the confidence subscale was the only one bearing a relationship to performance reliably different from zero points the same way. Two archery studies remain a thin base for a claim of that size, which is precisely the difficulty. The intervention literature carries a comparable weakness. Positive findings are reported with regularity from designs without controls. The one aggregated estimate is a small pooled effect moderated by athlete level and programme duration. The only randomised controlled trial in archery was null and was conducted in novices over three weeks at a distance well short of competition. Design and population are therefore confounded across the whole of that literature and the null cannot be read as evidence that the intervention does not work.

Taken together, the available evidence remains limited by methodological heterogeneity, small samples and inconsistent measurement approaches. Four study designs are outlined above that would address the principal gaps identified, and none of them is expensive or technically demanding. A randomised controlled trial in competitive archers at competition distance with an active control, a study giving one archer sample the revised state inventory, the Sport Anxiety Scale-2 and a dedicated confidence inventory together, a repeated measures design that continues sampling after competition has ended, and any of these conducted outside the two countries that supply the largest share of the present literature would answer more than a further set of surveys using the existing convention. Until that work is done, the findings summarised here should be read as what the available designs can support rather than as settled properties of the sport.

Declarations

Author Contributions- [Dr. Rina Poonia]¹ conceptualisation, methodology, investigation, data curation, formal analysis, writing of the original draft, and visualisation. [Mr. Priyanshu Poonia]² conceptualisation, methodology, supervision, validation, and writing through review and editing. [Dr. Ketul Kumar]³ investigation, screening of records, data curation, and writing through review and editing. [Mr. Mahindar Gora]⁴ investigation, verification of extracted data against source articles, and writing through review and editing. [Ms. Sareen Chhokar]⁵ formal analysis, appraisal of study quality, and writing through review and editing. [Dr. Anurag Sachan]⁶ validation, resources, and writing through review and editing. [Mr. Parth Chaudhary]⁷ supervision, project administration, and writing through review and editing. All authors have read and agreed to the submitted version of the manuscript.

Acknowledgements- The authors sincerely acknowledge Manipal University Jaipur for providing the necessary support, facilities, and academic environment to carry out this research.

Funding- This research received no external funding. The authors declare that no specific grant was received from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest- The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability- No new data were generated or analysed. All studies discussed are available from the cited sources.

Ethics- This is a review of previously published literature and involved no primary collection of human subject data.

Artificial Intelligence (AI) Statement- The authors declare that generative artificial intelligence (AI) tools were used solely to improve the English language, grammar, readability, and overall clarity of the manuscript. All scientific content, study design, data analysis, and conclusions were developed exclusively by the authors. The authors take full responsibility for the accuracy, originality, and integrity of the manuscript and have carefully reviewed and approved the final version prior to submission.