Anxiety, Stress, & Coping

An International Journal

ISSN: 1061-5806 (Print) 1477-2205 (Online) Journal homepage: www.tandfonline.com/journals/gasc20

To err again is human: exploring a bidirectional

relationship between pressure and performance failure
feedback

David J. Harris, Samuel J. Vine, Michael W. Eysenck & Mark R. Wilson

To cite this article: David J. Harris, Samuel J. Vine, Michael W. Eysenck & Mark R. Wilson
(2019) To err again is human: exploring a bidirectional relationship between pressure
and performance failure feedback, Anxiety, Stress, & Coping, 32:6, 670-678, DOI:
10.1080/10615806.2019.1643459

To link to this article: https://doi.org/10.1080/10615806.2019.1643459

Published online: 18 Jul 2019.

Submit your article to this journal

Article views: 1153

View related articles

View Crossmark data

Citing articles: 6 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=gasc20
ANXIETY, STRESS, & COPING
2019, VOL. 32, NO. 6, 670–678
https://doi.org/10.1080/10615806.2019.1643459

To err again is human: exploring a bidirectional relationship

between pressure and performance failure feedback
a
David J. Harris , Samuel J. Vinea, Michael W. Eysenckb and Mark R. Wilsona
a
College of Life and Environmental Sciences, University of Exeter, Exeter, UK; bDepartment of Psychology, Royal
Holloway University of London, London, UK

ABSTRACT ARTICLE HISTORY

Background and Objectives: While the potentially negative effects of Received 11 February 2019
pressure on skilled performance have been well studied in laboratory- Revised 22 May 2019
based research, theoretically driven questions based on real-world Accepted 28 May 2019
performance data are lacking.
KEYWORDS
Design: We aimed to test the predictions of the newly developed Anxiety; sport; choking;
Attentional Control Theory: Sport (ACTS), using archived play-by-play errors; clutch; dependence
data from the past seven seasons of the National Football League
(American Football).
Methods: An additive scoring system was developed to characterize the
degree of pressure on 212,356 individual offensive plays and a Bayesian
regression model was used to test the relationship between
performance, pressure and preceding negative outcomes, as outlined in
ACTS.
Results: There was found to be a clear increase in the incidence of failures
on high pressure plays (odds ratio = 1.20), and on plays immediately
following a previous play failure (odds ratio = 1.09). Additionally, a
combined interactive effect of previous failure and pressure indicated
that the feedback effect of negative outcomes was greater when
pressure was already high (odds ratio = 1.10), in line with the predictions
of ACTS.
Conclusions: These findings reveal the importance of exploring
momentary changes in pressure in real-world sport settings, and the
role of failure feedback in influencing the pressure-performance
relationship.

Introduction
Sport provides an almost perfect environment for examining performance under pressure. Skills that
have been honed and perfected during practice can break down just when the need to execute them
is greatest. In studying this paradoxical effect, Baumeister (1984) defined pressure as “any factor or
combination of factors that increases the importance of performing well” (pp. 610). The proposed
mechanism by which pressure exerts its effect on skilled performance is via increased anxiety, an
emotional response to threat, comprising cognitive worry and physiological arousal (Eysenck,
1992). While individual differences in response to pressure do exist (e.g., “clutch” performance; see
Otten, 2009), a large literature base has revealed that anxiety can have deleterious effects on sporting
performance by disrupting attention (see Payne, Wilson, & Vine, 2018 for a recent systematic review).
There is strong support for the role of attentional disruptions in leading to both increased self-moni-
toring and control (Baumeister, 1984; Beilock & Carr, 2001; Masters & Maxwell, 2008) and/or increased

CONTACT Mark R. Wilson [email protected]

© 2019 Informa UK Limited, trading as Taylor & Francis Group
 ANXIETY, STRESS, & COPING 671

distractibility (Eysenck, Derakshan, Santos, & Calvo, 2007; Wilson, 2008), but what is less well under-
stood is how and why competitive pressure leads to anxiety in the first instance. A new theoretical
development, Attentional Control Theory: Sport (ACTS; Eysenck & Wilson, 2016) seeks to address
just this question.
ACTS was developed to extend the predictions of Attention Control Theory (ACT; Eysenck et al.,
2007), to the effects of pressure on the relatively automated skills of sport performers. ACT suggests
that anxiety leads to an imbalance between goal-directed and stimulus-driven attentional systems,
creating increased attention to threat related cues and processing inefficiency. As a result, perform-
ance may suffer when compensatory strategies (e.g., increased effort) are unsuccessful (see Eysenck &
Wilson, 2016, and Wilson, 2012 for reviews in sporting tasks). While the relationship between anxiety,
attention and performance remains as previously outlined in ACT, it is the antecedents of anxiety that
receive more attention in ACTS. Specifically, ACTS suggests that a bidirectional relationship exists
between pressure and performance, based on feedback loops relating current, to desired perform-
ance (Eysenck & Wilson, 2016). It is the outcome of these feedback loops that influences perceptions
of threat, which in turn leads to the experience of anxiety (see Figure 1).
This effect is operationalized in ACTS in terms of Berenbaum’s two-phase model of worry, which
suggests that the initiation of anxiety (specifically its cognitive component worry) is influenced by the
perceived costs and perceived probability of future undesirable outcomes (Berenbaum, 2010; Beren-
baum, Thompson, & Pomerantz, 2007). First, undesirable outcomes (e.g., losing, or individual
examples of skill failure) are prominent in sporting contexts and the costs of these are greater in
high-pressure situations than low-pressure ones, because more is at stake (Baumeister, 1984).
However, the experience of pressure is rarely constant, and will depend on momentary reflections
on exactly what is at stake. Second, it is likely that the perceived probability of losing increases as
a function of the number of failure experiences during a match or competition and decreases as a
function of the number of success experiences. Indeed, a qualitative study of young, international
golfers by Nicholls, Holt, Polman, and James (2005) identified that three quarters of all stressors
could be grouped under themes related to either their own mental and physical errors or good per-
formance from opponents. Such negative performance feedback will increase the perceived prob-
ability of subsequent errors if an individual believes that performance exhibits dependence.

Figure 1. Schematic representation of the bi-directional pressure-performance relationship, as outlined in Attentional Control
Theory (Eysenck et al., 2007; dashed lines) and Attentional Control Theory: Sport (Eysenck & Wilson, 2016; solid lines). Of particular
importance is the fact that situational pressure does not necessarily result in increased anxiety, but that this is influenced by an
individual’s perception of the associated costs of failure (primarily influenced by the interpretation of momentary situational
pressure) and probability of failure (primarily influenced by the interpretation of preceding errors / negative performance feed-
back). An additional feedback loop between prior failure and attention reflects the direct influence that error monitoring has
on attention.
672 D. J. HARRIS ET AL.

Dependence reflects the belief that the probability of success on one play is influenced by pre-
vious plays and is most frequently associated with research examining performance streaks; colloqui-
ally referred to as the hot hand effect (e.g., Bar-Eli, Avugos, & Raab, 2006; Wetzels et al., 2016).
However, interpreting negative performance feedback (e.g., an error) as evidence that more mistakes
are likely, would also reflect dependence (e.g., Link & Wenninger, 2019). To summarize, ACTS predicts
that when both the perceived cost of failure (influenced by fluctuations in the current level of
pressure) and perceived probability of failure (influenced by previous unsuccessful performance feed-
back) are high, the interactive effect will lead to heightened anxiety, impaired attentional control and
negative consequences for performance (as summarized in Figure 1).
The current study sought to provide the first test of the basic performance effects proposed in
ACTS; by examining the potential interacting effects of preceding failure and situational pressure
on subsequent performance. The study sought to develop new knowledge in two ways. First,
while it is widely acknowledged that pressure can disrupt performance in many perceptual-cognitive
tasks (Payne et al., 2018), there is limited empirical evidence from real-world environments, where
pressure will fluctuate from moment to moment (e.g., Deutscher et al., 2018). Second, and as outlined
explicitly in ACTS, the pressure-performance relationship is likely to be more complex than the uni-
directional effect addressed by the blocked (low pressure vs high pressure) laboratory experimental
manipulations adopted in the vast majority of research examining sporting performance under
pressure (Eysenck & Wilson, 2016).
To explore the interacting effects of performance dependence and within game fluctuations in
pressure in a real-world environment (American Football), we examined every individual play from
all games in the National Football League from 2009 to 2016. As it was not possible to test the med-
iating interpretive processes leading to anxiety (Berenbaum, 2010), we restricted our focus to the pro-
posed relationship between the two input variables (momentary pressure and failure feedback) and
the output variable (current performance; see Figure 1). Based on the predictions of ACTS, it was
hypothesized that there would be: (i) an increased probability of play failure on high pressure
plays; (ii) an increased probability of one play failure following another (i.e., dependence); and (iii)
an additional interactive effect, such that the negative effect of negative performance feedback
would be exacerbated when pressure is already high (Eysenck & Wilson, 2016).

Methods
Every play from 2009 to 2016 in the National Football League was obtained from www.NFL.com using
the R package “nflscrapR” (Horowitz & Yurko, 2016). This data set provided 362,448 individual plays
outlining each play outcome, and game information such as field position (yards from the opposing
team’s in-goal area), game time remaining, and current score. The discrete nature of American Foot-
ball plays allows each instance to have a relatively clear positive/negative outcome, while retaining
relevance to previous plays (across the four “downs” – the available attempts to move the ball
forward 10 yards before possession is turned over). In order to assess dependence of performance
failure, only passing and running plays were analyzed, as kicking plays end a possession.

Performance failure
The analysis focused on play outcomes in relation to the team in possession, such that losing the ball
or failing to make ground were negative outcomes or examples of performance failure. Specifically,
these outcomes were operationally defined based on agreement between six University level Amer-
ican Football coaches, as plays resulting in: an incomplete pass (including interceptions); a sack (quar-
terback tackle behind the line of scrimmage); a fumble (player in possession loses control of the ball);
or making negative yards (receiver tackled behind line of scrimmage – the imaginary line separating
the teams before each play). Finally, plays immediately preceded by a failed play in the same drive
were then coded as “post-failure” plays.
 ANXIETY, STRESS, & COPING 673

Pressure
The occurrence of pressure was inferred based on match conditions that increased the importance of
performing well (Baumeister, 1984) and the cost of failure (Berenbaum et al., 2007). The scoring
system for the factors that increase pressure was developed based on: (1) previous literature exam-
ining performance pressure; (2) discussions with the same six American Football coaches; and (3)
agreement between three of the authors. Pressure was predicted to be greater when: the game
was close (e.g., Deutscher et al., 2018; Toma, 2017); there was less time remaining (Cao, Price, &
Stone, 2011; Solomonov, Avugos, & Bar-Eli, 2015; Toma, 2017); an error would confer a greater
cost (Berenbaum et al., 2007; Hickman & Metz, 2015); or the expectation of a score was higher (Solo-
monov et al., 2015). Therefore, a pressure score was assigned in a cumulative manner, based on
whether: (i) the play was 3rd or 4th down; (ii) the game score was close (within 8 points, i.e., a touch-
down and 2-point conversion); (iii) it was the final quarter; (iv) the team in possession was behind; (v)
the play began in the “red zone” (i.e., within 20 yards of the in-goal area). This resulted in a 6-point
pressure score ranging from 0 (low pressure) to 5 (high pressure).

Data analysis
Data analysis was conducted in RStudio 1.0.143 (R Core Team, 2017). Twenty-six plays with missing
data were removed, and only plays where dependence could be assessed were included (i.e., not the
first play in a drive), resulting in 212,356 plays for analysis. A logistic regression model was used to
examine the effect of scored pressure and prior failure on subsequent performance (binary
outcome), using a Bayesian estimation approach. Bayesian estimation attempts to identify the cred-
ible interval of a parameter (Kruschke, 2014) and is particularly appropriate for large data sets, where
the impact of sample size on p-values makes an examination of significance levels relatively uninfor-
mative (Royall, 1986). A Bayesian approach was chosen because it also provides a more intuitive
approach to estimating parameters and avoids binary decision criteria (see Kruschke, 2010, for
discussion).
Data were modeled as deriving from a Bernoulli distribution, with a logistic link function;
g = Bernoulli(m) with m = logistic(b0 + b1 + b2 + b3 ). Priors on b were set as a conventional non-
informative normal distribution (Kruschke, 2014). Markov Chain Monte Carlo simulations were run
using JAGS (Plummer, 2003) based on 50,000 steps. Chain diagnostics indicated good convergence,
and effective sample sizes exceeded 17,000. The reliability of observed effects was interpreted based
on the credible intervals of the regression coefficients, provided by the posterior distributions (i.e., do
the highest density intervals cross zero?). An odds ratio (OR) was also calculated as an unstandardized
effect size. All our data, analysis code for NFL plays, Bayesian modeling code and model checking stat-
istics are available from the Open Science Framework (osf.io/mjf5p/).

Results
The regression model indicated that increasing pressure score was a reliable predictor of perform-
ance failure (Table 1). The highest density interval (HDI) of the posterior distribution (Figure 2)

Table 1. Summary of estimated regression coefficients (and their 95% HDI) of predictors in the regression model.
Model
Predictor β HDI low HDI high OR
Constant −1.16 −1.19 −1.14 0.31
Pressure 0.18 0.17 0.19 1.20
Post failure 0.09 0.04 0.13 1.09
Interaction 0.09 0.07 0.12 1.10
Abbreviations: HDI: Highest density interval; OR: odds ratio.
674 D. J. HARRIS ET AL.

Figure 2. Posterior distributions of regression coefficients, with 95% highest density intervals (HDIs), based on 50,000 steps. These
distribution plots indicate that the credible values of the regression coefficients (i.e., the HDIs, resulting from the Markov Chain
Monte Carlo simulations) do not include zero.

represents the credible interval of a parameter, and indicates the presence of a reliable effect when
the credible values do not cross zero. The estimated pressure effect was modeled within a narrow
interval that did not cross zero (β = 0.18, 95%HDI [0.17, 0.19]), signifying a reliable effect. The com-
puted OR indicates that a one unit increase in the pressure score, entering the final quarter for
example, made an offensive play failure 1.2 times more likely. Prior negative performance feedback
also showed a non-zero effect (β = 0.09, 95%HDI [0.04, 0.13]), with the OR indicating that a failure on
the preceding play increased the chance of a further failure by 1.09 times. Additionally, an interaction
effect (β = 0.09, 95%HDI [0.07, 0.12]) explained further variance in play success, such that the effect of
a one unit increase in the pressure score was 1.1 times greater when the play was also preceded by a
play failure.1
To further illustrate the interactive effect of pressure score and negative performance feedback,
Figure 3 shows the mean rate of failure (with Bayesian credible interval) across the six levels of
pressure score, for plays following either a failure or a successful play. The interactive effect is
evident in the increasing difference between post-failure and post-success plays across increasing
pressure scores.

Figure 3. Mean play failure rate (with Bayesian 95% credible intervals) on plays immediately following failed or successful plays,
across increasing pressure scores.
 ANXIETY, STRESS, & COPING 675

To assess the effect of our scoring assumptions for pressure on these outcomes, a robustness
analysis (see Willink, 2008) was run by varying the scoring parameters. Assumptions such as closeness
of the game (4, 6, 10 or 12 points), distance from the end zone (10, 15, 25 or 30 yards) and number of
downs (3rd or 4th) were varied, and other predictors (“last quarter” and “in the red zone”) were
removed from the model. Results showed that slight variations to the assumptions had little
impact on the results and no effect on conclusions drawn (see supplementary materials for details
of these analyses: osf.io/mjf5p/).

Discussion
Despite the pressure-performance relationship being one of the most studied areas in sport psychol-
ogy, there is limited research that manages to both extend theoretical development and move
beyond the artificial confines of the laboratory. This study explored the relationship between
momentary pressure and performance dependence in an extensive, detailed, play-by-play data set
from elite sport. This is also the first study to explicitly test the main tenets of ACTS (Eysenck &
Wilson, 2016), a recent, sport-specific development of one of the most well-researched theories for
explaining how anxiety influences performance (ACT; Eysenck et al., 2007). Our three hypotheses –
based on the pressure-performance feedback effect outlined in ACTS – were supported, providing
key implications for both future research and practice.
First, an unsuccessful offensive play was more likely when the game situation dictated increased
pressure (e.g., the game was close, it was the final quarter, and the end zone was near). This finding
supports and extends earlier work in basketball by Cao et al. (2011), who found that, compared to
career averages, NBA free-throw shooting accuracy was significantly impaired during the final
seconds of close games. As such, the pressure-performance effect found here provides further com-
pelling evidence that detrimental effects can be observed in real-world elite sport, and not just in the
laboratory. While there were of course instances of successful offensive plays under pressure in this
data set (i.e., clutch performance; Otten, 2009), the strength of this analysis of more than 200,000
plays was that the average effect was one of performance impairment (i.e., choking; Baumeister,
1984; Beilock & Carr, 2001).
In line with our second hypothesis, the likelihood of an unsuccessful play was increased following
an unsuccessful play on the previous play of the drive. This supports the prediction of ACTS that per-
formance exhibits dependence and that errors can have detrimental feedback effects. Generally, the
support for dependence when examining “hot” performance streaks is mixed (Bar-Eli et al., 2006),
however, the current novel question suggests that negative dependence may have an important
influence on performance in pressurized environments (see also Gray & Allsop, 2013). Negative feed-
back (e.g., perceived errors) may provide a stronger input to subsequent performance expectancies
than positive feedback (e.g., hot streaks). This interpretation is supported by recent data in volleyball
decision making (Link & Wenninger, 2019) and the work of Baumeister and colleagues, who intimated
that “bad is stronger than good”, as a general principle across a broad range of psychological
phenomena (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001). Specifically, these authors noted
that bad events have longer lasting and more intense consequences than good events and that
the effects of good events dissipate more rapidly than those of bad events.
Of particular importance to the predictions of ACTS was support for our third hypothesis, that the
combination of an increased pressure score and a previous error would show an interactive effect.
The present finding indicates that, not only does prior failure increase the chance of further
failure, but this effect is larger under increasing levels of situational pressure. Indeed, Figure 3
reveals that at the highest levels of pressure (i.e., a pressure score of 4 or 5) there is a 50% probability
that one failure will be followed by another, compared to only a 27% probability at low levels of
pressure (i.e., a pressure score of 1). Importantly, ACTS provides an explanation as to why this stark
difference in performance might occur, based on Berenbaum’s (2010) initiation model of worry.
ACTS suggests that increased pressure will increase the perceived costs of failure and that negative
676 D. J. HARRIS ET AL.

performance feedback will increase the perceived probability of failure. These in turn will result in
increased cognitive anxiety, leading to disruptions to attention and subsequent performance as out-
lined in Figure 1.
It is important to note that in the current study, specific mediating pathways (e.g., anxiety, worry)
could not be directly tested, as we were unable to directly assess how players interpreted the failed
offensive plays or pressure. However, the pressure scoring system was based on factors that likely
increased the importance of performing (Baumeister, 1984) and was shown to be robust to modifi-
cations in its assumptions. We can also be confident that sportspeople do interpret negative perform-
ance feedback as a key stressor (e.g., Nicholls et al., 2005), although the fact that different players
within the team may respond to the same situation differently, provides additional complexity in
interpreting performance data from a team sport. Taken together, while important individual differ-
ences in interpretation were not measurable, and extraneous factors could not be controlled for (such
as the defence trying harder on some plays), an overall relationship was still found for this large data
set, which indicates that the predicted interacting effect of pressure and errors on performance held
true despite these individual variations.
Further research is therefore needed to assess the potential modulating effect of the underlying
psychological factors in domains where skills are performed under pressure (e.g., sport, military,
surgery, aviation). This research will need to explore novel experimental approaches so that the
online (or at least temporally proximal) measurement of felt pressure during performance can be con-
sidered in relation to ongoing performance expectancies – currently a limitation in most experimen-
tal work exploring the impact of state anxiety on performance (Eysenck & Wilson, 2016). Two
noteworthy examples of experimental studies that could guide future research, were carried out
by Gray and Allsop (2013) and Walters-Symons, Wilson, and Vine (2017). Gray and Allsop (2013)
found that pressurized performance in a baseball batting task was influenced by previous perform-
ance, and could be mediated by changes in attentional focus (as measured by secondary task per-
formance). Walters-Symons et al. (2017) measured objective attentional changes via eye tracking
technology in a golf putting task and examined how these measures changed following misses com-
pared to successful attempts. Participants were able to successfully recover from errors (i.e., missed
putts) through a refocusing of visual attention, but additional errors were made when attention
remained poor. Similar approaches examining moment-to-moment changes in objective perform-
ance markers may be required to understand how fluctuations in pressure and negative appraisals
may compound errors.
Despite the need for corroborating experimental data, there are a number of implications arising
from the findings of this novel study, and the predictions of ACTS in general. First, there may be
additional benefits related to the term “expertise-induced amnesia”, which is used to describe the
automatic and non-conscious nature of skilled performance (Beilock & Carr, 2001). Performers who
can forget their mistakes (or good plays from opponents) – especially when pressure is heightened
– are less likely to feel anxious and experience the disruption of attentional control associated with
choking. It may be that this is a key characteristic of performers who are described as clutch under
pressure (Otten, 2009; Solomonov et al., 2015). Second, practitioners seeking to help performers deal
more effectively in pressure situations could use ACTS to guide intervening at two stages; first by
reducing the likelihood that environmental pressure leads to anxiety, or second, by limiting
anxiety-induced impairments to effective attention control.
In the first instance, Berenbaum’s two phase model provides a useful structure: anxiety can be
limited if performers can reduce the perceived costs of failure, and do not associate mistakes with
an increased probability of further mistakes. Both of these strategies would fit within a framework
that either sought to maintain a rational interpretation of the competitive environment (e.g., Rational
Emotive Behavioral Therapy; Wood, Barker, Turner, & Sheffield, 2018), or one whereby mistakes are
accepted in a non-judgmental way (e.g., a Mindfulness-Acceptance-Commitment approach; Moore,
2009). Additionally, according to the model, it is possible intervene at a later stage by limiting the
impact of anxiety on attentional control, either by training individuals to maintain their focus on
 ANXIETY, STRESS, & COPING 677

key sources of information while they perform (e.g., quiet eye training; Vine, Moore, & Wilson, 2014),
or by training general functions of working memory implicated in attentional control (e.g., Ducrocq,
Wilson, Smith, & Derakshan, 2018; Ducrocq, Wilson, Vine, & Derakshan, 2016).
To conclude, situational pressure, performance failure and their interaction were all shown to be
reliable predictors of further performance failures, highlighting the importance of fluctuations in
pressure over time and the role of dependencies in performance. The current study is the first to
test the predictions of ACTS (Eysenck & Wilson, 2016) and reveal why it is important to adopt a
more fine-grained approach to studying the fluctuating nature of perceived pressure in real-world
settings, where the consequences of failure are meaningful. The combined effect of situational
pressure and the interpretation of failure (especially physical or mental errors) may have severe con-
sequences for subsequent performance, and future work should explore why such effects occur and
how they can be limited.

Note
1. Note, a frequentist approach, using a logistic regression gave almost identical regression coefficients and indi-
cated all effects to be significant at p < .001.

Disclosure statement
No potential conflict of interest was reported by the authors.

ORCID
David J. Harris http://orcid.org/0000-0003-3880-3856

References
Bar-Eli, M., Avugos, S., & Raab, M. (2006). Twenty years of “hot hand” research: Review and critique. Psychology of Sport and
Exercise, 7(6), 525–553. doi:10.1016/j.psychsport.2006.03.001
Baumeister, R. F. (1984). Choking under pressure: Self-consciousness and paradoxical effects of incentives on skillful per-
formance. Journal of Personality and Social Psychology, 46, 610–620.
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than good. Review of General
Psychology, 5(4), 323–370. doi:10.1037/1089-2680.5.4.323
Beilock, S. L., & Carr, T. H. (2001). On the fragility of skilled performance: What governs choking under pressure. Journal of
Experimental Psychology: General, 130(4), 701–725.
Berenbaum, H. (2010). An initiation–termination two-phase model of worrying. Clinical Psychology Review, 30(8), 962–975.
doi:10.1016/j.cpr.2010.06.011
Berenbaum, H., Thompson, R. J., & Pomerantz, E. M. (2007). The relation between worrying and concerns: The importance
of perceived probability and cost. Behaviour Research and Therapy, 45(2), 301–311. doi:10.1016/j.brat.2006.03.009
Cao, Z., Price, J., & Stone, D. F. (2011). Performance under pressure in the NBA. Journal of Sports Economics, 12(3), 231–252.
doi:10.1016/j.jebo.2015.04.007
Deutscher, C., Ötting, M., Langrock, R., Gehrmann, S., Schneemann, S., & Scholten, H. (2018). Very highly skilled individuals
do not choke under pressure: Evidence from professional darts. arXiv preprint arXiv:1809.07659.
Ducrocq, E., Wilson, M., Smith, T. J., & Derakshan, N. (2018). Adaptive working memory training reduces the negative
impact of anxiety on competitive motor performance. Journal of Sport and Exercise Psychology, 1–11. doi:10.1123/
jsep.2017-0217
Ducrocq, E., Wilson, M., Vine, S., & Derakshan, N. (2016). Training attentional control improves cognitive and motor task
performance. Journal of Sport and Exercise Psychology, 38(5), 521–533. doi:10.1123/jsep.2016-0052
Eysenck, M. W. (1992). Anxiety: The cognitive perspective. Hove, UK: Lawrence Erlbaum Associates.
Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive performance: Attentional control
theory. Emotion, 7(2), 336–353. doi:10.1037/1528-3542.7.2.336
Eysenck, M. W., & Wilson, M. R. (2016). Sporting performance, pressure and cognition: Introducing attentional control
theory: Sport. In D. Groome, & M. Eysenck (Eds.), An introduction to applied cognitive psychology (2nd ed., pp. 329–
350). London: Routledge.
678 D. J. HARRIS ET AL.

Gray, R., & Allsop, J. (2013). Interactions between performance pressure, performance streaks, and attentional focus.
Journal of Sport and Exercise Psychology, 35(4), 368–386. doi:10.1123/jsep.35.4.368
Hickman, D. C., & Metz, N. E. (2015). The impact of pressure on performance: Evidence from the PGA TOUR. Journal of
Economic Behavior & Organization, 116, 319–330. doi:10.1016/j.jebo.2015.04.007
Horowitz, M., & Yurko, R. (2016). nflscrapR: Scrapes NFL play-by-play and boxscore data across full seasons. R package
version 1.4.0. Retrieved from github.com/maksimhorowitz/nflscrapR
Kruschke, J. K. (2010). What to believe: Bayesian methods for data analysis. Trends in Cognitive Sciences, 14(7), 293–300.
Kruschke, J. K. (2014). Doing Bayesian data analysis: A tutorial with R, JAGS, and stan. New York, NY: Academic Press.
Link, W., & Wenninger, S. (2019). Performance streaks in elite beach volleyball - does failure in one sideout affect attacking
in the next? Frontiers in Psychology. doi:10.3389/fpsyg.2019.00919
Masters, R., & Maxwell, J. (2008). The theory of reinvestment. International Review of Sport and Exercise Psychology, 1(2),
160–183. doi:10.1080/17509840802287218
Moore, Z. E. (2009). Theoretical and empirical developments of the Mindfulness-Acceptance-Commitment (MAC)
approach to performance enhancement. Journal of Clinical Sport Psychology, 3(4), 291–302. doi:10.1123/jcsp.3.4.291
Nicholls, A. R., Holt, N. L., Polman, R. C. J., & James, D. W. G. (2005). Stress and coping among international adolescent
golfers. Journal of Applied Sport Psychology, 17(4), 333–340. doi:10.1080/10413200500313644.
Otten, M. (2009). Choking vs. clutch performance: A study of sport performance under pressure. Journal of Sport and
Exercise Psychology, 31(5), 583–601. doi:10.1123/jsep.31.5.583
Payne, K. L., Wilson, M. R., & Vine, S. J. (2018). A systematic review of the anxiety-attention relationship in far-aiming skills.
International Review of Sport and Exercise Psychology, 1–31. doi:10.1080/1750984X.2018.1499796
Plummer, M. (2003). JAGS: A program for analysis of Bayesian graphical models using Gibbs sampling. Proceedings of the
3rd international workshop on distributed statistical computing 124, 125.
R Core Team. (2017). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical
Computing. Retrieved from https://www.R-project.org/
Royall, R. M. (1986). The effect of sample size on the meaning of significance tests. The American Statistician, 40(4), 313–
315. doi:10.2307/2684616
Solomonov, Y., Avugos, S., & Bar-Eli, M. (2015). Do clutch players win the game? Testing the validity of the clutch player’s
reputation in basketball. Psychology of Sport and Exercise, 16, 130–138. doi:10.1016/j.psychsport.2014.10.004
Toma, M. (2017). Missed shots at the free-throw line: Analyzing the determinants of choking under pressure. Journal of
Sports Economics, 18(6), 539–559. doi:10.1177/1527002515593779
Vine, S. J., Moore, L. J., & Wilson, M. R. (2014). Quiet eye training: The acquisition, refinement and resilient performance of
targeting skills. European Journal of Sport Science, 14(1), S235–S242. doi:10.1080/17461391.2012.683815
Walters-Symons, R., Wilson, M. R., & Vine, S. J. (2017). The Quiet Eye supports error recovery in golf putting. Psychology of
Sport & Exercise, 31, 21–27. doi:10.1016/j.psychsport.2017.03.012
Wetzels, R., Tutschkow, D., Dolan, C., van der Sluis, S., Dutilh, G., & Wagenmakers, E.-J. (2016). A Bayesian test for the hot
hand phenomenon. Journal of Mathematical Psychology, 72, 200–209. doi:10.1016/j.jmp.2015.12.003
Willink, R. (2008). What is robustness in data analysis? Metrologia, 45(4), 442–447. doi:10.1088/0026-1394/45/4/010
Wilson, M. (2008). From processing efficiency to attentional control: A mechanistic account of the anxiety–performance
relationship. International Review of Sport and Exercise Psychology, 1(2), 184–201.
Wilson, M. R. (2012). Anxiety: Attention, the brain, the body and performance. In S. Murphy (Ed.), The Oxford handbook of
sport and performance psychology (pp. 173–190). New York, NY: Oxford University Press.
Wood, A. G., Barker, J. B., Turner, M. J., & Sheffield, D. (2018). Examining the effects of rational emotive behavior therapy on
performance outcomes in elite paralympic athletes. Scandinavian Journal of Medicine & Science in Sports, 28(1), 329–
339. doi:10.1111/sms.12926