Achieving one’s best performance under high pressure is one of the most important challenges for athletes. In sport psychology, pressure has been defined as a factor or a combination of factors that increases the importance of performing well under certain situations (Baumeister, 1984). In the context of sports, many such factors exist, including competition, the presence of an audience, and a reward or punishment contingency (Baumeister & Showers, 1986). Therefore, in the field of sport psychology, attempts have been made to elucidate the underlying mechanisms of pressure-induced errors from motor control and emotional/cognitive perspectives (Baumeister, 1984; Eysenck & Calvo, 1992; Eysenck et al., 2007; Hanin, 2000). The understanding of these mechanisms has motivated research on interventions that can be used to acquire motor skills resistant to pressure, and their effectiveness has been supported (Beilock & Carr, 2001; Beilock et al., 2004; Ganesh et al., 2019; Gröpel & Mesagno, 2019; Masters, 1992). Therefore, clarification of the mechanisms of pressure-induced errors is important for maximizing athletic performance.
From the perspectives of motor control and learning, conscious processing (Baumeister, 1984) and the reinvestment hypothesis (Masters, 1992) have long been favored as the leading models to explain pressure-induced errors. They posit that poor performance under pressure is caused by regression to early stages of learning caused by overt conscious processing (explicit monitoring) in motor control. Specifically, pressure elevates self-consciousness and anxiety about performing correctly, which increases the amount of attention paid to the details of movements and their systematic control (Baumeister, 1984; Beilock et al., 2001; Lewis & Linder, 1997). This systematic level of attention to execution is thought to disrupt well-learned or automated performance (Lewis & Linder, 1997; Masters, 1992), as if it was a regression to an earlier stage of learning. This regression is evidenced by the high internal (novice-like) variability of movements observed under pressure (Gray, 2004; Gray & Allsop, 2013; Tanaka & Sekiya, 2011). Neuroscientific research has also proven that there exists a causal relationship between conscious processing and regression to an earlier learning phase (Ganesh et al., 2019; Rietschel et al., 2011; Zhu et al., 2011).
However, contrary to the above findings, some studies note that regression does not necessarily occur in pressure situations. For example, Higuchi (2000) investigated changes in throwing performance and kinematics under pressure after participants learned to throw a ball toward a specified target for 150 trials. If the participants had regressed to an early learning stage, then they should have exhibited novice-like throwing kinematics when under pressure. However, the results showed that, although throwing accuracy decreased, the kinematics differed from those in the early stages of learning, with conscious processing of the learned reference point (i.e., the release position). Although the performance kinematics were different, the manner of this difference was not akin to that demonstrated when learning the task in the early phase; rather, the change related to some other set of altered kinematics that were also ineffective. An important implication of this study is that regression does not always occur under pressure, as conscious processing would predict. Therefore, the mechanisms that explain the effects on motor control must be considered from perspectives other than the abovementioned theories.
The theory of ironic processes of mental control (Wegner, 1994, 2009; Wegner et al., 1987) has recently attracted attention as a possible framework for explaining the causes of errors (Gray et al., 2017; Tanaka & Karakida, 2019; Woodman et al., 2015). This theory was originally proposed following the discovery of an ironic cognitive phenomenon in which, when instructed “not to think” about particular content, participants would nevertheless think about that content (Wegner et al., 1987). Additionally, in motor tasks, instructions on what “not to do” have been shown to induce prohibited behavior (ironic errors; e.g., Liu et al., 2015; Wegner et al., 1998).
Woodman et al. (2015) tested the effect of pressure on ironic errors in motor tasks using a shooting task that presented target zones (where participants should aim) and ironic zones (where participants should not aim) to field hockey players. As a result, shots into the ironic zone increased under pressure conditions. Other examples of increased ironic errors under pressure have been found in golf putting (Woodman & Davis, 2008), soccer shooting (Barlow et al., 2016), dart throwing (Barlow et al., 2016; Woodman et al., 2015), rifle shooting (Gorgulu, 2019a), basketball free throws (Gorgulu, 2019b), tennis serves (Gorgulu, 2019c), and catching balls (Gorgulu et al., 2019). If motor control were only impaired by conscious processing, then errors would have no specific direction and would increase random variability. Thus, an increase in errors specifically limited to ironic directions could account for the performance decline under pressure that conscious processing theories cannot explain.
In this regard, Gray et al. (2017) proposed that the above errors (random vs. biased error) that occur in conscious processing and ironic processes might be caused by different processes (action execution vs. selection). That study was conducted by dividing baseball pitchers into two groups: a target-only group, in which only the batter’s weak course (target zone into which the ball should be thrown) was presented, and an ironic group, in which the batter’s strong course (ironic zone into which the ball should not be thrown) was presented along with the weak course. In the target-only group, as predicted by conscious processing, pressure decreased, and the number of throws toward the weak and various courses increased, while also increasing the variability of the movement, indicating regression to the early phase of learning (e.g., greater variability in their lead foot landing location and pitching elbow flexion angle). However, in the ironic group, as predicted by the ironic process theory, participants exhibited a decreased and increased number of throws toward the weak and strong (ironic) courses under pressure, respectively, although the variability of throwing movements was identical (i.e., expert-like consistent kinematics) across pressure and nonpressure conditions. Accordingly, the authors suggested that, in situations where the ironic zone is not presented, pressure causes errors in action execution due to conscious processing. In situations where the ironic zone is presented, attention is directed toward undesirable outcomes (i.e., ironic zone), resulting in action selection errors. The abovementioned study, which supports the ironic process theory, generated important findings that could explain a novel mechanism of performance decline from a motor control perspective.
However, in contrast with the predictions of the ironic process theory, the implicit overcompensation hypothesis (de la Peña et al., 2008) posits that a negatively worded instruction, such as “not to do,” rather increases the opposite behavior that avoids ironic error (i.e., overcompensation error). They argued that a negative instruction discouraging a particular action induces an implicit interpretation that it is better to perform the opposite of the instruction, resulting in the execution of movements that excessively avoid the discouraged behavior. For example, in their study using golf putting, they reported that the instruction “don’t putt it short” induced overly long putting (and vice versa). In their view, the instruction “don’t putt it short” triggered implicit messages such as “short is a failure,” emphasizing a negative meaning and producing a compensatory misinterpretation (e.g., over is still better). They argued that the motor program is shaped by this implicit and compensatory interpretation, resulting in overcompensation error (i.e., over putting) that increases the execution of a movement that avoids the instructions (e.g., do not putt short). Research has reported that when under pressure, players tend to adopt strategies to avoid failure (Murayama & Sekiya, 2015); thus, overcompensation errors, which are the polar opposite of ironic errors, are expected to also increase under pressure. Gorgulu (2019c) is the only scholar to have dealt with both ironic and overcompensation errors in the task of tennis serving, reporting an increase in ironic errors (i.e., serving in a particular prohibited direction) under pressure. However, since the direction evaluated as an overcompensation error was not set as directly opposite from the direction of ironic error, further examination is needed to determine which type of error is induced under pressure.
Apart from the above, it is interesting to note that, in the aforementioned study by Gray et al. (2017), ironic instructions did not undermine the kinematics of expert performance under pressure in error production, whereas the effects of pressure without the ironic instruction resulted in a retreat to more novice levels of performance owing to reinvestment/conscious processing. This suggests that presenting the ironic zone on a certain occasion may lead to avoidance of pressure-induced errors caused by conscious processing. Wegner (1994) proposed that, during the performance of cognitive and motor skills, the operating process, which consciously directs attention toward desired outcomes, and the monitoring process, which unconsciously directs attention toward undesirable outcomes, work in parallel within attentional capacity. In addition, ironic instructions, such as instructions on what “not to do,” are through to increase the allocation of attention to the monitoring process. Accordingly, the presentation of the ironic zone is thought to preoccupy and consume most of an individual’s attention during the monitoring process. Thus, the allocation of attention to the conscious (operation) processing of movement could be suppressed. For example, as Gray et al. (2017) stated, in pressure situations, coaches can falsely teach pitchers that the course is the batter’s strong course, even though it is truly their weakest, so that pitchers can avoid conscious processing, which in turn increases the frequency of pitches to the batter’s weakest course. Based on the above, appropriate use of the ironic zone (i.e., negatively worded instruction) can be a useful way to avoid pressure-induced errors in actual situations. Considering this, further evidence is needed that supports the observation that conscious processing does not occur under the presentation of the ironic zone.
Hence, in the current study, we focus on the sense of weakness regarding one’s own play. Notably, Murayama and Sekiya (2015) suggested that, under pressure, a sense of weakness is induced, which in turn disrupts motor control. This sense of weakness is likely to lead to poor performance due to conscious processing. For example, in a study on basketball free throws, performance was found to be impaired when participants were primed with information that a group was not good at that task (Krendl et al., 2012). Additionally, it has been proposed that these types of performance declines can be attributed to conscious processing (Beilock & McConnell, 2004). Therefore, the subjective sense of one’s own weakness with respect to a motor task, which is the focus of this study, is possibly also caused by conscious processing under pressure, which may lead to a decrease in performance. However, if the presentation of the ironic zone draws attention to the monitoring process and gives rise to the avoidance of conscious processing, performance decline caused by perceived weakness may be alleviated or even eliminated.
The two conceptual frameworks described above predict quite different results when the ironic zone is presented under pressure; however, no study to date has compared the two simultaneously in a neutral situation. Further, the possibility of potential merit in ironic presentations would be useful to examine in various situations. Thus, the purpose of this study was to clarify the influence of pressure and one’s perceived sense of weakness on ironic/overcompensation errors, as well as to test whether presenting the ironic zone can help reduce the influence of conscious processing on performance. Accordingly, a tennis stroke task was used in the current study. Tennis rallies require players to return the ball using both their forehand and backhand, with many players often having weakness in hitting one or the other. Therefore, the participants were asked to perform a stroke task under both low and high pressure as well as less- and more-weakness conditions. In accordance with the abovementioned previous research, the participants were presented with the target and ironic zones before being asked to return the ball to the target zone as frequently as possible. It was deemed that, if pressure or perceived weakness did increase the focus of attention to the ironic zone, then ironic errors or overcompensation errors would be increased in the high-pressure and more-weakness conditions. However, if it did not, then random errors were expected to increase, regardless of the zone, as predicted by conscious processing. Additionally, if the presentation of the ironic zone resulted in reduced conscious processing, then an increase in random errors due to high pressure and more weakness would be avoided.
Methods
Participants
The participants were seven male and five female players (age: 19.5 ± 1.5 years, playing career: 10.5 ± 3.5 years) who belonged to a university tennis team. Participants’ levels ranged from having competed in national tournaments in high school at the varsity level to having played mainly at the regional level in varsity tournaments. That is, skilled rather than novice players were chosen as the participants because conscious processing theory assumes that pressure violates the automated movements characteristic of skilled individuals. We estimated the sample size by a power analysis (G*Power 3.1; Faul et al., 2007). We selected an ANOVA with repeated measures within the factors option from the F tests family. Analysis indicated that a total of 12 participants would be necessary to detect differences in pressure- and weakness-induced bias of six returned-ball zones that were qualitatively categorized (Table 1) with a power >0.80, assuming α = .05. This was based on the large effect size (
Number of Returned Balls in Each Zone With Each Condition, M ± SD
| Condition | Ironic | Near-ironic | Target | Near-target | Overcompensation | Near-overcompensation | Others |
|---|---|---|---|---|---|---|---|
| Low pressure, less weakness | 5.3 ± 2.2 | 0.5 ± 0.8 | 14.9 ± 3.6 | 1.0 ± 1.2 | 4.0 ± 2.4 | 0.6 ± 1.2 | 1.6 ± 1.8 |
| Low pressure, more weakness | 4.0 ± 1.8 | 0.8 ± 1.5 | 13.4 ± 3.1 | 1.3 ± 1.1 | 4.3 ± 1.4 | 1.0 ± 1.7 | 2.0 ± 1.4 |
| High pressure, less weakness | 2.6 ± 2.3 | 0.3 ± 0.6 | 14.8 ± 4.7 | 0.8 ± 0.8 | 8.7 ± 4.4 | 0.7 ± 1.4 | 0.6 ± 1.0 |
| High pressure, more weakness | 2.8 ± 2.0 | 0.4 ± 0.9 | 13.6 ± 4.3 | 0.9 ± 1.0 | 8.3 ± 2.5 | 1.2 ± 1.8 | 0.5 ± 0.6 |
Note. Near zone includes both sides of each ironic, target, and overcompensation zone.
—Illustration of the target and ironic zones in the tennis-stroke task. Participants were required to hit a ball coming from a ball-projection machine back to the target zone. They were also instructed not to hit back into the ironic zone. Half of the participants performed the task with the ironic zone in front of the target and the other half with the ironic zone behind the target. These tasks were performed with relatively less or more weakness (backhand or forehand shots) and by a team of three. In the low-pressure condition, the players were asked to return as many balls as possible to the target zone (30 balls per player). In the high-pressure condition, 100 yen was added for each ball returned to the target zone. If the ball was returned to the ironic zone, the prize money earned thus far was reset (i.e., 0 yen). If the ball was returned to any other zone, including the overcompensated zone, it was counted as +0 yen.
Citation: Journal of Sport & Exercise Psychology 46, 3; 10.1123/jsep.2023-0042
Experimental Task and Apparatus
The experimental task was a tennis stroke task in which a ball was launched from a ball projection machine and hit back toward a target zone in the cross direction of an opposite side of the court with relatively less or more weakness (backhand or forehand shots; Figure 1). The ball projection machine (TQ-2000 H II, Tanaka Electric Co.) was positioned 1 m from the point of contact between the center service line and the service line in the direction of the sideline. Additionally, yellow line markers were placed on the court (as shown in Figure 1, solid line) so that participants could see both the target zone, to which they were supposed to return the ball, and the ironic zone, to which they were told not to return the ball. To avoid the position of the ironic zone influencing the results, half of the participants performed the task in a situation where the ironic zone was in front of the target and the other half in a situation where the ironic zone was behind the target. Before the task, the participants were verbally informed of the target and ironic zones. These tasks were performed under low- and high-pressure conditions (see below for details). Two video cameras were set up to record the landing positions of the balls returned by the participants. One video camera was placed on the side of the tennis court to record the entire foreside, while another video camera covered the entire backside.
Procedure
First, a visual analog scale (VAS) was used to investigate the degree to which each participant experienced backhand and forehand weakness. Specifically, participants were presented with a sheet comprising a 10-cm horizontal line, along which were written the phrases “I am not good at forehand,” “I am not good at backhand,” and “I am neither of them” at the right end, left end, and in the center, respectively (Figure 2). The participants were then asked to mark the point that best corresponded to their perceived weakness. Participants were also instructed to mark the center if they did not feel weakness in either their forehand or backhand abilities. The length from the center was evaluated as the degree of weakness and assigned to the weakness condition, with the more-weakness condition assigned to the nonpreferred hitting style and the less-weakness condition to the preferred hitting style. As a result, six participants were classified as having weakness with the forehand and six as having weakness with the backhand (Figure 2). Before the experimental task, the participants were divided into four groups, one of which consisted of three participants with similar perceived weakness, as shown in Figure 2.
—Visual analog scale results for each participant’s perceived weakness and the experimental grouping based on the visual analog scale results. Participants who marked the left of the center line were classified as having weakness with the backhand, and those who marked the right were classified as having weakness with the forehand. The midpoint meant that there was no sense of weakness in either forehand or backhand. The groups, in performing the experimental task, were categorized according to the direction and degree of their perceived weakness.
Citation: Journal of Sport & Exercise Psychology 46, 3; 10.1123/jsep.2023-0042
Subsequently, after warming up as usual, including rallies, the players performed a stroke task under low and high pressure in less- and more-weakness conditions. In this study, we manipulated pressure by combining social responsibility (i.e., other people are depending on you to succeed) and reward (i.e., score and/or monetary incentive; Beilock & Carr, 2001; Carr, 2015). In the low-pressure condition, three players in the group were asked to alternately hit a series of balls projected from the ball projection machine and return as many balls as possible to the target zone. The members were also told that the total number of balls returned to the target zone would be the group’s score and that they should try to score more points than the other groups. In addition to this, the players were instructed to avoid hitting toward the ironic zone. The number of trials was 30 trials per person, with a total of 90 trials for the group. In the high-pressure condition, monetary prizes were awarded instead of scores, following the study of Woodman et al. (2015), and the procedures were otherwise identical to those in the low-pressure condition. Specifically, participants were told that, if they returned to the target zone, they would receive +100 yen (approximately, 1 USD), for other zones +0 yen, and if they returned to the ironic zone, all amounts accrued up until that time would be reset. To increase the competitive pressure, participants were informed that these monetary awards would only be given to the team with the highest number of points. Thus, the success or failure of each participant influenced that of the entire group. The experimental day for each group was conducted on a separate date without them being informed of the results achieved by other groups.
Thus, participants performed in the less- and more-weakness conditions, consecutively, under low and high pressure (4 conditions × 30 trials per participant for a total of 120 trials). The group members remained the same throughout the experiment to avoid factors other than pressure and weakness affecting the performance depending on the combination of members (e.g., motivational changes resulting from differences in the skill levels of the members within the group) and to equalize or minimize the effect of the presence of the others across conditions. A 5-min break was provided between conditions to eliminate the effects of fatigue. The participants were tennis players who commonly practiced for long periods and had a deep understanding of their own physical condition. Therefore, after the 5-min break, each group member was asked, at a subjective level, if it was acceptable to start the next condition in terms of fatigue. When all members stated that fatigue was not a problem, the task for the next condition was initiated. The order of the two pressure and two weakness conditions were counterbalanced among groups. In addition, we instructed the participants to hit the ball back to the target zone as accurately as possible throughout the experiment. This is because the movement speed is modulated under pressure, which may affect performance accuracy (e.g., Nieuwenhuys & Oudejans, 2010). Therefore, only accuracy was emphasized so as not to affect such natural changes.
Measurements and Data Analysis
The frequencies of balls returned to each zone in each condition were determined for each participant based on the video footage recorded during the experimental task. For this purpose, the ball-landing positions were divided into six qualitatively different zones (see dotted line in Figure 1). Specifically, the left and right sides of each zone were labeled as near-target, near-ironic, and near-overcompensation zones in addition to the target, ironic, and overcompensation zones. Moreover, for each of the six zones, the number of returned balls was calculated (Figure 3). The near zone was calculated as the average of the number of balls that landed in both the left and right zones (Table 1). Any zone outside of the six zones, including outside the court or a net ball, was counted as a different zone. Since the positions of the ironic zone were counterbalanced across participants, as mentioned above, the depth direction was reversed, and the data were calculated in such a way that both groups were processed equally. The foreside and backside positions were also inverted and counted in the left-right direction so that they were placed in the same position.
—Ratio of return location for each condition. Values in parentheses indicate the SD. Gray indicates zones that are significantly lower in the same zone between conditions, and black indicates zones that are significantly higher in the Friedman test, which used the average number of balls in each zone. Note that participants were divided into two groups: one with the ironic zone placed in front of the target and one with the ironic zone placed behind the target. The data for the latter group were therefore inverted so that the placement in both groups was the same. Thus, if the ball landed in the zone in front of the target, it was classified as an overcompensation error.
Citation: Journal of Sport & Exercise Psychology 46, 3; 10.1123/jsep.2023-0042
To estimate the effects of pressure and sense of weakness on error type in the tennis stroke task, we conducted a conventional group-level analysis with a sample size of 12 participants and a zone-level analysis with the sample size being the number of balls returned to each zone. For the group-level analysis, parametric distribution was not visible in the 12 data points. Therefore, a nonparametric Friedman test was conducted to investigate differences in the number of returned balls in each zone among experimental conditions. We used the Wilcoxon paired test and reported effect size (r) for the multiple comparisons between experimental conditions. This analysis was performed in accordance with previous research on ironic errors, allowing for comparison between previous findings and the results of this study. For zone-level analysis, the total number of returned balls in each zone was counted in each experimental condition regardless of the participant, and the return rate (risk) was determined for each zone relative to the target zone. In addition, risk ratios (RRs [95% CIs]) in the ironic zone, overcompensation zone, and other zones were calculated as the effect sizes of task conditions (pressure and/or perceived weakness) by comparing each condition and showed the increase or decrease in the rate of return of the ball in each zone. CIs were considered significant if they did not contain 1. This analysis allowed the magnitude of the effect of pressure and sense of weakness on error type to be shown visually and easily interpreted as a change in return likelihood. Statistical analysis software (IBM, SPSS for Windows version 22) was used for the above statistical analysis, and the significance level was set at less than 5%. Effect sizes were reported using r and RR. Effect sizes, r, were interpreted as .1 for small, .3 for medium, and .5 for large, following Cohen (1992). RR was interpreted as 1.22 for small, 1.86 for medium, and 3.00 for large, following Olivier et al. (2017).
The data supporting the findings of this study are openly available in figshare at http://doi.org/10.6084/m9.figshare.19398188.
Results
Table 1 illustrates the average number of balls returned to each zone under each condition for all participants (see also Figures 3 and 4). As mentioned above, participants were divided into two groups: one with the ironic zone placed in front of the target and one with the ironic zone placed behind the target. The data for the latter group were therefore inverted so that the placement was as shown in Figure 3. Thus, if the ball landed in the zone in front of the target, it was classified as an overcompensation error. The individual data shown in Figure 4 were also inverted in the same way.
—Individual data of each error type. The other errors include all errors aside from the ironic and overcompensation errors.
Citation: Journal of Sport & Exercise Psychology 46, 3; 10.1123/jsep.2023-0042
To clarify the influence of the task condition on the characteristics of errors using group-level analysis, a nonparametric Friedman test was applied to the number of returned balls in each zone separately. First, there was no significant difference in the number of balls returned to the target zone between conditions, χ2(3) =1.49, p = .69. Conversely, however, there was a significant difference in the effect of pressure and weakness level on the number of balls returned to the ironic zone, χ2(3) = 11.31, p = .01. Post hoc analysis with Wilcoxon signed-rank tests showed that the number of balls returned to the ironic zone in the low-pressure with less-weakness condition, M (interquartile range [IQR]) = 5 (4–6), was significantly higher than that in the high pressure with less weakness with medium effect size, M (IQR) = 2 (1.25–2.75), p = .04, r = .43, and in the more-weakness condition with a large effect size, M (IQR) = 3 (0.25–5), p = .01, r = .51.
The number of balls returned to the overcompensation zone also differed significantly between conditions, χ2(3) = 21.47, p < .001. Post hoc analysis demonstrated that the number of balls in the high-pressure with less-weakness condition, M (IQR) = 7 (6.25–8.75), was significantly higher than that in the low-pressure with less-weakness condition with a medium effect size, M (IQR) = 4 (2–6), p = .02, r = . 50, and in the more-weakness condition with large effect size, M (IQR) = 4.5 (3.25–5.75), p = .004, r = .59. Similar results were observed in the high-pressure with more-weakness condition, M (IQR) = 8 (6.25–10). Significant differences were observed between the previous condition and the low-pressure with less-weakness, p = .002, r = .63, and more-weakness conditions, p = .002, r = .63.
Based on zone-level analysis, Figure 5 demonstrates the increase or decrease in the probability of a ball being returned to each zone by pressure (upper panel) and weakness (bottom panel). First, high pressure significantly increased the probability of a ball being returned to the overcompensation zone by 1.75 times, 95% CI [1.31, 2.35], in the less-weakness condition and by 1.55 times, 95% CI [1.17, 2.05], in the more-weakness condition. In contrast, the probability of a ball being returned to the ironic zone (less-weakness condition: 0.60, 95% CI [0.41, 0.88], more-weakness condition: 0.70, 95% CI [0.46, 1.05]) and other zones (less-weakness condition: 0.76, 95% CI [0.55, 1.05], more-weakness condition: 0.75, 95% CI [0.58, 0.97]) tended to decrease significantly under high pressure.
—Increase/decrease in return rate as a function of pressure (upper panel) and perceived weakness (bottom panel). RR = risk ratio; CI = confidence interval.
Citation: Journal of Sport & Exercise Psychology 46, 3; 10.1123/jsep.2023-0042
Second, weakness tended to increase the probability of the ball being returned to the other zone by 1.35 times, 95% CI [1.05, 1.74] in the low-pressure condition and 1.34 times, 95% CI [0.97, 1.85] in the high-pressure condition, while no significant increase or decrease was found in the ironic and overcompensation zones.
Discussion
This study aimed to determine the influence of pressure and a sense of one’s own weakness on ironic and overcompensation errors, as well as to test whether presenting the ironic zone helped to reduce the influence of conscious processing on performance. To this end, the participants were asked to perform a tennis stroke task under four conditions with varying degrees of pressure and perceived weakness. The main findings were that there were no differences between conditions in the rate of return to the target zone, pressure decreased the rate of return to the ironic zone and increased the rate of return to the overcompensation zone (Figure 3 and Table 1), and perceived weakness increased the rate of return to the other zones (Figure 5).
A striking finding in this study was that, unlike the findings reported in previous studies (i.e., that pressure increases ironic errors; Barlow et al., 2016; Gray et al., 2017; Woodman et al., 2015), we found that pressure rather decreases ironic errors and increases overcompensation errors. In the present study, ironic errors significantly decreased under pressure and were replaced by a significant increase in overcompensation errors (Figure 3). More precisely, pressure decreased the probability of a ball being returned to the ironic zone by a factor of 0.60–0.70, whereas it increased the probability of a ball being returned to the overcompensation zone by a factor of 1.55–1.75 (Figure 5). Although their study did not deal with pressure, Binsch et al. (2009) found a mix of participants who exhibited ironic errors and those who exhibited overcompensation errors, with 33.3%–51.9% of participants exhibiting overcompensation errors (Binsch et al., 2009; Russell & Grealy, 2010). Contrastingly, in the current study, pressure increased overcompensation errors in nine (75%) of the participants under the less-weakness condition, and in 12 (100%) of the participants under the more-weakness condition. Binsch et al. (2009) reported differences in gaze location as the difference between those who made ironic errors and those who made overcompensation errors: those who gazed at the ironic zone made ironic errors and those who gazed at the overcompensation zone made overcompensation errors. Considering the above, it is possible that the increase in overcompensation errors in this study was associated with an increase in gazing at the overcompensation zone under pressure. Although actual gaze behavior needs to be investigated in this regard, the increase in overcompensation errors when participants are under pressure is a new finding.
Several possible reasons could explain the different results observed in the previous study, in which pressure increased ironic errors, in stark contrast with the current study, in which pressure increased overcompensation errors. One possibility is that risk-averse behavior might have been more strongly encouraged in the present study. In the previous study, a gain (+1 point) in the target zone was equal to a loss (−1 point) in the ironic zone (e.g., Woodman et al., 2015), whereas in the current study, a gain in the target zone was +100 yen and the prize money was reset if the ball was returned to the ironic zone under the high-pressure condition. In this case, it is possible that the risk in the ironic zone was too high, and so participants hit the ball harder (or softer) overall to avoid this zone, resulting in a higher probability of the ball being returned to the overcompensation zone (+0 yen), which was farther away than the ironic zone. While this interpretation appears to be plausible at first glance, it would not result in an overall shift in the direction of overcompensation, as can be seen in Figure 3. Specifically, if the strategy were to hit the ball harder or softer overall to avoid the ironic zone, then errors in the target zone, near-target, and near-overcompensation zones should have also increased in addition to those in the overcompensation zone. However, the results showed that pressure increased only +0.9% on the left side of the near-overcompensation zone and +1.4% on the right side of the near-overcompensation zone in the less- and more-weakness conditions, respectively. One exception was the overcompensation zone, where it was either equaled or decreased with respect to the other zones. Moreover, risk-seeking has been observed to occur in situations where the gain and loss resulting from a behavior are inequivalent and the gain could be zero, as in the current study, without risk-aversive behavior occurring in those situations (Ota et al., 2015, 2016, 2020). In such situations, competitive pressure could enhance risk-seeking behavior (Tanae et al., 2021). Therefore, the increase in overcompensation errors under pressure found in this study is unlikely to be the result of simply encouraging risk-averse behavior.
Another possible explanation is differences in the mode of motor control induced by the task demands. Increased ironic errors have been reported by studies using field hockey shooting (Barlow et al., 2016; Woodman et al., 2015), baseball fastball throwing (Gray et al., 2017), or tennis serving at maximal speed (Gorgulu, 2019c) tasks. These types of high-velocity movements require a specific sequence to generate distal end velocity, and this sequence would strongly constrain motion in the early phase (Hirashima et al., 2008; Koike et al., 2019; Putnam, 1991). Furthermore, it has been hypothesized that, in fast, short-duration movements such as hitting and pitching, the motor program selected before the onset of movement contributes strongly to motor control (preprogrammed control; Schmidt, 1975). Considering the above, with this type of motor control, movement is difficult to adjust through online control, resulting in the dominantly preselected motor program affecting the outcome. Thus, ironic errors resulting from failures in action selection (Gray et al., 2017) are expected to have a stronger impact on the outcome of movements that require preprogrammed control.
Conversely, the tennis stroke task in this study did not require participants to hit the ball at maximum velocity, but only to hit the target zone as accurately as possible. In this case, since the hitting movement is relatively slow, and given the submaximal velocity, online adjustment of the movement (feedback control) is likely to be performed with sensory feedback such as visual information of the ball position and one’s own kinesthesia information. Toner et al. (2013), using a golf putting task that is thought to be performed by feedback control, investigated the kinematics involved in overcompensation errors. They reported that unskilled golfers who exhibited overcompensation errors were a result of the putter path heading in the opposite direction to the target at the point of impact, even though it was heading in the target direction before impact. In other words, overcompensation errors are thought to be caused by overcompensated online adjustment in the late phase of movement and likely to be more pronounced in tasks where feedback control is dominant, as in the present study. Consistent with this, most studies that reported increased overcompensation errors (but not in the case of pressure) have dealt with golf putting tasks, which require distance or position adjustments through feedback control (Beilock et al., 2001; Binsch et al., 2009; de la Peña et al., 2008; Toner et al., 2013). Thus, the degree to which online control is determined by mechanical factors, and in situations where feedforward control dominantly operates, ironic errors due to action selection errors before the initiation of movement are expected to increase. Contrastingly, in situations where feedback control dominantly operates, overcompensation errors due to excessive online adjustments after the initiation of movement are expected to increase.
Lawrence et al. (2013) investigated the effects of pressure on feedforward (preprogrammed) and feedback control using an aiming task in which the participants moved their hands to a target point at a specified time. They found that the increase in errors under pressure depends on feedback control, rather than on feedforward control. Considering these results, the increase in overcompensation errors under pressure in the present study may depend on the feedback control induced by the precise stroke requirement. However, to support this claim, it is necessary to verify whether an increase in ironic errors is observed when strokes are made at the maximum speed under the same conditions, as in the present study, and whether the number of ironic errors changes systematically when the ball return speed is experimentally manipulated. Moreover, kinematics measurements, as used in Toner et al. (2013) and Lawrence et al. (2013), are required.
In addition to the above explanations, self-control strength may explain the differences in error types that occur under pressure observed in the present and previous studies. Englert and Bertrams (2015) proposed an integrated model of attention control theory (Eysenck et al., 2007) and a strength model of self-control (Baumeister et al., 1998). According to the attention control theory, pressure-induced anxiety causes distraction, which results in a lack of additional resources that can be used for a task, negatively affecting cognitive tasks and perceptual-motor control, both of which require attention regulation. However, this negative effect is believed to depend on the ability to counteract distraction (Eysenck et al., 2007; Nieuwenhuys & Oudejans, 2012). Englert and Bertrams (2015) proposed that the abovementioned ability is actually the strength model of self-control, defined as “a process in which predominant impulses can be volitionally overridden to achieve a specific goal” (Baumeister et al., 1998). More precisely, failure to invest additional self-control resources to counter automatic bottom-up information processing (which leads to distraction) impairs the ability to direct attention to task-relevant stimuli and reduces performance. Conversely, if self-control strength is sufficient, then this negative effect on attentional regulation can be prevented and performance decline can be avoided. Considering the aforementioned, it can be argued that, depending on the level of self-control strength, athletes with fully available resources might act in line with the overcompensation hypothesis, as they have the resources to follow the instructions (i.e., follow the instructions on what “not to do”), while athletes with lower resources might act in line with the ironic error hypothesis, as they do not have the resources necessary to follow the instructions. Thus, it is possible that the tasks in the current and previous studies had different effects on this self-control. Given that self-control strength can be weakened by preceding mental tasks, such as resisting temptation (Baumeister et al., 1998; Inzlicht & Schmeichel, 2012), the above-mentioned hypothesis could be tested by determining whether the manipulation of depleting self-control capacity before the main task affects the error type (ironic and overcompensation errors).
Furthermore, in the present study, pressure increased overcompensation errors, whereas self-perceived weakness increased errors made to other zones by a factor of 1.35 (Figure 4). Additionally, as shown in Figure 3, the increase in errors due to self-perceived weakness was more pronounced around the ironic zone in the low-pressure condition and around the overcompensation zone in the high-pressure condition. This means that the task in the more-weakness condition under the presentation of the ironic zone increased random errors around the ironic and overcompensation zones in the bias toward those zones. Since random errors are attributed to conscious processing (Gray, 2004; Gray & Allsop, 2013; Tanaka & Sekiya, 2011), when there is weakness under the presentation of the ironic zone, as in the present study, both errors can be explained by the ironic process theory under the nonpressure condition, while those that can be explained by the implicit overcompensation theory under pressure are considered to have been shifted to errors due to conscious processing.
In addition to the abovementioned gaze behavior, the mechanism by which ironic errors occur has also been explained in terms of attentional capacity (Wegner, 1994). Attentional capacity is allocated between motor and nonmotor events within that capacity. Furthermore, and as mentioned earlier, attention to movement is divided into an operating process and a monitoring process. Ironic errors are thought to be caused by the increased attention required for the monitoring process and the decreased attention required for the operating process. According to this explanation, in the low-pressure condition of the present study, errors due to conscious processing occurred instead of ironic errors because the attention required for operating processing was amplified with the increase in conscious processing toward the movement in which the participants experienced weakness. The overcompensation errors that occurred in the less-weakness task under the high-pressure condition, which also shifted to random errors, could also have been caused by conscious processing induced by self-perceived weakness.
Gray et al. (2017) reported that conscious processing did not occur under conditions in which the ironic zone is presented under pressure. Therefore, the current study aimed to examine whether the presentation of ironic zones would inhibit performance decline due to conscious processing. We expected that random errors would not increase even when perceived weakness was strengthened under the presentation of the ironic zone. However, perceived weakness increased random errors. At first glance, this would appear to indicate that the present results do not provide further support for the notion that the presentation of ironic zones is effective in avoiding errors due to conscious processing. However, no differences in random errors were observed between the high- and low-pressure conditions. Several previous studies have reported that high pressure increases the variability of movement and performance outcomes (e.g., Beilock & McConnell, 2004; Ganesh et al., 2019; Gray, 2004; Tanaka & Sekiya, 2011). Thus, the results of this study are consistent with the finding that the presentation of the ironic zone suppresses conscious processing. We manipulated not only pressure but also participants’ sense of weakness to induce conscious processing, whereas Gray et al. (2017) manipulated only pressure. Thus, when attention is more strongly focused on movement (i.e., the operating process), as in the present study, the positive effects of the ironic zone are not likely to be evident. In any case, the presentation of the ironic zone may be useful as a method to suppress conscious processing, which is a factor in performance decline, although further validation is required.
In summary, the present study demonstrates that pressure-induced performance errors for submaximal velocity movements involve overcompensation errors, as predicted by the implicit overcompensation hypothesis. Furthermore, the results suggest that the perceived weakness of this type of movement may promote errors through conscious processing. However, in this study, no differences were found in the rate of return to the target zone among conditions, although the low-pressure with less-weakness condition had the highest value (Figure 3). Therefore, it should be noted that the results obtained from this study must be interpreted as follows. In movements in which successes and errors occur together over repeated trials, the presentation of the ironic zone transformed the type of error from ironic to overcompensation under pressure, and the presence of weakness transformed ironic and overcompensation errors into conscious processing errors, while the success rate did not change.
This study is useful in that we have demonstrated decreasing and increasing specific types of errors under pressure conditions instead of random errors that are predicted by conscious processing; all of this is set in the context of skilled tennis players who have well-learned or automated performances. The rationale for testing skilled athletes was due to previous findings showing that conscious processing increases as the participant’s level of expertise increases (Baumeister, 1984; Beilock et al., 2001; Lewis & Linder, 1997). However, the limited sample size does increase the risk of Type I and II errors in our experiments. The clear statistical findings, along with the large effect sizes, provide some reassurance regarding the veracity of the findings. Although the individual participant data points suggest, most of the time, that the manipulations had systematic effects on most, if not all, of the participants, future research should demonstrate the replicability of the findings and test how widely they might be generalized to participants of lesser skill in the task in question.
Moreover, in this study, the same pressure-evoking method as that used in a previous study was employed; however, the intensity was not considered sufficiently strong to cause a decrease in performance. The affective demands induced by pressure in the environment differ from those that occur in actual competition situations, and this difference also affects the resulting behavior (e.g., Maloney et al., 2018). Therefore, interpretation of the results in this study is limited to pressure at levels that do not cause performance decrement. Additionally, it has been reported that those who make ironic errors have higher anxiety than those who make overcompensation errors (Russell & Grealy, 2010). However, the present study did not measure state anxiety during the experiment. State anxiety is associated with trait anxiety under pressure (Horikawa & Yagi, 2012). Therefore, future research should also examine the relationship between pressure-induced stress intensity and these errors by measuring anxiety and arousal levels induced by the pressure.
Conclusion
While previous studies have shown that pressure induces ironic errors and conscious processing errors, the results of this study indicate that pressure induces overcompensation errors depending on the motor control mode, providing useful information for avoiding errors under pressure. Therefore, in practical situations, it appears possible to estimate the error tendency based on the movement type or to control the information under pressure to avoid the performance decrement. Additionally, the thought of avoiding a particular outcome can occur even without explicit instruction (Murayama & Sekiya, 2015), as shown in the current study. Thus, even seemingly random errors that occur under pressure may depend on prohibited thoughts that arise during task execution and may be the result of systematic errors such as those observed in this study. Therefore, future studies should investigate participants’ thoughts regarding the discouraged behavior during the experiment to further clarify the mechanism of pressure-induced performance decline.
Acknowledgments
This research was supported by grants from Japan Society for the Promotion of Science KAKENHI (Number JP19H04001 to HN).
References
Barlow, M., Woodman, T., Gorgulu, R., & Voyzey, R. (2016). Ironic effects of performance are worse for neurotics. Psychology of Sport and Exercise, 24, 27–37.
Baumeister, R.F. (1984). Choking under pressure: Self-consciousness and paradoxical effects of incentives on skillful performance. Journal of Personality and Social Psychology, 46(3), 610–620.
Baumeister, R.F., Bratslavsky, E., Muraven, M., & Tice, D.M. (1998). Ego depletion: Is the active self a limited resource? Journal of Personality Social Psychology, 74(5), 1252–1265.
Baumeister, R.F., & Showers, C.J. (1986). A review of paradoxical performance effects: Choking under pressure in sports and mental tests. European Journal of Social Psychology, 16(4), 361–383.
Beilock, S.L., Afremow, J.A., Rabe, A.L., & Carr, T.H. (2001). “Don’t miss!” The debilitating effects of suppressive imagery on golf putting performance. Journal of Sport & Exercise Psychology, 23(3), 200–221.
Beilock, S.L., Bertenthal, B.I., McCoy, A.M., & Carr, T.H. (2004). Haste does not always make waste: Expertise, direction of attention, and speed versus accuracy in performing sensorimotor skills. Psychonomic Bulletin & Review, 11(2), 373–379.
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.
Beilock, S.L., & McConnell, A.R. (2004). Stereotype threat and sport: Can athletic performance be threatened?. Journal of Sport & Exercise Psychology, 26(4), 597–609.
Binsch, O., Oudejans, R.R.D., Bakker, F.C., & Savelsbergh, G.J.P. (2009). Unwanted effects in aiming actions: The relationship between gaze behavior and performance in a golf putting task. Psychology of Sport and Exercise, 10(6), 628–635.
Carr, T.H. (2015). Strengths and weaknesses of reflection as a guide to action: Pressure assails performance in multiple ways. Phenomenology and the Cognitive Sciences, 14(2), 227–252.
Cohen, J. (1992). Statistical power analysis. Current Directions in Psychological Science, 1(3), 98–101.
de la Peña, D., Murray, N.P., & Janelle, C.M. (2008). Implicit overcompensation: The influence of negative self-instructions on performance of a self-paced motor task. Journal of Sports Sciences, 26(12), 1323–1331.
Englert, C., & Bertrams, A. (2015). Integrating attentional control theory and the strength model of self-control. Frontiers in Psychology, 6, Article 824.
Eysenck, M.W., & Calvo, M.G. (1992). Anxiety and performance: The processing efficiency theory. Cognition and Emotion, 6(6), 409–434.
Eysenck, M.W., Derakshan, N., Santos, R., & Calvo, M.G. (2007). Anxiety and cognitive performance: Attentional control theory. Emotion, 7(2), 336–353.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191.
Ganesh, G., Minamoto, T., & Haruno, M. (2019). Activity in the dorsal ACC causes deterioration of sequential motor performance due to anxiety. Nature Communications, 10, Article 4287.
Gorgulu, R. (2019a). An examination of ironic effects in air-pistol shooting under pressure. Journal of Functional Morphology and Kinesiology, 4(2), Article 20.
Gorgulu, R. (2019b). Counter-intentional errors of basketball free throw shooting under elevated pressure: An educational approach of task instruction. Journal of Education and Learning, 8(2), 89–97.
Gorgulu, R. (2019c). Ironic or overcompensation effects of motor behaviour: An examination of a tennis serving task under pressure. Behavioral Science, 9(2), Article 21.
Gorgulu, R., Cooke, A., & Woodman, T. (2019). Anxiety and ironic errors of performance: Task instruction matters. Journal of Sport & Exercise Psychology, 41(2), 82–95.
Gray, R. (2004). Attending to the execution of a complex sensorimotor skill: Expertise differences, choking, and slumps. Journal of Experimental Psychology: Applied, 10(1), 42–54.
Gray, R., & Allsop, J. (2013). Interactions between performance pressure, performance streaks, and attentional focus. Journal of Sport & Exercise Psychology, 35(4), 368–386.
Gray, R., Orn, A., & Woodman, T. (2017). Ironic and reinvestment effects in baseball pitching: How information about an opponent can influence performance under pressure. Journal of Sport & Exercise Psychology, 39(1), 3–12.
Gröpel, P., & Mesagno, C. (2019). Choking interventions in sports: A systematic review. International Review of Sport and Exercise Psychology, 12(1), 176–201.
Hanin, Y.L. (2000). Individual Zones of Optimal Functioning (IZOF) model: Emotion-performance relationship in sport. In Y.L. Hanin (Ed.), Emotions in sport (pp. 65–89). Human Kinetics.
Higuchi, T. (2000). Disruption of kinematic coordination in throwing under stress. Japanese Psychological Research, 42(3), 168–177.
Hirashima, M., Yamane, K., Nakamura, Y., & Ohtsuki, T. (2008). Kinetic chain of overarm throwing in terms of joint rotations revealed by induced acceleration analysis. Journal of Biomechanics, 41(13), 2874–2883.
Horikawa, M., & Yagi, A. (2012). The relationships among trait anxiety, state anxiety and the goal performance of penalty shoot-out by university soccer players. PLoS One, 7(4), Article e35727.
Inzlicht, M., & Schmeichel, B.J. (2012). What is ego depletion? Toward a mechanistic revision of the resource model of self-control. Perspectives on Psychological Science, 7(5), 450–463.
Koike, S., Ishikawa, T., Willmott, A.P., & Bezodis, N.E. (2019). Direct and indirect effects of joint torque inputs during an induced speed analysis of a swinging motion. Journal of Biomechanics, 86, 8–16.
Krendl, A., Gainsburg, I., & Ambady, N. (2012). The effects of stereotypes and observer pressure on athletic performance. Journal of Sport & Exercise Psychology, 34(1), 3–15.
Lawrence, G.P., Khan, M.A., & Hardy, L. (2013). The effect of state anxiety on the online and offline control of fast target-directed movements. Psychological Research, 77(4), 422–433.
Lewis, B.P., & Linder, D.E. (1997). Thinking about choking? Attentional processes and paradoxical performance. Personality and Social Psychology Bulletin, 23(9), 937–944.
Liu, S., Eklund, R.C., & Tenenbaum, G. (2015). Time pressure and attention allocation effect on upper limb motion steadiness. Journal of Motor Behavior, 47(4), 271–281.
Maloney, M.A., Renshaw, I., Headrick, J., Martin, D.T., & Farrow, D. (2018). Taekwondo fighting in training does not simulate the affective and cognitive demands of competition: Implications for behavior and transfer. Frontiers in Psychology, 9, Article 25.
Masters, R.S.W. (1992). Knowledge, knerves and know-how: The role of explicit versus implicit knowledge in the breakdown of a complex motor skill under pressure. British Journal of Psychology, 83(3), 343–358.
Murayama, T., & Sekiya, H. (2015). Factors related to choking under pressure in sports and the relationships among them. International Journal of Sport and Health Science, 13, 1–16.
Nieuwenhuys, A., & Oudejans, R.R.D. (2010). Effects of anxiety on handgun shooting behavior of police officers: A pilot study. Anxiety, Stress, & Coping, 23(2), 225–233.
Nieuwenhuys, A., & Oudejans, R.R.D. (2012). Anxiety and perceptual-motor performance: Toward an integrated model of concepts, mechanisms, and processes. Psychological Research, 76(6), 747–759.
Olivier, J., May, W.L., & Bell, M.L. (2017). Relative effect sizes for measures of risk. Communications in Statistics—Theory and Methods, 46(14), 6774–6781.
Ota, K., Shinya, M., & Kudo, K. (2015). Motor planning under temporal uncertainty is suboptimal when the gain function is asymmetric. Frontiers in Computational Neuroscience, 9, Article 88.
Ota, K., Shinya, M., & Kudo, K. (2016). Sub-optimality in motor planning is retained throughout 9 days practice of 2250 trials. Scientific Reports, 6(1), Article 37181.
Ota, K., Tanae, M., Ishii, K., & Takiyama, K. (2020). Optimizing motor decision-making through competition with opponents. Scientific Reports, 10(1), Article 950.
Putnam, C.A. (1991). A segment interaction analysis of proximal-to-distal sequential segment motion patterns. Medicine & Science in Sports Exercise, 23(1), 130–144. https://www.ncbi.nlm.nih.gov/pubmed/1997807
Rietschel, J.C., Goodman, R.N., King, B.R., Lo, L.C., Contreras-Vidal, J.L., & Hatfield, B.D. (2011). Cerebral cortical dynamics and the quality of motor behavior during social evaluative challenge. Psychophysiology, 48(4), 479–487.
Russell, C., & Grealy, M.A. (2010). Avoidant instructions induce ironic and overcompensatory movement errors differently between and within individuals. Quarterly Journal of Experimental Psychology, 63(9), 1671–1682.
Schmidt, R.A. (1975). A schema theory of discrete motor skill learning. Psychological Review, 82(4), 225–260.
Tanae, M., Ota, K., & Takiyama, K. (2021). Competition rather than observation and cooperation facilitates optimal motor planning. Frontiers in Sports and Active Living, 3, Article 637225.
Tanaka, Y., & Karakida, K. (2019). Application of ironic processing theory to motor performance overcompensation errors. Japanese Journal of Sport Psychology, 46(1), 27–39.
Tanaka, Y., & Sekiya, H. (2011). The influence of monetary reward and punishment on psychological, physiological, behavioral and performance aspects of a golf putting task. Human Movement Science, 30(6), 1115–1128.
Toner, J., Moran, A., & Jackson, R. (2013). The effects of avoidant instructions on golf putting proficiency and kinematics. Psychology of Sport and Exercise, 14(4), 501–507.
Wegner, D.M. (1994). Ironic processes of mental control. Psychological Review, 101(1), 34–52.
Wegner, D.M. (2009). How to think, say, or do precisely the worst thing for any occasion. Science, 325(5936), 48–50.
Wegner, D.M., Ansfield, M., & Pilloff, D. (1998). The putt and the pendulum: Ironic effects of the mental control of action. Psychological Science, 9(3), 196–199.
Wegner, D.M., Schneider, D.J., Carter, S.R., & White, T.L. (1987). Paradoxical effects of thought suppression. Journal of Personality and Social Psychology, 53(1), 5–13.
Woodman, T., Barlow, M., & Gorgulu, R. (2015). Don’t miss, don’t miss, d’oh! Performance when anxious suffers specifically where least desired. The Sport Psychologist, 29(3), 213–223.
Woodman, T., & Davis, P.A. (2008). The role of repression in the incidence of ironic errors. The Sport Psychologist, 22(2), 183–196.
Zhu, F.F., Poolton, J.M., Wilson, M.R., Maxwell, J.P., & Masters, R.S. (2011). Neural co-activation as a yardstick of implicit motor learning and the propensity for conscious control of movement. Biological Psychology, 87(1), 66–73.
