Scandinavian Journal of Psychology, 2019 DOI: 10.1111/sjop.

12571

Personality and Social Psychology

Psychological characteristics in women football players: Skills, mental
toughness, and anxiety
HAFRUN


KRISTJANSD

OTTIR,1

KAMILLA R. JOHANNSD

OTTIR,2
MIGUEL PIC3 and JOSE M. SAAVEDRA1
1
Physical Activity, Physical Education, Sport and Health (PAPESH) Research Centre, Sports Science Department, School of Social Sciences, Reykjavik
University, Reykjavik, Iceland
2
Department of Psychology, School of Social Sciences, Reykjavik University, Reykjavik, Iceland
3
Group of Research and Teaching Innovation in Physical Activity and Sport, University of La Laguna, La Laguna, Spain

Kristj
ansd
ottir, H., J
ohannsd
ottir, K. R., Pic, M. & Saavedra, J. M. (2019). Psychological characteristics in women football players: Skills, mental
toughness, and anxiety. Scandinavian Journal of Psychology.

Women’s football has been far less studied than men’s. This work’s objectives were to: (1) analyze the differences in psychological skills, mental
toughness (MT), and anxiety in women football players according to their level (national team, first division, and second division); and (2) predict those
three levels (using a multivariate model) according to the players’ psychological skills, mental toughness, and anxiety. One hundred and forty-two
Icelandic women football players (23.5  3.5 years) participated in the study. They were classified into three groups according to their level: national
team, and first and second divisions. Three questionnaires were used: the Test of Performance Strategies Questionnaire, the Sport Mental Toughness
Questionnaire, and the Sport Anxiety Scale-2 questionnaire. A one-way ANOVA with Bonferroni post hoc correction was used to examine differences
between teams. Applying a classification tree analysis, the participants were classified into three groups according to their level. There were few differences
between the three groups in psychological skills, but in mental toughness and anxiety the national team had the highest and lowest values respectively, and
the first and second division players differed in relaxation in competition (TOPS), total score and confidence (SMTQ), and worry (SAS-2). The
classification tree correctly classified 54.9% of the sample with the variables total score (SMTQ) and activation in practice (TOPS). Therefore, given the
relevance that psychological attributes appear to have for women football players’ performance, it would seem indispensable to incorporate the figure of
the sports psychologist into national and club teams.
Key words: team sports, psychology, soccer, performance.
Jose M. Saavedra, Physical Activity, Physical Education, Health and Sport Research Centre (PAPESH), School of Social Sciences, Reykjavik University,
Menntavegur 1, Nautholsvık, 101 Reykjavık, Iceland. E-mail: [email protected]

INTRODUCTION Olivier, 2010), although they may also include other skills such as
activation, attentional control, and automaticity (Thomas, Murphy
Football (soccer) is one of the world’s most popular sports (Kurt, & Hardy,1999). Notwithstanding this amplitude of PS, one must
€
Catikkas, Omurlu & Atalag, 2012). Performance in football is bear in mind that their application during training differs from
based on such factors as physical fitness, technique, tactics, that in competition. Instruments like the Test of Performance
endurance, agility, and coordination. Football players also have to Strategies Questionnaire (TOPS) make this differentiation explicit
be able to jump, kick, tackle, and sprint to succeed (Stølen, (Thomas et al., 1999). Several studies have shown that the most
Chamari, Castagna & Wisløff, 2005). The factors affecting successful athletes have better PS than those who are less
performance in the sport are not only individual, but also successful, and that this is so for both non-Olympic (Neil,
collective (tactics, relationship between players, inter alia) and Mellalieu & Hanton, 2006) and Olympic (Gould & Maynard,
environmental (the team’s town, budget, inter alia). The 2009; Taylor, Gould & Rolo, 2008) athletes.
individual factors include anthropometric and body composition For MT however, researchers have unfortunately been unable
factors (height, weight, body, fat, inter alia), physical fitness to agree upon a common definition of this state. The different
factors (aerobic and anaerobic endurance, power, inter alia), definitions that have been proposed usually refer to the athlete’s
specific skills (kicking speed, passing, inter alia) and ability to concentrate, rebound from failure, cope with pressure,
psychological factors (Reilly, Williams, Nevill & Franks, 2000). and face adversity, as well as mental resilience, commitment, and
In particular, such psychological factors as motivation, confidence (Bull, Shambrook, James & Brooks, 2005). In a recent
psychological skills (PS), and mental toughness (MT) have been review, it was found that athletes competing at a high level had
considered essential for football performance (Jooste, Steyn & greater MT than those competing at lower levels. A vast majority
Van Den Berg, 2014). In the current literature, there are many (90%) of studies that have examined the relationship between MT
studies concerning the relation of physical factors with football and performance have found that athletes with a high level of MT
performance, but studies of the influence of psychological factors perform better than those with lower levels (Cowden, 2017).
are somewhat lacking (Slimani, Bragazzi, Tod et al., 2016). There are differences, however, according to gender and age, with
The PS are trainable psychological abilities that can enhance an MT values being higher in men than in women, and in older
athlete’s performance. Usually, they refer to goal-setting skills, athletes (over 55 years old) than in younger athletes (18–35 years
imagery, self-talk, and relaxation (McCarthy, Jones, Harwood & old) (Zeiger & Zeiger, 2018).

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd
2 H. Kristjansdottir et al. Scand J Psychol (2019)

Researchers and athletes agree that the ability to control one’s football players found differences in confidence, but only between
own emotions, anxiety for example, and to cope with stress both the first and third divisions (Danielsen, Rodahl, Giske &
in and out of competition is an important psychological skill for Høigaard, 2017).
athletes (Mace & Carroll, 1986; Mellalieu, Neil, Hanton & Given this context of so few studies on this topic in women’s
Fletcher, 2009). That being said, it is important to note that the football, the objectives of the present study were to: (1) analyse
relationship between anxiety and performance can be quite the differences in psychological skills, mental toughness, and
complicated. The inverted-U theory states that both high and low anxiety in women football players according to their level
levels of anxiety will affect performance negatively in a similar (national team, first division, and second division); and (2) predict
way. The optimal performance will accrue when the athlete is in the aforementioned levels of the players (using a multivariate
the middle of the curve, i.e., is experiencing a moderate level of model) in terms of their psychological skills, mental toughness,
anxiety (Landers & Arent, 2010). Anxiety levels are related to and anxiety.
other aspects that are part of MT, such as confidence (Kang &
Jang, 2018). Influencing one of them will bring about changes in
the other. A study of elite handball players found women to MATERIAL AND METHODS
generally have higher levels of anxiety than men (Kristj
ansd
ottir,
Erlingsd
ottir, Sveinsson & Saavedra, 2018). Participants
There have been few studies of PS, MT, and anxiety in relation One hundred and forty-two Icelendic women football players
to football performance levels. Almost all studies that do exist (23.5  3.5 years) participated in the study. The players were
were conducted with men football players, especially young male classified into three groups according to their level: the national
players. Their results can be somewhat confusing. In one study, team (n = 32; 25.1  4.7 years), the first division (n = 54;
there was no difference in PS between the selected and de- 23.4  3.8 years), and the second division (n = 56;
selected youth male players in talent development programs 22.8  3.4 years). There were no differences by age between the
(Huijgen, Elferink-Gemser, Lemmink & Visscher, 2014). Similar three levels (p = 0.580). It has to be noted that the national team
results were found in another study (Jooste et al., 2014). players play mostly in the country’s first division, so that their
Concentration was the only variable that was associated with group could also have been called “top first-division players.”
performance level among 14- to 18-year-old players. Other However, the commonly used term is national team (Castagna &
studies have found a somewhat stronger relationship between Castellini, 2013; Jacobson & Tegner, 2007). All the participants
performance level, PS, and anxiety. A study of amputee male were playing in the 2016/2017 season. In a normal week, they
football players who were competing at the World Cup found trained for at least five 90-minute sessions. The study respected
moderate positive correlations of their attentional control, imagery the principles of the Declaration of Helsinki. The data were
use, and relaxation in training, and of their activation in treated anonymously so that there was no possibility of
competition with self-evaluated performance (Lowther, Lane & associating results with the player’s name. All the players
Lane, 2002). Another study (Coetzee, Grobbelaar & Grid, 2006) participated voluntarily in the study without receiving any
found the most important PS discriminators between successful compensation in exchange, and after signing an informed consent.
and less successful male football teams were concentration, It was also approved by the Ethics Committee of Reykjavik
mental preparation, a mental rehearsal, achievement motivation, University.
and activation control. This same study also found that the less
successful teams had less worry than the successful teams
(Coetzee et al., 2006). To this has to be added that professional Measurements
male football players were found to differ in somatic anxiety All the participants completed a battery of three questionnaires.
according to the level of the team they played in – those who The questionnaires were presented in their [deleted for blinding
played in the super league had less anxiety than those who played purposes] versions during training camp (national team) or via
in the first league (Kolayisß & Cß elik, 2017). In the same line were email/facebook (first and second divisions). All players were
findings that highly promotion-worthy young players (based on informed of the objective of the study and that their participation
coach evaluation) had less worry than promotion-worthy players was voluntary. It was also made clear that there were no “correct”
and partly promotion-worthy players (H€oner & Feichtinger, or “incorrect” answers, and that they should respond to the items
2016). Regarding MT, a very recent study (Guillen & Santana, sincerely. The questionnaires are described in the following three
2018) found that older (U19) male football players had higher paragraphs.
MT than younger players (U16), and that there was some The Test of Performance Strategies Questionnaire (TOPS)
difference in MT in relation to the players’ performance level. (Thomas et al., 1999) was used to measure the skills and
Even though women’s football is a major participatory sport for strategies used by sportspersons in practice and competition.
women worldwide, studies on PS, MT, and anxiety among female There are eight sub-scales: goal setting, relaxation, automaticity,
football players are almost non-existent (Gledhill, Harwood & activation, imagery, self-talk, emotional control, and attentional
Forsdyke, 2017). One of the few that do exist found that there control (concentration). The participants responded to the items
was a difference in post-game confidence among collegiate female on a Likert scale from (1) “never” to (5) “always.” The total score
football players between starters and non-starters, favoring the for each sub-scale was taken to be the mean of the responses to
starters (Haneishi, Fry, Moore, Schilling, Li & Fry, 2007). In a its items. A high score indicates greater use of that skill or
similar line, another study conducted with Norwegian women strategy except for negative thinking for which a higher score

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd
Scand J Psychol (2019) Psychological characteristics in football players 3

indicates more problems in that sub-scale. The psychometric on each sub-scale), and to determine which variables
properties of the English version of the questionnaire are known discriminated the teams. This technique basically consists of
to be good (Thomas et al., 1999), and, while those of the applying a binary tree algorithm which segments the data from
Icelandic version of the questionnaire are unknown, the test’s the study’s questionnaires (independent variables) to create
internal consistency was acceptable in this study (Cronbach’s homogeneous subsets with respect to the dependent variable
alpha = 0.85). The Cronbach’s alpha values of each sub-scale in (team) (Alkhasawneh, Ngah, Tay, Isa, Ashidi & Al-Batah, 2014;
practice were: goal setting (a = 0.87), relaxation (a = 0.75), Breiman, Friedman, Olshen & Stone, 1984; Knable, Barci,
automaticity (a = 0.55), activation (a = 0.59), imagery Bartko, Webster & Torrey, 2002). The search parameters started
(a = 0.73), self-talk (a = 0.76), emotional control (a = 0.77), and from levels of statistical significance (p < 0.05), and the depth of
attentional control (a = 0.65). For the sub-scales in competition, the nodes varied between 25 and 50 interactions. Lost cases were
the values were: goal setting (a = 0.81), relaxation (a = 0.83), eliminated and replaced in the construction of the classification
automaticity (a = 0.51), activation (a = 0.71), imagery tree which consisted of a maximum of five classification levels
(a = 0.64), self-talk (a = 0.72), emotional control (a = 0.80), and and two nodes per level. The measure of improvement had to be
negative thinking (a = 0.79). at least 0.001. Finally, cross-validation was applied, using 10% of
The Sport Mental Toughness Questionnaire (SMTQ) (Sheard, the cases for training and the other 90% for the final test.
Golby & van Wersch, 2009) was used to measure mental
toughness. There are three sub-scales: confidence (6 items),
constancy (4 items), and control (4 items). The participants RESULTS
responded to the items on a Likert scale from (1) “not at all true” Table 1 lists the basic descriptive statistics (mean and standard
to (4) “very true.” The total score for each sub-scale was taken to deviation) and the one-way ANOVA with Bonferroni’s post hoc
be the mean of the responses to its items. A higher score indicates test for all the TOPS sub-scales, both in practice and in
more mental toughness. The psychometric properties of the competition. There were differences between national team
English version of the questionnaire are known to be good players and second division players in ten of the sub-scales
(Sheard et al., 2009), and, while those of the Iceandic version of studied. In only four sub-scales (goal setting in practice,
the questionnaire are unknown, its internal consistency was activation in practice, self-talk in practice, and activation in
acceptable in this study (Cronbach’s alpha = 0.82). The competition) were there differences between national team players
Cronbach’s alpha values of each sub-scale were: confidence and first division players.
(a = 0.84), constancy (a = 0.56), and control (a = 0.72). Table 2 lists the basic descriptive statistics (mean and standard
The Sport Anxiety Scale 2 questionnaire (SAS-2) (Smith, deviation) and the one-way ANOVA with Bonferroni post hoc
Smoll, Cumming & Grossbard, 2006) was used to measure test for all the MT and anxiety sub-scales. There were differences
cognitive and somatic anxiety in sport performance. There are between each of the three groups of players in three sub-scales:
three sub-scales: worry, somatic anxiety, and concentration confidence in SMTQ, total in SMTQ, and worry in anxiety. There
disruption with five items each. The participants responded to the were also differences between national team players and first
items on a Likert scale from (1) “not at all” to (4) “very much.” division players in the constancy sub-scale in SMTQ,
The total score for each sub-scale was taken to be the mean of the concentration disruption sub-scale in anxiety, and total anxiety.
responses to its items. A higher score indicates more anxiety. The Regarding the multivariate model, the classification tree
psychometric properties of the English version of the correctly classified 54.9% of the sample. This percentage was
questionnaire are known to be good (Smith et al., 2006), and, distributed differently in the three levels of the dependent variable
while those of the Icelandic version of the questionnaire are (national team: 53.1%, first division: 50%, second division:
unknown, its internal consistency was acceptable in this study 61.1%). Node 0 has the number and percentage of players
(Cronbach’s alpha = 0.82). The Cronbach’s alpha values of each belonging to each team (Figure 1). The first predictor variable
sub-scale were: worry (a = 0.94), somatic anxiety (a = 0.79), and was the total score in SMTQ, taking ≤36,500 as threshold,
concentration disruption (a = 0.84). leading to the compositions of node 1 (FD = 20, 37%; SD = 33,
61.1%; At = 1, 1.9%) and node 2 (FD = 36, 40.9%; SD = 21,
23.9%; NT = 31, 35.2%). Starting from node 2 with values
Data analysis >36,500, the variable activation in practice (TOPS) emerged to
All the variables satisfied the tests of homoskedasticity (Levene’s form node 3 (FD = 28, 50%; SD = 14, 25.%; NT = 14, 25.%)
homogeneity test) and normality (Kolmogorov–Smirnov test). The and node 4 (FD = 8, 25%; SD = 7, 21.9%; NT = 17, 53.1%).
basic descriptive statistics (mean and standard deviation) were The normalized importances of each variable were asymmetric –
calculated. A one-way ANOVA was used to examine whether while the total score in SMTQ explained 100% of the model, the
there were differences between teams. The Bonferroni post hoc variable activation in practice (TOPS) reached 2.3%.
test was used to determine the teams between which there were
such differences. The eta-squared (g2) statistic, which describes
the proportion of the variance attributable to a given factor, was DISCUSSION
calculated. A multivariate classification and regression tree This study has analyzed PS, MT, and anxiety in female football
technique was used to predict which team each player would players according to level (national team, first division, and
belong to (national team, first division, second division) based on second division). In general, there were few variables
the variables studied (total score on each questionnaire and score differentiating the national team and first division players in PS.

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd
4 H. Kristjansdottir et al. Scand J Psychol (2019)

Table 1. Basic descriptors (mean and standard deviation) of PS (The Test of Performance Strategies Questionnaire – TOPS), F-value (one-way analysis
of variance), p-value, and eta squared (g2), and differences between the national, first, and second division teams for each sub-scale

National team First division Second division

n = 32 (A) n = 54 (F) n = 56 (S) F p g2 Differences

Practice
Goal setting 3.48  0.72 3.02  0.84 2.83  0.93 5.806 0.004 0.079 A>F,S
Relaxation 2.40  0.77 2.27  0.71 2.14  0.67 1.272 0.284 0.019 n.s.
Automaticity 3.43  0.31 3.31  0.49 3.31  0.41 0.945 0.391 0.014 n.s.
Activation 3.33  0.57 2.86  0.55 2.83  0.64 7.789 0.001 0.105 A>F,S
Imagery 2.98  0.89 2.81  0.76 2.74  0.78 0.841 0.442 0.012 n.s.
Self-talk 3.61  0.70 3.18  0.70 3.20  0.74 4.258 0.016 0.059 A>F,S
Emotional control 4.46  0.57 3.25  0.55 3.03  0.67 5.152 0.007 0.070 A>S
Attentional control 3.70  0.60 3.44  0.57 3.52  0.49 2.251 0.109 0.032 n.s.
Competition
Goal setting 3.91  0.78 3.52  0.82 3.30  0.91 5.191 0.007 0.071 A>S
Relaxation 3.63  0.55 3.49  0.69 3.11  0.56 8.471 <0.001 0.112 A,F>S
Automaticity 3.00  0.48 3.11  0.59 2.97  0.55 0.889 0.413 0.013 n.s.
Activation 4.37  0.42 4.05  0.55 4.04  0.54 4.741 0.010 0.065 A>F,S
Imagery 3.48  0.76 3.23  0.75 3.03  0.61 4.170 0.017 0.057 A>S
Self-talk 3.39  0.76 3.04  0.71 3.03  0.75 2.919 0.057 0.041 n.s.
Emotional control 3.70  0.50 3.49  0.71 3.27  0.80 3.743 0.026 0.051 A>S
Negative thinking 3.77  0.61 3.62  0.65 3.34  0.75 4.458 0.013 0.061 A>S

n.s., not significant.

Table 2. Basic descriptors (mean and standard deviation) of MT (the Sport Mental Toughness Questionnaire – SMTQ) and anxiety (the Sport Anxiety
Scale 2 questionnaire – SAS-2), F-value (one-way analysis of variance), p-value, and eta squared (g2), and differences between the national, first, and
second division teams for each sub-scale

National team First division Second division

n = 32 (A) n = 54 (F) n = 56 (S) F p g2 Differences

SMTQ
Confidence 16.03  2.66 14.25  2.98 12.29  3.09 15.947 <0.001 0.192 A>F>S
Constancy 14.74  1.34 13.47  1.74 12.87  1.91 11.541 <0.001 0.148 A>F,S
Control 12.61  1.91 11.38  2.74 10.13  2.81 9.217 <0.001 0.119 A,F>S
Total 43.43  4.36 39.94  6.16 35.15  6.38 18.993 <0.001 0.226 A>F>S
SAS-2
Somatic anxiety 8.38  2.12 9.25  3.18 9.85  2.62 2.883 0.059 0.040 n.s.
Worry 8.03  1.99 10.59  3.67 12.11  3.94 13.444 <0.001 0.163 A<S<F
Concentration disruption 5.56  1.19 7.02  2.59 7.04  2.03 5.850 0.004 0.079 A<F,S
Total 21.94  3.71 26.95  8.17 29.06  7.33 9.999 <0.001 0.128 A<F,S

n.s., not significant.

In MT and anxiety, however, both the total and all sub-scales that presented differences between national team and first division
except for control in MT and somatic anxiety showed differences players was “activation,” emphasizing that the importance of
between national team and first division players, and between being “alert” is a characteristic feature of higher level players.
these and the second division players. Regarding the multivariate There have been no studies that have analyzed PS in women’s
model, the classification tree correctly classified 55% of the football. There has only been one work carried out with 15 male
sample with the variables total score in MT and activation in amputee football players in which a correlation was found
practice. To the best of our knowledge, this has been the first between the player’s self-evaluated performance and most of the
study conducted with a sample comprising only women football sub-scales, with the strongest correlation being in imagery in
players that analyses the aforementioned variables. Likewise, this practice (r = 0.69, p < 0.05) and attentional control in practice
is the first time that decision trees have been applied to a sample (r = 0.61, p < 0.05) (Lowther et al., 2002). For another team
of football players from a psychological perspective. sport (rugby), there were also found to be few differences
The PS univariate analysis showed differences between the between elite and non-elite players (Neil et al., 2006), coinciding
national team and second division players, mainly in the with the present study’s results. A further two studies point in the
competition sub-scales. Differences between national team and same line, one on golf players (Hayslip, Petrie, MacIntire &
first division players were few, however. This suggests that, once Jones, 2010) and the other on a sample formed by different sports
a certain level has been reached, the players make a similar use of (Nicholls, Polman, Levy & Backhouse, 2009) which did not find
PS. It has to noted, however, that the only competition sub-scale any differences in PS depending on the level of skill (least skilled,

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd
Scand J Psychol (2019) Psychological characteristics in football players 5

Fig. 1. Regression tree generated by decision tree analysis applied to the study's predictor variables (scales and total of each questionnaire) according to
the player's team (national, and first and second divisions). STMQ-total = total score on the Sport Mental Toughness Questionnaire; TOPS pract
activation = score on the activation practice sub-scale in the Test of Performance Strategies Questionnaire.

moderately skilled and most skilled) in the former, and Finally, with respect to the univariate analysis of anxiety, the
achievement level in the latter. national team scored lower than the first and second division
For MT in the univariate analysis, there were higher confidence players except on the “somatic anxiety” sub-scale. This is
and total scores in national team than in first division players, and coherent with a previous study conducted with South Korean
in the latter higher than in second division players. These data professional men players in which the reduction of anxiety levels
partially agree with those of a study of Norwegian women improved performance in their away games (Kang & Jang, 2018).
football players for whom there were differences in the Similarly, a study of women collegiate football players showed
“confidence” sub-scale, but only between the first and third the starters to have lower anxiety levels than non-starters
division players (Danielsen et al., 2017). Also in the same line as (Haneishi et al., 2007). It is necessary here, however, to highlight
the present results are those of a study which found differences in that, although these studies point in the same line (lower level of
MT in male players of 14–19 years in age depending on their anxiety, better performance), the performance criteria were
team’s classification in the league, with highest scores in the top- different – winning vs. losing, and starter vs. non-starter. It has to
ranked team (Guillen & Santana, 2018). This last work used a be added that anxiety levels can be influenced by the coach’s
different mental toughness index questionnaire (Gucciardi, style (Das & Chatterjee, 2018), but, in spite of this, these levels
Hanton, Gordon, Mallett & Temby, 2015), and indeed this variety have also been used to pick out talent in young male players of
of instruments used is one of the main problems in studying MT 12 years old (H€oner & Feichtinger, 2016).
(Cowden, 2017). However, the questionnaire used in the present With the development of the multivariate classification tree
study, SMTQ, has shown a good correlation (r < 0.709, model, its application offered an alternative perspective to that of
p < 0.001) with other similar instruments such as the previous analyses. For example, the itinerary described by
Psychological Performance Inventory (Loehr, 1986), with men partitioning the data into nodes was revealed. That is, the
football players who play or had played for national teams possibilities were specified of grouping together the predictor
scoring slightly higher on the mental hardiness scores (Wieser & variables according to the three levels of the women’s football
Thiel, 2014). teams. The model thereby correctly classified 55% of the sample

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd
6 H. Kristjansdottir et al. Scand J Psychol (2019)

with the variables total score in MT and activation in practice, “team resilience” (Gucciardi, Crane, Ntoumanis et al., 2018).
although only for values greater than 36,500 did the second Anxiety levels also influence performance, although the focus is
variable activation in practice emerge. These data concur with a the cognitive perspective of anxiety, since its somatic perspective
recent review which indicated that MT represents one of the most showed no differences between the three teams. Therefore, given
important psychological attributes influencing athletic success the relevance that psychological attributes appear to have for
(Cowden, 2017). This relationship between MT and success in women football players’ performance, it would seem
sport is not only found when performance is assessed objectively, indispensable to incorporate the figure of the sports psychologist
but also when the athlete assesses their own performance into national and club teams, which would allow these teams to
subjectively (Meggs, Chen & Koehn, 2019). It should be noted, improve the diagnosis of psychological characteristics, the
however, that MT may be influenced by other psychological interventions applied, and the processes of talent detection


attributes such as task and ego orientation (Alvarez, Walker & (Musculus & Lobinger, 2018).
Castillo, 2018). With respect to studies carried out on women
footballers (Danielsen et al., 2017; Guillen & Santana, 2018;
Wieser & Thiel, 2014), there seems to be a relationship between REFERENCES
MT and performance, although this relationship is influenced by
Alkhasawneh, M. S., Ngah, U. K., Tay, L. T., Isa, M., Ashidi, N. &
the player’s level and age. Again, which criterion is chosen for Al-Batah, M. S. (2014). Modeling and testing landslide hazard using
“performance” might influence that relationship (Cowden, 2017). decision tree. Journal of Applied Mathematics, 2014, 929768, 1–9.
The second variable selected by the multivariate model was https://doi.org/10.1155/2014/929768.


Alvarez, O., Walker, B. & Castillo, I. (2018). Examining motivational
activation in practice, reflecting the relevance of being prepared to
correlates of mental toughness in Spanish athletes. Cuadernos de
face the daily tasks of training. The decision tree results point to
Psicologıa del Deporte, 18, 141–150.
the need to have high levels of MT and activation in practice to Breiman, L., Friedman, J. H., Olshen, R. A. & Stone, C. J. (1984).
reach the national team. In this sense, it seems necessary to train Classification and regression trees. Belmont, TN: Wadsworth.
these constructs. However, although MT is improved by training Bull, S. J., Shambrook, C. J., James, W. & Brooks, J. E. (2005). Towards
(Dohme, Bloom, Piggott & Backhouse, 2019), it must be kept in an understanding of mental toughness in elite English cricketers.
Journal of Applied Sport Psychology, 17, 209–227.
mind that MT has shown a relationship with aspects of
Castagna, C. & Castellini, E. (2013). Vertical jump performance in Italian
personality that are innate, such as perfectionism (Cowden, Crust, male and female national team soccer players. Journal of Strength and
Jackman & Duckett, 2019). Conditioning Research, 27, 1156–1161.
The present work has a series of limitations. The first is that it Coetzee, B., Grobbelaar, H. & Grid, C. C. (2006). Sport psychological
was cross-sectional but used variables which fluctuate over time. skills that distinguish successful from less successful soccer teams.
Journal of Human Movement Studies, 51, 383–401.
With variables of this type, it is recommended that longitudinal
Cowden, R. G. (2017). Mental toughness and success in sport: A review
studies to be made. Hence, the conclusions should be interpreted and prospect. The Open Sports Sciences Journal, 10, 1–14.
with this in mind. Given the scarcity of studies of women’s Cowden, R. G., Crust, L., Jackman, P. & Duckett, T. R. (2019).
football however, this work may be taken as a first approach to Perfectionism and motivation in sport: The mediating role of mental
this topic. The second is that there are external factors such as the toughness. South African Journal of Science, 115, 57–63. https://doi.
org/0.17159/sajs.2019/5271.
type of coach (Das & Chatterjee, 2018), and internal factors such
Danielsen, L., Rodahl, S., Giske, R. & Høigaard, R. (2017). Mental
as the player’s personality type (Kristj
ansd
ottir et al., 2018) that toughness in elite and sub-elite female soccer players. International
might influence the results. The third is that the TOPS Journal of Applied Sports Sciences, 29, 77–85.
questionnaire only purports to measure the amount an individual Das, M. & Chatterjee, M. (2018). Relationship between perception of
uses a psychological skill, not the effectiveness of that use (Neil coach efficacy and competitive state anxiety among players – an
exploration. International Journal of Physical Education and Sports,
et al., 2006) nor whether the players usually train those skills.
3, 12–18.
Dohme, L. C., Bloom, G. A., Piggott, P. & Backhouse, S. (2019).
Development, implementation, and evaluation of an athlete-informed
CONCLUSIONS mental skills training program for elite youth tennis players. Journal of
In the competitive field, performance cannot be explained solely Applied Sport Psychology. https://doi.org/10.1080/10413200.2019.
1573204.
by physical condition or technique. It is also necessary to identify
Gledhill, A., Harwood, C. & Forsdyke, D. (2017). Psychosocial factors
the psychological attributes that affect it (Singh, Thapa, Baro & associated with talent development in football: A systematic review.
Singh, 2016). The differences in PS between national team and Psychology of Sport and Exercise, 31, 93–112.
first division players were few and far between, but, in general, Gould, D. & Maynard, I. (2009). Psychological preparation for the
some differences in MT and anxiety were found between the Olympic Games. Journal of Sports Sciences, 27, 1393–1408.
Gucciardi, D. F., Crane, M., Ntoumanis, N., Parker, S. K., Thøgersen-
three teams. The multivariate classification tree model correctly
Ntoumani, C., Ducker, K. J. et al. (2018). The emergence of team
classified 55% of the sample with the variables total score in resilience: A multilevel conceptual model of facilitating factors.
SMTQ and activation in practice in TOPS. The results seem to Journal of Occupational and Organizational Psychology, 91, 729–
indicate that there is indeed a relationship between MT and the 768.
team’s level. Thus, the ability to concentrate, rebound from Gucciardi, D. F., Hanton, S., Gordon, S., Mallett, C. J. & Temby, P.
(2015). The concept of mental toughness: Tests of dimensionality,
failure, cope with pressure, and face adversity, as well as mental
nomological network and traitness. Journal of Personality, 83, 26–44.
resilience, commitment, and confidence (Bull et al., 2005) Guillen, F. & Santana, J. (2018). Exploring mental toughness in soccer
become determinants of sports performance. In this sense, a players of different levels of performance. Revista Iberoamerica de
recent study goes further and speaks of the emerging concept of Psicologıa del Ejercicio y el Deporte, 13, 297–303.

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd
Scand J Psychol (2019) Psychological characteristics in football players 7

Haneishi, K., Fry, A. C., Moore, C. A., Schilling, B. K., Li, Y. & Fry, M. Meggs, J., Chen, M. A. & Koehn, S. (2019). Relationships between flow,
D. (2007). Cortisol and stress responses during a game and practice in mental toughness, and subjective performance perception in various
female collegiate soccer players. Journal of Strength and Conditioning triathletes. Perceptual and Motor Skills, 12, 241–252.
Research, 21, 583–588. Mellalieu, S. D., Neil, R., Hanton, S. & Fletcher, D. (2009). Competition
Hayslip, B., Petrie, T. A., MacIntire, M. M. & Jones, G. M. (2010). The stress in sport performers: Stressors experienced in the competition
influences of skill level, anxiety, and psychological skills use on environment. Journal of Sports Sciences, 27, 729–744.
amateur golfers’ performances. Journal of Applied Sport Psychology, Musculus, L. & Lobinger, B. H. (2018). Psychological characteristics in
22, 123–133. talented soccer players-Recommendations on how to improve coaches’
H€oner, O. & Feichtinger, P. (2016). Psychological talent predictors in assessment. Frontiers in Psychology, 9, 41. https://doi.org/10.3389/
early adolescence and their empirical relationship with current and fpsyg.2018.00041
future performance in soccer. Psychology of Sport and Exercise, 25, Neil, R., Mellalieu, S. D. & Hanton, S. (2006). Psychological skills usage
17–26. the competitive anxiety response as a function of skill level in rugby
Huijgen, B. C. H., Elferink-Gemser, M. T., Lemmink, K. A. P. M. & union. Journal of Sports Science and Medicine, 5, 415–423.
Visscher, C. (2014). Multidimensional performance characteristics in Nicholls, A. R., Polman, R. C. J., Levy, A. R. & Backhouse, S. H. (2009).
selected and deselected talented soccer players. European Journal of Mental toughness in sport: Achievement level, gender, age,
Sport Science, 14, 2–10. experience, and sport type differences. Personality and Individual
Jacobson, I. & Tegner, Y. (2007). Injuries among Swedish female elite Differences, 47, 73–75.
football players: A prospective population study. Scandinavian Reilly, T., Williams, A. M., Nevill, A. & Franks, A. (2000). A
Journal of Medicine and Science in Sports, 17, 84–91. multidisciplinary approach to talent identification in soccer. Journal of
Jooste, J., Steyn, B. J. M. & Van Den Berg, L. (2014). Psychological Sports Sciences, 18, 695–702.
skills, playing positions and performance of African youth soccer Sheard, M., Golby, J. & van Wersch, A. (2009). Progress toward construct
teams. South African Journal for Research in Sport, Physical validation of the Sports Mental Toughness Questionnaire (SMTQ).
Education and Recreation, 36, 85–100. European Journal of Psychological Assessment, 25, 186–193.
Kang, H. & Jang, S. (2018). Effects of competition anxiety on self- Singh, L. S., Thapa, S. K., Baro, M. & Singh, O. J. (2016). Mental
confidence in soccer players: Modulation effects of home and away toughness between selected outdoor and indoor games athletes.
games. Journal Men's Health, 14, e62–e68. International Journal of Physical Education, Sports and Health, 3,
Knable, M. B., Barci, B. M., Bartko, J. J., Webster, M. J. & Torrey, E. F. 271–273.
(2002). Molecular abnormalities in the major psychiatric illnesses: Slimani, M., Bragazzi, N. L., Tod, D., Dellal, A., Hue, O., Cheour, F.
Classification and regression tree (CRT) analysis of post-mortem et al. (2016). Do cognitive training strategies improve motor and
prefrontal markers. Molecular Psychiatry, 7, 392–404. positive psychological skills development in soccer players? Insights
Kolayisß, H. & C ß elik, N. (2017). Examination of motivation, anxiety and from a systematic review. Journal of Sports Sciences, 34, 2338–2349.
imagery levels of footballers from different leagues. Revista de Smith, R. E., Smoll, F. L., Cumming, S. P. & Grossbard, J. R. (2006).
Psicologıa del Deporte, 26, 23–27. Measurement of multidimensional sport performance anxiety in
Kristj
ansd
ottir, H., Erlingsd
ottir, A. V., Sveinsson, G. & Saavedra, J. M. children and adults: The Sport Anxiety Scale-2. Journal of Sport &
(2018). Psychological skills, mental toughness and anxiety in elite Exercise Psychology, 28, 479–501.
handball players. Personality and Individual Differences, 134, 125– Stølen, T., Chamari, K., Castagna, C. & Wisløff, U. (2005). Physiology of
130. Soccer. Sports Medicine, 35, 501–536.
€
Kurt, C., Catikkas, F., Omurlu, I. K. & Atalag, O. (2012). Comparison of Taylor, M. K., Gould, D. & Rolo, C. (2008). Performance strategies of US
loneliness, trait, anger-anger expression style, self-esteem attributes olympians in practice and competition. High Ability Studies, 19, 19–
with different playing position in soccer. Journal of Physical 36.
Education and Sport, 12, 39–43. Thomas, P. R., Murphy, S. M. & Hardy, L. (1999). Test of performances
Landers, D. M. & Arent, S. M. (2010). Arousal-performance relationships. strategies: Development and preliminary validation of a comprehensive
In J. M. Williams (Ed.), Applied sport psychology; Personal growth to measure of athletes’ psychological skills. Journal of Sports Sciences,
peak performance (6th ed., pp. 221–246). Dubuque, IA: McGraw-Hill. 17, 697–711.
Loehr, J. E. (1986). Mental toughness training for sports: Achieving Wieser, R. & Thiel, H. (2014). A survey of “mental hardiness” and
athletic excellence. Lexington, MA: Stephen Green. “mental toughness” in professional male football players. Chiropractic
Lowther, J., Lane, A. & Lane, H. (2002). Self-efficacy and psychological and Manual Therapies, 22, 17. https://doi.org/10.1186/2045-709X-22-
skills during the amputee soccer World Cup. Athletic Insight. The 17
Online Journal Sport Psychology, 4, 23–33. Zeiger, J. S. & Zeiger, R. S. (2018). Mental toughness latent profiles in
Mace, R. & Carroll, D. (1986). Stress inoculation training to control endurance athletes. PLoS ONE, 13, e0193071. https://doi.org/10.1371/
anxiety in sport: Two case studies in squash. British Journal of Sports journal.pone.0193071.
Medicine, 20, 115–117.
McCarthy, P. J., Jones, M. V., Harwood, C. G. & Olivier, S. (2010). What Received 14 January 2019, accepted 1 July 2019
do young athletes implicitly understand about psychological skills?
Journal of Clinical Sport Psychology, 4, 158–172.

© 2019 Scandinavian Psychological Associations and John Wiley & Sons Ltd