Mental toughness in sport

A literature search was conducted using major computerized databases (e.g. PubMed, ScienceDirect and Scopus) and library holdings for peer-reviewed articles in the English language and were rechecked by another two co-authors. The keywords used in this review were mental toughness, sport and athlete. A manual search of the reference lists in the relevant studies found in the computerized search was also performed.

The inclusion of articles followed a three-phased approach (Fig. 1) using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al., 2015). In the first phase, the 1311 records were initially obtained through extensive database searching. In all, 57 duplicates were identified and removed in this phase. In the second phase, the titles of 1254 records were screened, and records were removed if they did not refer to the following terms (or close variants of such): mental toughness, sport, athlete. This process resulted in the removal of 1093 records. In the third phase, the full-texts of the final 161 records were examined. Records that met the inclusion criteria were studies that: (a) used the mental toughness in the study; (b) were conducted in a sport context; and (c) were written in English. A total of 14 studies met the inclusion criteria, and these were included in the review.

The studies included in this review have been divided into four major broad categories. The first category comprises studies which involved the early conceptualization of mental toughness. The second category consists of studies that utilized qualitative approaches on mental toughness. The third category includes studies that used quantitative approaches. This review considers both qualitative and quantitative approaches to the study of mental toughness with the specific focus on the models and the development of the measurement of this construct (all studies summarized in Table 1). Finally, the forth category reviewed research on the relationship between mental toughness and other psychological variables.

The first academic reference to the concept of mental toughness was proposed by Cattell, Blewett, and Beloff (1955) who suggested ‘tough-mindedness’ as a culturally or environmentally determined personality trait fundamental to individual success. Purported to be one of sixteen primary traits that described personality, Cattell (1957) viewed tough-minded individuals as self-reliant, realistic and responsible, and contrasted this with emotional sensitivity. This position was supported by suggestions that “the athlete who is mentally tough is somewhat insensitive to the feelings and problems of others” (Tutko & Richards, 1971), and that “being able to handle pressure off the field can help you be mentally tough on it” (Tapp, 1991). Whilst others supported the notion that mental toughness was a personality trait (Kroll, 1967; Werner & Gottheil, 1966), others have challenged this (Dennis, 1978), with some purporting that the construct is simply a state of mind (Gibson, 1998) or even just a set of psychological characteristics (Bull, Albinson, & Shambrook, 1996).

Most elite athletes contended that at least 50% of their superior athletic performance was the result of mental or psychological factors that reflected the phenomenon of mental toughness (Loehr, 1982, 1986), whereas 83% of wrestling coaches rated it as the most important psychological characteristics for determining competitive success (Gould et al., 1987). The extensive work of Loehr (1982, 1986) who advocated that mental toughness is an attribute of those who respond to problems, pressure, making mistakes and competition with the right attitude. According to Loehr (1986), mentally tough performers are disciplined thinkers who respond to pressure in ways which enables them to remain feeling relaxed, calm and energized simply because of the ability to sustain positive energy flow despite adversity. In addition, Loehr (1986) published a model of mental toughness that included seven characteristics: self-confidence, negative energy, attention control, visual and imagery control, motivation, positive energy, and attitude control. Although this model is conceptually appealing, Loehr did not provide a rationale for the selection of the seven mental toughness factors. However, Loehr (1995) pointed out that mental toughness can be developed and acquired, and not just innate genetic traits (Gucciardi et al., 2009).

Crust (2007) noted that one point of contention in the literature had been related to whether mental toughness is conceptualized as an inherited, innate personality characteristic or if the acquisition of mental toughness is believed to be due more to environmental factors and learning. In addition, these studies are not based on rigorous theoretical and scientific methods. The limitations of earlier studies reflect the development of mental toughness in the future. Therefore, contemporary research begins to explore the definition and characteristics of mental toughness with qualitative and quantitative research patterns. The next section focuses on the qualitative approaches to the study of mental toughness.

In a pioneering qualitative study of mental toughness, Fourie and Potgieter (2001) analyzsd written responses from 131 expert coaches and 160 elite athletes. Coaches and elite athletes responded to a series of open-ended questions requiring them to provide their best definitions and descriptions of mental toughness. They identified twelve components of mental toughness, including motivation level, coping skills, confidence maintenance, cognitive skills, discipline and goal-directedness, competitiveness, possession of prerequisite physical and mental requirements, team unity, preparation skills, psychological hardiness, and ethics. Drawing from these responses, they recognized the subjective interpretation of the way coaches and athletes described characteristics of mental toughness. For this reason, the researchers concluded that further work was needed to bring about a more objective description and model of mental toughness (Fourie & Potgieter, 2001).

In another qualitative study, Jones et al. (2002) employed Kelly’s (1955) personal construct theory to understand how mental toughness is construed. In brief, this theory focuses on both the uniqueness of the individual and the processes common to all people. Furthermore, personal construct theory proposes that individuals strive to understand, interpret, anticipate, and control the world of experience in order to deal effectively with it (Kelly, 1955). Using data from ten elite international athletes, Jones et al. (2002) proposed that mental toughness is having the natural or developed psychological edge that enables athletes to (1) generally cope better than the opponents with the many sports demands (competition, training, lifestyle); and (2) specifically more consistent and better than the opponents in staying determined, focused, confident, and in control under pressure.

As a result of an inductive thematic content analysis, Jones et al. (2002) identified twelve key attributes of mental toughness including having an unshakable self-belief in one’ ability to achieve your competition goals; bouncing back from performance setbacks as a result of increased determination to succeed; having an unshakable self-belief that one’s possess unique qualities and abilities; having an insatiable desire and internalized motives to succeed; remaining fully-focused on the task at hand in the face of competition-specific distractions; regaining psychological control following competition-specific unexpected and uncontrollable events; pushing back the boundaries of physical and emotional pain, while still maintaining technique and effort under distress (in training and competition); accepting that competition anxiety is inevitable and knowing that one can cope with it; thriving on the pressure of competition; not being adversely affected by others’ good and bad performances; remaining fully-focused in the face of personal life distractions; and switching a sport focus on and off as required. These related to performance and lifestyle-related focus, self-belief, desire and motivation and how a mentally tough performer deals with the pressure (external), anxiety (internal) and the hardship associated with superior performance (i.e., physical and emotional pain).

A limitation in early mental toughness research is its general conceptualization of the construct of mental toughness. One of the studies that attempt to address this is the study of mental toughness specific to cricket players (Bull et al., 2005). The two main issues Bull et al. (2005) addressed in their study were the following: to obtain a better understanding of what mental toughness is for cricketers and to identify how cricketers developed their mental toughness. They used qualitative procedures to interview 12 mentally tough cricketers. The participants were drawn from a total of 101 English cricketers whom were identified by coaches as being the most mentally tough cricketers of the previous 20 years in English cricket. From their results, they presented a model of mental toughness that included four structural categories, each containing a number of themes related to overall mental toughness. These include environmental influence: parents, childhood, the need to earn success, opportunities to survive early setbacks, exposure to foreign cricket; tough character: resilient confidence, independence, self-reflection, competitiveness with self as well as others; tough attitudes: never-say-die mindset, go-the-extra-mile mindset, thrive on competition, belief in making a difference, exploit learning opportunities, willing to take risks, belief in quality preparation, determination to make the most of ability, self-set challenging targets; and tough thinking: thinking clearly–making good decisions, keeping perspective, honest self-appraisal and robust self-confidence–overcoming self-doubts, feeding off physical conditioning, maintaining self-focus (Bull et al., 2005).

In another study involving male professional soccer players, Thelwell et al. (2005) asked the players to compare their soccer-specific definition and their understanding of mental toughness with that proposed by Jones et al. (2007). The soccer players viewed mental toughness as enabling players to “always” cope better than their opponents rather than “generally” cope better. Likewise, they identified only ten attributes as opposed to the twelve attributes of mental toughness by Jones et al. (2007). Those attributes included the following: having total self-belief at all times that one will achieve success; wanting the ball/wanting to be involved at all times; having the ability to react to situations positively; having the ability to hang on and be calm under pressure; knowing what it takes to grind oneself out of trouble; having the ability to ignore distractions and remain focused; controlling emotions throughout performance; having a presence that affects opponents; having everything outside of the game in control, and enjoying the pressure associated with performance (Thelwell et al., 2005).

In 2007, Jones et al. conducted a follow-up study using a sample of super-elite sports performers (i.e., Olympic/World Champions) to expand the mental toughness knowledge base and broadened the scope by including the perceptions of coaches and sport psychologists who had coached and consulted at that level. The results mirrored their earlier definition of mental toughness. Moreover, they also extended the list of attributes considered essential to the make-up of mental toughness to 30. These were subsequently categorized into 13 subcomponents of mental toughness, which were then organized into a framework of mental toughness comprising four dimensions; a general Attitude/mindset dimension, and three time-specific dimensions, Training, Competition, and Post-competition (Jones et al., 2007).

In an attempt to propose a more specific definition of mental toughness, Gucciardi et al. (2009) employed Kelly’s (1955, 1991) Personal Construct Psychology (PCP) and proposed mental toughness as “a collection of experientially developed and inherent sport-specific and sport-general values, attitudes, behaviours, and emotions that influence the way in which an individual approaches, responds to, and appraises both negatively and positively construed pressures, challenges and adversities to consistently achieve his or her goals” (p. 278). Whilst Gucciardi et al. (2008) did not offer a definitive perspective on the key values, attitudes, cognitions and emotions, investigations into the sport-specific components of mental toughness related to Australian rules football (Gucciardi et al., 2008), cricket (Gucciardi & Gordon, 2009) and soccer (Coulter, Mallett, & Gucciardi, 2010), highlighting the emergence of a core group of key mental toughness facets that do not vary significantly by sport (e.g., self-belief, self-motive, attention control, resilience).

One of the key differences between a qualitative and a quantitative approach is the used of instruments to quantify the data. In this regard, following his conceptualization of mental toughness, Loehr (1986) constructed the Psychological Performance Inventory (PPI). Loehr (1982) suggested that mentally tough athletes learned or developed two important skills: first, the ability to increase their flow of positive energy when faced with adversity or a crisis; and second, to think in ways that promote the right attitudes to solve problems, or to deal with pressure, mistakes, or competition. The PPI contains 42 items and measures mental toughness which is conceptualized to have seven dimensions; Self-confidence, Negative Energy, Attention Control, Visualization and Imagery Control, Motivation, Positive Energy, and Attitude Control. Each subscale contains six items, each scored on a 5-point Likert scale, with scores for each subscale ranging from 6 to 30, and for total mental toughness ranging from 42 to 210. A number of studies (e.g., Golby, Sheard, & Lavallee, 2003, 2004; Kuan & Roy, 2007; Lee, Shin, Han, & Lee, 1994) have employed the PPI as a measure of mental toughness.

Despite its widespread usage, evidence has emerged suggesting limitations of its psychometric properties. For example, Golby et al. (2007) and Middleton et al. (2004) tested the construct validity of the PPI. For example, using responses from 263 young elite athletes, Middleton et al. (2004) revealed inadequate fit between the hypothesized seven-factor model and the data as well as an improper solution (i.e. factor correlations >1) using confirmatory analysis (CFA). However, some criticism has been levelled at Middleton et al. (2004) study for using a somewhat small sample size (N = 263) for testing construct validity and the limited age range of the sample (12–17 years old). Addressing these limitations, Golby et al. (2007) used a larger sample (N = 408) with a broader age range (12–63 years old) to evaluate the psychometric properties of the PPI, and they reported a similar lack of support for the factorial structure of the PPI. Besides, Gucciardi (2011) also concluded that the psychometric evidence for the hypothesized measurement model of the PPI is not encouraging for its future use.

With the analyses revealing a lack of support for the hypothesized factor structure of the original PPI (1986), Golby et al. (2007) subsequently developed the Psychological Performance Inventory-Alternative (PPI-A), which represents four factors of MT, namely determination, self-belief, positive cognition, and visualization. Golby et al. (2007) used the responses from their original PPI study to generate the PPI-A involving 408 respondents. Using principal component analysis to find structure in their data, they used confirmatory factor analysis to assess the psychometric structure of the model. Collectively satisfying absolute and incremental fit index benchmarks, the inventory possesses satisfactory psychometric properties, with adequate reliability and convergent and discriminant validity. The results lend preliminary support to the factorial validity and reliability of the model.

Sheard (2009) used the PPI-A to investigate national differences in mental toughness between rugby league players in the United Kingdom and Australia. The results from this study indicated that significant differences in mental toughness were apparent between national teams. Although these findings are based on small sample size, Sheard (2009) concluded that these findings provided evidence for the divergent (or discriminant) validity (i.e. does not correlate too much with similar but distinct constructs) of the PPI-A. As alternatives to PPI and PPI-A, Clough et al. (2002) developed the Mental Toughness Questionnaire 48 (MTQ 48) consistent with their model of mental toughness. Reflecting the name, the MTQ 48 contains 48 items that are scored on a 5-point Likert scale ranging from (1) strongly agree to (5) strongly disagree (Crust & Clough, 2005). It has an overall test–retest coefficient of 0.9, with the internal consistency of the subscales (Control, Commitment, Challenge and Confidence) found to be 0.73, 0.71, 0.71, and 0.8 respectively (Clough et al., 2002). Connaughton et al. (2008), however, advised that the MTQ 48 psychometric properties reported should be interpreted with caution because the rationale for the conceptualization of MT is essentially based on hardiness and confidence constructs. Furthermore, there was no demonstration of validity even if a sound conceptualization was apparent.

Middleton et al. (2005) constructed the Mental Toughness Inventory (MTI) 36-item based on a model of mental toughness developed from themes that emerged from their qualitative study. The MTI was designed to measure the twelve characteristics Middleton et al. (2004) proposed, namely: self-efficacy, future potential, mental self-concept, task familiarity, value, personal best motivation, goal commitment, task-specific attention, perseverance, positivity, positive comparisons, and stress minimization, which are summed to produce a global mental toughness score. The self-report MTI is an 8‑point response scale where respondents rate statements from 1 (not like me) to 8 (like me). Middleton, Martin, and Marsh (2011) reported that the MTI is strong on conceptual grounds with strong psychometric properties and high reliability. However, Sheard (2012) suggested that although the development of MTI is based on a sound theoretical framework, the MTI required independent testing to assess its psychometric properties.

Sheard et al. (2009) then developed the Sport Mental Toughness Questionnaire (SMTQ). Two independent studies supported a three-factor (Confidence, Constancy, and Control) 14-item model for the SMTQ. With a sample of 633 athletes (427 males, 206 females; mean [M] age = 21.5 years; standard deviation [SD] = 5.48), drawn from 25 sports, of international, national, county and provincial, or club and regional standards, the first study focused on item development and exploratory factor analytic techniques to establish the psychometric properties of the SMTQ. The second study sought to confirm the structure obtained in the first study using confirmatory factor analytic techniques with an independent sample of 509 sports performers (351 males, 158 females; M age = 20.2 years; SD = 3.35), competing at the aforementioned standards, and representative of 26 sports. Confirmatory analysis using structural equation modelling confirmed the overall structure. A single factor underlying mental toughness was identified with higher-order exploratory factor analysis using the Schmid-Leiman procedure. Collectively, satisfying absolute and incremental fit-index benchmarks, the inventory was shown to possess satisfactory psychometric properties, with adequate reliability, divergent validity, and discriminative power. The results revealed promising features of the SMTQ, lending preliminary support to the instrument’s factorial validity and reliability. The results revealed promising features of the SMTQ, lending preliminary support to the instrument’s factorial validity and reliability.

Another mental toughness questionnaire is based on the work of Gucciardi et al. (2009). They develop a pool of 60 items for their 11-factor model of mental toughness. Confirmatory factor analysis with 418 footballers revealed that the 11-factor model did not fit the data well. The authors then performed a series of principal component analysis (PCA) using both varimax and promax rotations to explore the usefulness of three-, four-, and five-factor solutions. These analyses led to a 24-item four-factor model (thrive through challenge, sport awareness, tough attitude, desire success), which they labelled the Australian Football Mental Toughness Inventory (AfMTI). However, Gucciardi et al. (2009) found that the 24-item AfMTI did not fit the data with a sample of 350 youth Australian footballers. Gucciardi (2011) concluded that further work is required to address these concerns.

Recognizing the need to capture the context-specific dimension of mental toughness, Gucciardi et al. (2009) conducted a series of studies within a mixed-methods framework to develop their Cricket Mental Toughness Inventory (CMTI). Interviews with 16 cricketers from two countries, five of whom were still involved in international cricket at the time of study to develop another model of mental toughness for cricketers. A six-factor model emerged from the interviews, and two independent focus groups with nine Australian cricketers resulted in minor modifications to the wording and the addition of eight items. Gucciardi et al. (2009) administered the six-factor inventory to a general sample of cricketers from international leagues (n = 570) and a sample of Australian cricketers (n = 433). There was an equal split of the total sample for either a calibration or cross-validation analysis. One factor (i.e., cricket smarts) was removed following a series of confirmatory factor analyses with the calibration sample, which provided support for the 15-item, five-factor model. Gucciardi (2011) found support for the validity of the CMTI measurement model by confirmatory factor analysis and internal reliability analysis. Gucciardi (2011) and Gucciardi et al. (2009) provided preliminary support for the factor structure, internal reliability, and construct validity of the CMTI. Gucciardi (2011) stated that the use of a male sample created some practical issues in the generalizability of the measurement tool across gender and suggested that further conceptual and statistical work would be beneficial before using the CMTI in mental toughness research.

Despite the emergence of alternatives to measure general and sport-specific mental toughness, the existing questionnaire are showing mixed results when tested in a different sample from which they are originally constructed. These mixed findings warrant further investigation into how mental toughness is to be measured.

The studies included in this review covered a period of 41 years (1978–2019). The studies involved participants from a range of competitive levels, such as school athletes, college athletes, professional athletes, and the general population. In terms of countries distribution, the vast majority of studies were from the European setting. As for sport distribution, there was a varied representation of sports, namely swimming, sprinting, artistic and rhythmic gymnastics, trampolining, middle-distance running, triathlon, golf, netball, boxing, athletics, judo, triathlon, rowing, pentathlon, squash, cricket, soccer, rugby, roller skating, basketball, canoeing, golf, soccer and Australian football. Seven studies involved multiple sports and five studies involved specific sports.

The study of mental toughness has advanced since the adoption of more scientifically rigorous approaches, but there are still a number of limitations and theoretical description that should be considered when interpreting their findings. Although both qualitative (e.g., Bull et al., 2005; Fourie & Potgieter, 2001; Gucciardi et al., 2008; Jones et al., 2002; Middleton et al., 2004; Thelwell et al., 2005) and quantitative approaches (e.g., Clough et al., 2002; 2005; Golby et al., 2007; Gucciardi & Gordon, 2009; Gucciardi et al., 2009; Loehr, 1986; Middleton, 2007; Sheard et al., 2009) have been used to understand mental toughness, there are also differences and agreement in regard to conceptual issues and measurement. Mental toughness appears to be multidimensional and most often associated with unshakeable self-belief, the ability to rebound after failures (resilience), persistence or refusal to quit, coping effectively with adversity and pressure, and retaining concentration in the face of many potential distractions. Most contemporary researchers suggest that an individual’s mental toughness will be determined by both inherited characteristics and by learning, experience, and environments influences (Bull et al., 2005; Gordon & Sridhar, 2005; Jones et al., 2002; Thelwell et al., 2005). Research into the relationship between mental toughness and performance has consistently shown that better performances of both cognitive and motor skills are associated with higher levels of mental toughness (Clough et al., 2002; Crust & Clough, 2005) and those elite athletes have higher mental toughness than lower level performers (Golby et al., 2003; Thomas, Schlinker, & Over, 1996). One of the key advances toward a greater understanding of mental toughness appears to be the development of valid and reliable measurement instruments. Past studies used other psychological variables such as emotions, affections, perceptions or pains as a potential mechanism for psychological mental toughness in competitive sports situations. Therefore, this issue should be considered as the future direction of study because there is still room for further development of potential mechanisms for confirmation.