The consensus that mental toughness (MT), as a psychological construct, is an indispensible factor for athletic achievement, success, and excellence (Clough, Earle, & Sewell, 2002; Crust, 2007; Jones, Hanton, & Connaughton, 2007; Loehr, 1986) has increased interest towards not only developing a comprehensive understanding of the construct, but developing appropriate instruments to assess MT. Accurate MT appraisals have prospective implications for examining areas warranting improvement as well as assessments of MT following intervention efforts, particularly due to the developmental trajectory of MT over time and through experiences (Connaughton, Hanton, & Jones, 2010) as well as through direct psychological training and intervention efforts (Bell, Hardy, & Beattie, 2013).

Although several general (e.g., Mental Toughness Questionnaire 48 [MTQ48]; Clough et al., 2002) and sport-specific MT inventories (e.g., Cricket Mental Toughness Inventory [CMTI]; Gucciardi & Gordon, 2009) have been developed, some (e.g., Psychological Performance Inventory [PPI]; Loehr, 1986) have demonstrated inadequate psychometric properties and have been devoid of a theoretical basis for item inclusion (Middleton et al., 2004). In addition, there are suggestions that MT develops differently according to sport-type involvement (Gucciardi, 2009) and has distinct requirements based on particular types of sports or selected periods of sport participation (Bull, Shambrook, James, & Brooks, 2005). This has resulted in the development of sport-specific MT inventories to adequately appraise MT according to sport-type MT manifestations and demands.

One sport that has received insufficient attention and for which a sport-specific MT instrument has yet to be developed is tennis. Considering the conditions, nature of participation, and demands associated with tennis differ markedly from other sports, it is important that an instrument be examined and validated for use with tennis athletes. In this study, the decision to select an inventory that has been afforded prior psychometric support is based on two reasons. The first is that instruments that have been validated using athletes from a variety of sporting backgrounds may not exhibit the same factor structure or item loadings among a group of athletes from one type of sport, especially if MT differs according to sport type. The second is that although the instrument that has been selected for use in this study (i.e., Sport Mental Toughness Questionnaire [SMTQ]; Sheard et al., 2009) is a multidimensional measure of MT and has undergone extensive psychometric validation and support, subsequent examinations have resulted in the suggestion of further psychometric evaluation (Crust & Swann, 2011). Therefore, accurate sport-specific appraisals of athlete MT require the assessment and validation of individual instruments for appropriate adaptation and use in such contexts. Hence, the purpose of the present study is to examine the psychometric properties of a previously validated MT instrument for prospective adaptation and use amid competitive tennis players.

Method

Participants

A total of 365 competitive tennis athletes participated in the study. The participant demographics are outlined in Table 1. The participants were required to have engaged in some form of tennis competition within the last two weeks and must have engaged in tennis participation for a minimum of five years prior to their participation in the study. These two criteria were considered necessary in order to operationally define the term competitive, which would more likely enable the evaluation of a specific sub-group of tennis participants and maintain congruence among all participants. Additionally, the criteria were considered important for reducing the potential for confounds to influence the results and subsequent interpretation of the findings. For instance, if the requirement to include athletes currently competing was omitted, the degree to which participant scores on the various measures were influenced by the act of engaging in competition or absence of competitive engagement could not be determined.

Participants were recruited through local and national tennis organisations, universities, high schools, professional coaches, tournaments, and county clubs. University ethical approval from the University of Kwazulu-Natal's Institutional Review Board was obtained before conducting the study. All participants were provided and completed an informed consent document prior to their participation.

Materials

The Sports Mental Toughness Questionnaire (SMTQ; Sheard et al., 2009) was used to assess participants' MT. The SMTQ is a multidimensional measurement of MT that is based on the qualitative constituents of MT consistently identified in the MT literature (Sheard, 2013). The primary reason for selecting the SMTQ was based on the demonstrated construct validity of the inventory resulting from exploratory and confirmatory factor analyses, adequate internal reliability, divergent validity, and discriminative ability findings. In fact, Sheard (2013) asserts that the SMTQ is only one of two MT instruments (the other being the PPI-A) that have been extensively evaluated using EFA and CFA and acquired support from the analyses.

The SMTQ is comprised of three subscales (confidence, constancy, and control) that combine for a general measure of MT. Sheard et al. (2009), in their CFA, evidenced strong support for the hierarchical three-factor model, with a goodness-of-fit index (GFI) of .95 suggesting good model fit. The coefficients between the higher-order factor of total MT and second-order factors of confidence (r = .72), constancy (r = .71), and control (r = .66) were considered acceptable. Correlations between confidence and control, confidence and constancy, and constancy and control were reportedly .28, .31, and .31, respectively, all of which were statistically significant (Sheard et al., 2009). Internal consistency for global MT on the SMTQ was strong ([alpha] = .81). Additionally, internal consistency for the subfactors was considered moderate to strong (confidence = .79, constancy = .76, control = .72).

Providing evidence for the divergent validity of the measure, correlations between the SMTQ and the subscales on the Life Orientation Test, Personal View Survey II1-R, and the Positive and Negative Affect Schedule were moderate and ranged from .23-.38,. 14-.33, and .12-.49, respectively (Sheard et al., 2009). The researchers also reported the discriminative power of the SMTQ based on statistically meaningful differences between athletes of dissimilar competitive levels, age, and gender, with higher scores for more advanced competitive levels, for older athletes, and males.

The SMTQ contains 14-items rated on a Likert-type scale, anchored between 1 (not at all true) and 4 (very true). Sample items on the three subfactors include "I interpret potential threats as positive opportunities" (confidence), "I give up in difficult situations" (constancy), and "I am overcome by self-doubt" (control).

Procedure

Permission letters were obtained from various tennis organisations to access the database of tennis athletes or the contact information of the athletes affiliated with such organisations. The informed consent document and questionnaire were placed into an online survey format for completion and relevant local, provincial, and national organisations were provided the survey link and requested to forward the link to appropriate competitive tennis players attached to such organisations.

The online survey link initially delineated informed consent elements and requested the participants' agreement to the informed consent elements prior to proceeding with the questionnaire items. The participants were provided details of the study along with ethical obligations and procedures during the informed consent process, and, once informed consent agreement was attained, the participants' proceeded through the questionnaire items that mirrored the hard copy completion process.

Appropriate tennis tournaments were also identified and approached in order to obtain additional participants. Permission was obtained to attend the tournaments to engage the tennis participants and obtain the required data, with the questionnaires administered to the players in groups of convenient sizes based on their availability (approximately 5 to 10 players at a time). Informed consent was discussed and obtained from the participants, following which the self-administered questionnaire was distributed to the participants. This questionnaire completion process required approximately 15 to 20 minutes from the participants.

Results

Principal component analysis (PCA) was selected to evaluate the factor structure of the SMTQ. The scree plot indicated the presence of three components prior to the flattening of the line across the points. These three points also evidenced eigenvalues of greater than one. As a result, the number of factors that were specified for extraction in the subsequent factor structure analysis was fixed to three.

PCA with a factor extraction limit of three (using a Direct Oblimin rotation) was conducted with the original 14 questionnaire items. The pattern matrix was examined to determine items that did not possess clear or adequate loadings onto specific factors. One of the more common approaches to assessing the appropriateness of item-factor loadings is to apply loading cut-off criteria to the rotated factor solution results. That is, in order for an item to be retained in the factor structure, the highest item-factor loading should fall above the established criterion. According to Matsunaga (2010), a loading criterion of an absolute value of .4 is a liberal yet appropriate cut-off value, which was selected for application in this study, primarily because the questionnaire had undergone prior validation and is an established instrument. The pattern matrix with item loadings is displayed in Table 2. Although a number of items evidenced factor loadings close to an absolute value of .4, the application of this cutolf criterion resulted in the retention of each item. The retention of all the SMTQ items was followed by further examination of the PCA results.

The Barlett's Test of Sphericity was statistically significant, [[chi square].sub.(91)] = 1010.796, p < .001. This rejection of the null hypothesis suggests the presence of correlations based on the items included in the analysis, and, as a result, denotes the appropriateness of the data for proceeding with PCA. In addition, the Kaiser-Meyer-Olkin (KMO) value was calculated to evaluate the sampling adequacy of the dataset. In this instance, KMO was .796, which, according to the general rules outlined by Hutcheson and Sofroniou (1999), is within an acceptable range for proceeding with PCA.

The total variance explained by Factor 1, Factor 2, and Factor 3 were 26.05%, 13.21%, and 8.21%, respectively, with a total variance of 47.48% explained by the three factors. Examining the pattern matrix that converged with 18 iterations, six items loaded on Factor 1, six items loaded onto Factor 2, and two items loaded onto Factor 3 (see Table 2). The PCA item-factor loadings for each scale and subscale were used to generate the revised measures, with the items that loaded onto each factor summated for subscale totals. The three subfactors were labelled confidence/self-efficacy, emotional/cognitive control, and positive perspective, respectively. Cronbach's alpha for total MT was .771 (all 14 items), with internal consistency estimates (and mean inter-item correlations) for confidence/self-efficacy (6 items), emotional/ cognitive control (6 items), and positive perspective (2 items) .720 (.299), .739 (.324), and .420 (.266), respectively.

Pearson correlation results between the total MT scale subscales are outlined in Table 3. In particular, total MT exhibited moderate to strong positive and statistically significant correlations (r = .41 to .82) with each of the revised MT subscales. In addition, there were moderate positive and statistically correlations (r = .19 to .34) between each of the MT subscales.

Discussion

The purpose of the present study was to examine the factor structure and psychometric properties of a previously validated MT instrument for adaptation and use amongst competitive tennis players in South Africa. Although the SMTQ has been developed and validated previously (Sheard et al., 2009), recent studies involving the SMTQ have recommended further validation efforts towards the refinement of the instrument (e.g., Crust & Swann, 2011). In this study, the SMTQ PCA analysis results confirmed the original three factor structure of the instrument (Sheard et al., 2009). However, the item-factor loadings in this study indicated that the items included in each of the subfactors diverge to some extent from the original SMTQ. Specifically, the results revealed that 6 items loaded onto confidence/self-efficacy, 6 items loaded onto emotional/cognitive control, and 2 items onto positive perspective. This contrasts the original SMTQ item-factor loadings of six, four, and four items on the confidence, constancy, and control subscales, respectively.

Compared to the original SMTQ items included on each factor, the present results differ in several ways. Two of the items that previously loaded onto the confidence subscale ("I interpret potential threats as positive opportunities" and "I have an unshakeable confidence in my ability") loaded onto positive perspective, two of the items that previously loaded onto the constancy subscale ("I am committed to completing the tasks I have to do" and "I take responsibility for setting myself challenging targets") loaded onto confidence/self-efficacy, and two of the items that previously loaded onto the constancy subscale ("I give up in difficult situations" and "I get distracted easily and lose my concentration") loaded onto emotional/ cognitive control.

The SMTQ was previously validated using samples of competitive athletes from a range of sports (26 different sports), whereas this study solely included competitive tennis players. The divergence between the present and past findings may also be influenced, in part, by the geographic and cultural context in this study, as the original factor structure investigation did not involve South African tennis players. Prior studies have evidenced discrepancies between athletes' perceptions of item relevance and importance in the context of their particular sport. For instance, elite tennis athletes reported "not giving up in difficult situations" as the most important MT facet (Cowden, 2012), whereas "a winning mentality and desire" was considered the most important component in soccer (Coulter, Mallett, & Gucciardi, 2010). The context, participation factors, and demands associated with particular sports may emphasise certain MT components as opposed to others, which may explain perceptual differences and the item-factor loading differences in this study.

The multicultural, multiracial, and broad range of tennis players included in the current study may also provide some explanation for the results obtained. It is recognised that certain psychological constructs, such as intelligence, are conceptualised differently between African and Western societies (Serpell & Haynes, 2001). Perhaps these distinctions also apply to MT, with the possibility that certain aspects of MT, as it applies to South Africa, are not included on the SMTQ, and, as a result, were not examined in this study. Prior studies have suggested cultural differences in MT among athletes involved in the same sports, such as Bull et al.'s (2005) report of a number of components associated with MT in English cricketers (e.g., willingness to take risks) that were not outlined in Gucciardi and Gordon's (2009) study involving Australian cricketers. Although MT studies in tennis are limited, it is conceivable to suggest that tennis culture factors, along with social influences, have an impact on the MT manifested within South African tennis players.

Another reason for the current results may be based on the factor structure evaluation method that was selected. That is, Sheard et al. (2009) used CFA to evaluate the EFA identified factor structure of the SMTQ, whereas PCA was used in this study. In addition, Crust and Swann (2011) suggested that the some of the SMTQ subscale internal consistency scores denoted that certain subscales were assessing more than one construct of MT. With the potential influence of cultural perceptions and interpretations in the current sample, the apparent heterogeneity in the item content of selected SMTQ subscales may signify that, at least in the sample of tennis players in this study, some of the SMTQ items may more appropriately load onto the factors delineated.

In addition, the discrepancy between the original and current findings may relate to the common underlying factor of MT and overall similarities among the items, denoting that the items are all measuring a similar underlying construct. This is supported by the strong subscale-global factor correlations in the present study (see Table 3). These findings compare favourably with Sheard et al.'s (2009) report of strong subscale-global factor correlations.

Although certain items loaded differently in this study as compared to the original SMTQ subscales, a number of the items were retained on their original scales. Specifically, four items from the original confidence and control subscales, respectively, loaded significantly onto the same factors in this study, possibly denoting that particular aspects of MT are central to the construct, regardless of culture or sport type. Indeed, Gucciardi, Gordon, and Dimmock's (2009a) definition of MT connotes the relevance of inherent facets as well as the experientially developmental nature of MT, with a core MT component/s that is relatively stable throughout situations and sport involvement. Perhaps the core MT component/s extends cultures, genders, and sport type involvement, with non-core MT components more likely to fluctuate based on background, experiences, and sporting requirements. This perspective coincides with Horsburgh, Schermer, Veselka, and Vernon's (2009) report that MT is dependent on both biological influences and environmental experiences, with the supposition that some aspects of MT may be more amenable to adaptation than others. Determining which aspects of MT are more heritable or adaptable than others, however, is an area that needs to be explored.

As denoted, the subscale terms exhibit conceptual overlap with the original SMTQ subscales, except for the subscale of constancy, which was not adjudicated as an appropriate label for any of the factors revealed in this study. In a prior study, Crust and Swann (2011) suggested that the constancy subscale items appear to assess more than one construct (e.g., personal responsibility and concentration), which was posited as a potential reason for the lower internal consistency estimate for the subscale that was found in their study. This may afford some explanation for some of the items included on the original constancy subscale demonstrating superior loadings with other factors.

Although the third factor, positive perspective, has limited similarity with the original SMTQ subscales, it does possess some degree of congruence with Clough et al.'s (2002) MTQ48 instrument, in which a subscale--challenge--is included as a component of MT. Based on the findings in this study, positive perspective may be likened to the MTQ48 challenge subscale because of common thread of possessing an optimistic outlook and interpretation of the self and events or situations. These similarities may suggest that, despite the findings of different SMTQ factor structure contents in this study, the components the SMTQ appears to be measuring overlap with other measures as well as the components of MT identified in the literature.

Contextualising the positive perspective component amid the other MT components outlined in the study, it would appear that this optimistic and hopeful attitude has important implications for high self-efficacious thoughts and behaviours, such as self-perceived confidence, the ability to control thoughts, emotions, and behaviours, to overcome adversity, and to perceive difficulties as challenges rather than threats. Therefore, it would appear that possessing a positive perspective is an indispensible characteristic for being mentally tough. Providing support for this, Nicholls, Polman, Levy, and Backhouse (2008) evidenced greater levels of optimism and lower levels of pessimism at higher levels of athlete MT. Due to the capacity for optimism to be augmented (Seligman, 1990), regardless of whether optimism is a component of MT or simply a characteristic mentally tough individuals are more likely to possess, interventions aimed at enhancing optimism amongst athletes may provide benefits for improving the MT of athletes (Nicholls et al., 2008).

The internal consistency for the total SMTQ scale computed in this study ([alpha] = .77) compares favourably with previous research ([alpha] = .70 to .81--Crust & Swann, 2011; Sheard et al., 2009). The internal consistency the confidence/self-efficacy, emotional/cognitive control, and positive perspective subscales were .72, .74, and .42, respectively. Although confidence/ self-efficacy and emotional/cognitive control exhibited acceptable internal consistency, Cronbach's alpha for positive perspective was low. With indications that, for scales with low item quantities, mean inter-item correlations be reported and fall between .2 and .4 for reliability to be considered acceptable (Briggs & Cheek, 1986), the positive perspective subscale mean inter-item correlation was .266, which resulted in the decision to retain the subscale.

Based on the divergence between the present results and the original SMTQ findings reported by Sheard et al. (2009), coupled with the lower internal consistency for the positive perspective subscale, further SMTQ validation investigations are warranted to provide additional evidence of the factor structure, item loadings, and support for the use of the SMTQ amongst athletes in general and sport-specific domains. Perhaps, Bull et al.'s (2005) supposition that different sports require or emphasise different types of MT may have particular relevance to the distinctions between the original SMTQ factor loadings and those found in the present study. Indeed, researchers have begun to examine and develop sport-specific MT instruments (e.g., Australian rules Football [AfMTI]; Gucciardi, Gordon, & Dimmock, 2009b), positing that MT may be best understood and measured within particular sporting domains as opposed to more generally. The AfMTI assesses four components of MT, including tough attitude, desire success, sport awareness, and thrive through challenge. The CMTI, on the other hand, assesses affective intelligence, resilience, attentional control, desire to achieve, and self-belief.

Though there is apparent distinction between the facets of MT examined through each of the sport-specific instruments, there are similarities between them, such as the need or desire for success. In addition, there are some commonalities between these instruments and more general measures of MT, such as control (subscales on the MTQ48 and SMTQ) and self-belief (similar to confidence on the MTQ48 and SMTQ). The factor structure outlined in this study demonstrated analogies between the composition of MT in competitive tennis players and the sport-specific MT measures that have been developed. For instance, the confidence/self-efficacy and emotional/cognitive control subscales are similar to the self-belief and attentional control subscales included in the CMTI, whereas the positive perspective subscale may be likened to the thrive through challenge subscale included on the AfMTI. This may further support the notion of a core set of MT components across all spheres and subtle differences based on sport-specific requirements or experiences. As such, it may be important to determine the applicability of the SMTQ as a measure of MT within specific or general athletic populations, or whether instruments should be developed for particular use with a single sporting group. This is one area of the MT literature that requires attention and clarification. Nonetheless, regardless of the proposed direction for measuring MT, the current findings suggest that the SMTQ is an appropriate measure of MT, but that greater caution should be used when interpreting results based on the subfactors as compared to global MT scores.

Limitations and Future Research

Though extending current MT instrument validation efforts, there are several limitations associated with the present study. In particular, the sample included South African tennis players, making it challenging to generalise the findings to tennis players from other countries. In addition, despite including tennis players across a range of competitive standards, ethnic groups, and age ranges, the sample included considerably greater quantities of University Team or League players, White tennis players, and younger tennis players. It is indeterminable as to whether these demographics are representative of the target population in general, and these factors should be considered when using the revised SMTQ outlined in this study.

Notwithstanding the selected limitations associated with the present study, the findings provide preliminary psychometric evidence for the applicability and use of the SMTQ amongst competitive tennis players. The findings further the suggestions that MT applications and requirements may differ according to sport type (e.g., Bull et al., 2005), with supplementary efforts encouraged to evaluate the relevance and appropriate use of the SMTQ and other MT instruments amid athletes from a variety of sporting domains. It is recommended that researchers examine the nature and manifestation of MT based on sport type and develop sport-specific MT instruments.

References

Bell, J., Hardy, L., & Beattie, S. (2013). Enhancing mental toughness and performance under pressure in elite young cricketers: A 2-year longitudinal intervention. Sport, Exercise, and Performance Psychology, 2(4), 281-297. doi: 10.1037/a0033129doi: 10.1037/a0033129

Briggs, S., & Cheek, J. (1986). The role of factor analysis in the evaluation of personality scales .Journal of Personality, 54, 106-148. doi: 10.1111/j. 1467-6494.1986.tb00391.x

Bull, S., Shambrook, C., James, W., & Brooks, J. (2005). Towards an understanding of mental toughness in elite English cricketers. Journal of Applied Sport Psychology, 17, 209-227. doi: 10.1080/10413200591010085

Clough, R, Earle, K., & Sewell, D. (2002). Mental toughness: The concept and its measurement. In I. Cockerill (Ed.), Solutions in sport psychology (pp. 32-45). London, England: Thomson.

Connaughton, D., Hanton, S., & Jones, G. (2010). The development and maintenance of mental toughness in the world's best performers. The Sport Psychologist, 24, 168-193.

Coulter, T., Mallett, C., & Gucciardi, D. (2010). Understanding mental toughness in Australian soccer: Perceptions of players, parents, and coaches. Journal of Sports Sciences, 28, 699-716. doi:10.1080/02640411003734085

Cowden, R. (2012). Differences between college athletes and their coaches' appraisal of the athletes' degree of mental toughness in competitive tennis (Unpublished Master's thesis). Middle Tennessee State University, Murfreesboro.

Crust, L. (2007). Mental toughness in sport: A review. International Journal of Sport and Exercise Psychology, 5, 270-290. doi:10.1080/1612197X.2007.9671836

Crust, L., & Swann, C. (2011). Comparing two measures of mental toughness. Personality and Individual Differences, 50, 217-221. doi:10.1016/j.paid.2010.09.032

Gucciardi, D. (2009). Do developmental differences in mental toughness exist between specialized and invested Australian footballers? Personality and Individual Differences, 47, 985-989. doi: 10.1016/j.paid.2009.08.001

Gucciardi, D., & Gordon, S. (2009). Development and preliminary validation of the Cricket Mental Toughness Inventory (CMTI). Journal of Sports Sciences, 27, 1293-1310. doi: 10.1080/02640410903242306

Gucciardi, D., Gordon, S., & Dimmock, J. (2009a). Advancing mental toughness research and theory using personal construct psychology. International Review of Sport and Exercise Psychology, 2, 54-72. doi: 10.1080/17509840802705938

Gucciardi, D., Gordon, S., & Dimmock, J. (2009b). Development and preliminary validation of a mental toughness inventory for Australian football. Psychology of Sport and Exercise, 10, 201-209. doi:10.1016/j.psychsport.2008.07.011

Horsburgh, V., Schermer, J., Veselka, L., & Vernon, P. (2009). A behavioural genetic study of mental toughness and personality. Personality and Individual Differences, 46, 100-105. doi: 10.1016/j.paid.2008.09.009

Hutcheson, G., & Sofroniou, N. (1999). The multivariate social scientist: Introductory statistics using generalized linear models. London, England: Sage Publication.

Jones, G., Hanton, S., & Connaughton, D. (2007). A framework of mental toughness in the world's best performers. The Sport Psychologist, 21, 243-264.

Loehr, J. (1986). Mental toughness training for sports: Achieving athletic excellence. Lexington, MA: Stephen Greene Press.

Matsunaga, M. (2010). How to factor-analyze your data right: Do's, don'ts, and how-to's. International Journal of Psychological Research, 3, 97-110.

Middleton, S., Marsh, H., Martin, A., Richards, G., Savis, J., Perry, C., & Brown, R. (2004). The Psychological Performance Inventory: Is the mental toughness test tough enough? International Journal of Sport Psychology, 35, 91-108.

Nicholls, A., Polman, R., Levy, A., & Backhouse, S. (2008). Mental toughness, optimism, pessimism, and coping among athletes. Personality and Individual Differences, 44, 11821192. doi: 10.1016/j paid.2007.11.011

Seligman, M. (1990). Learned optimism. New York, NY: Knopf.

Serpell, R., & Haynes, B. (2004). The cultural practice of intelligence testing: Problems of international export. In R. Sternberg & E. Grigorenko (Eds.), Culture and competence: Contexts of life success (pp. 163-185). Washington, DC: American Psychological Association.

Sheard, M. (2013). Mental toughness: The mindset behind sporting achievement (2nd Ed.). New York, NY: Routledge/Taylor & Francis Group.

Sheard, M., Golby, J., & van Wersch, A. (2009). Progress toward construct validation of the Sports Mental Toughness Questionnaire. European Journal of Psychological Assessment, 25, 186-193. doi: 10.1027/1015-5759.25.3.186

Richard G. Cowden

North-West University

Anna Meyer-Weitz

University of Kwazulu-Natal

Address correspondence to: Richard G. Cowden, Email: richardgregorycowden@gmail.com; Phone: (+27) 018 299 1724, Institute of Psychology and Wellbeing, Potchefstroom Campus, North-West University, Private Bag X6001,Potchefstroom 2520, South Africa

Table 1
Participant Demographics

                               Age             Years of Participation

Variable                 n      M      SD       n       M      SD

Gender                  365   28.80   13.68    365    16.75   12.03
Male                    191   31.32   15.28    191    19.33   13.59
Female                  174   26.03   11.08    174    13.91   9.29
Ethnicity               365    --      --      365     --      --
Black                   80    23.46   6.29     80     9.84    5.27
Coloured                18    25.28   8.18     18     10.56   3.29
Indian                   9    24.44   7.52      9     11.11   2.67
White                   258   30.85   15.24    258    19.52   13.01
Participation Type      365    --      --      365     --      --
County Club             63    41.21   14.61    63     24.37   14.07
Local County            24    30.96   12.80    24     18.21   12.58
  Tournaments
National Tournaments    83    32.58   15.70    83     20.04   13.68
International           35    29.69   17.11    35     19.06   16.54
  Tournaments
University Team /       160   21.44   2.81     160    11.32   4.05
  League

Note, n = Number of participants,
M = Mean, SD = Standard Deviation.

Table 2
PCA Pattern Matrix of SMTQ Item Factor Loadings

Item    Factor 1   Factor 2   Factor 3

3         .793       .052      -.403
12        .712       .086      -.024
6         .609      -.050       .233
14        .558      -.121       .319
11        .493      -.022       .145
1         .442      -.112       .220
2        -.179      -.712      -.017
9        -.059      -.691       .100
4         .047      -.657       .200
7        -.073      -.651       .015
10        .181      -.632      -.201
8         .339      -.554      -.156
13        .032       .014       .780
5         .237      -.044       .496

Note. The item numbers correspond with the items
included on the original SMTQ validation study.

Table 3
Pearson Correlations between the Mental
Toughness Scale and Subscales

Variable                           (1)     (2)      (3)      (4)

(1) Total Mental Toughness         --     .12 **   .82 **   .41 **
(2) Confidence/Self-Efficacy       --       --     .34 **   .26 **
(3) Emotional/Cognitive Control    --       --       --     .19 *
(4) Positive Perspective           --       --       --
M                                 41.22   18.94    16.53     5.67
SD                                4.67     2.50     3.11     0.85

Note. * p = .001 (two-tailed); ** p < .001 (two-tailed).