Mental toughness in sport: testing the goal-expectancy-self-control (GES) model among runners and cyclists using cross-sectional and experimental designs

ABSTRACT

The Goal-Expectancy-Self-Control (GES) model provides a novel framework to study mental toughness in sport. This model proposes that mental toughness is a state-like multidimensional concept comprising three resources – challenging goals, self-efficacy, and self-control – that operate when athletes encounter a stressor that puts their goal achievement at risk. These resources are proposed to lead to optimal performance through four psychological mechanisms. These include attention, effort, perseverance, and strategies. The purpose of this research was to test this model in endurance sports within the confines of two studies (cross-sectional and experimental). Our samples consisted of 649 runners (Study 1) and 74 trained cyclists (Study 2). Overall, results support the GES model. Taken together, results indicate that mental toughness resources are positively related to endurance performance through the four psychological mechanisms. These findings contribute to a better understanding of mental toughness, as well as underline the importance for athletes to learn how to set challenging goals, attain and sustain high self-control and self-efficacy levels to optimally deploy their psychological mechanisms and reach their goals. Applied implications for athletes, coaches, and mental performance consultants are discussed.

KEYWORDS:

• Athlete
• performance
• self-efficacy
• stressors
• ego depletion

Acknowledgements

We thank Dr. Guy Thibault (associate professor, Université de Montréal) for providing assistance with the methodology, validation of the experimental study, and analysis discussions. We thank Dr. François Billaut (Université Laval), Dr. Alexandre Gareau (Université Laval). Thank you to Richard Bradet and Bei Feng for their support in the analysis, as well as Bruno Langlois (kinesiologist, cycling coach) for methodology discussions and resources for the experimental study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this research are available from the corresponding author, [CBT], upon reasonable request.

Ethics statement

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional committee (Comité d’éthique de la recherche de l’Université Laval: Approval No 2019-078) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Credit authorship contribution statement

Christiana Bédard-Thom: Conceptualisation, Methodology, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, review & editing, Visualisation, Project administration. Frédéric Guay: Conceptualisation, Methodology, Validation, Formal analysis, Resources, Data curation, Writing – original draft, review & editing, Supervision, Funding acquisition. Christiane Trottier: Conceptualisation, Methodology, Validation, Resources, Data curation, Writing – original draft, review & editing, Supervision, Funding acquisition.

Notes

1 McDonald omega value (MOV) is an alternative to Cronbach alpha value (CAV) when it comes to assess the internal consistency or reliability of a scale. Essentially, MOV is a more precise estimate of a scale reliability because this computational method does not assume that all pairs of items of a given scale has the exact same correlation, which is a strong assumption in CAV that is referred to tau-equivalence. Specifically, essential tau-equivalence would be satisfied for two items if two people who differ by a given amount of X differ from each other by the same amount in their response to the two items intended to measure X. Because this rarely the case in practice, MOV offer a better estimate of reliability of a given construct. MOV has the same cut-off value than CAV, which is around .70 (Hayes & Coutts, 2020).

2 The hybrid CFA/B-ESEM model contained 152 free parameters for 649 participants leading to an approximate ratio of four participants by free parameters (4:1). Often cited is a ratio of 20:1 (e.g., Kline, 2016), but this suggested ratio depends greatly on the expected magnitude of loadings and the number of factors used in the model. The hybrid CFA/BESEM solution contained 7 factors with the number of indicators ranging between 3 and 7. If we use a conservative criteria that loadings will be around .50 with 7 latent factors and a minimum of 4 indicators per factor our ratio of 4:1 appears adequate in light of past findings showing that sample size requirements reduce as the number of latent factors increase and the number of indicators (Wolf et al., 2013). Moreover, the SEM part of the analyses contains only 20 free parameters leading to a ratio of 32:1 which is adequate for regression models.

Additional information

Funding

Our work was funded by Fonds de Recherche du Québec - Société et Culture: [Grant Number 00 00271687].