Relationships Between Athletes' Self-Reported Grit Levels and Coach-Reported Practice Engagement Over One Sport Season.
Grit is a personality characteristic related to persistent hard work and associated with success in various achievement domains (e.g., Duckworth, Peterson, Matthews, & Kelly, 2007). Defined broadly as passion and perseverance for long-term goals (Duckworth et al., 2007), it comprises two facets: consistency of interests (CI; tendency to remain interested in the same goals over time) and perseverance of effort (PE; tendency to work hard despite adversity).To assess grit, Duckworth et al. (2007) developed a self-report questionnaire, the Grit Scale,which contains two subscales, one assessing PE and the other assessing CI. To obtain an overall grit score, Duckworth et al. (2007) combined the scores from both subscales. In sport, PE was associated with athletes'weekly hours of quality, deliberate practice and with more frequent attendance at mandatory and optional practice; CI protected athletes from thinking about switching or quitting their sport (Tedesqui & Young, 2017). When contrasted against self-control and conscientiousness variables, grit variables best predicted quality practice and sport commitment outcomes (Tedesqui & Young, 2018). Grit has also been associated with better decision-making (Larkin, O'Connor, & Williams, 2015), higher skill level (Meyer, Markgraf, & Gnacinski, 2016), and higher sport engagement (Martin, Byrd, Watts, & Dent, 2015).
Despite these works, research has yet to investigate the impact of athletes' grit levels on their practice engagement longitudinally. Also, no studies have examined whether athletes' self-reported grit scores can be reliably associated with their coach's ratings for practice quality. Cross-validating athletes' grit scores with coach's ratings for their practice is important given that astute coach's judgments (though subjective) might represent ecologically valid measures in a talent development/selection domain (Hendry, Williams, & Hodges, 2018). In this study, we examine whether an athlete's self-reported grit levels during early season might be considered proxy indicators for their level and consistency of practice engagement over the season.
Most sport studies on grit use done composite score of the CI and PE subscales that constitute the Grit Scale (e.g., Larkin et al., 2015; Martin et al., 2015), rather than examining differential associations with the facets. This is an important oversight as predictive validity tends to increase with use of facet subscales (Paunonen & Ashton, 2001) and differential effects have been attributed to PE and CI in sport (Tedesqui & Young, 2017) and in other domains (e.g., Crede, Tynam, & Harms, 2017). Furthermore, in their meta-analysis, Crede et al. (2017) raised concerns about the construct of grit and its predictive validity. Crede et al. questioned whether grit should be assessed as a single construct or, instead, two associated facets and suggested that the utility of the construct may be attributed primarily to the perseverance of effort facet. Given the mentioned gaps and the issues raised regarding its construct and predictive validity, this study aimed to: (a) assess relationships between athletes' grit levels and coach-reported quality practice indicators, longitudinally; and (b) assess which measure of grit (i.e., PE or CI facets, or composite grit) showed greater predictive validity for quality practice indicators.
Survey data were collected over three time points across a season (start of outdoor season around May). At T1 (August/September), athletes completed a demographic questionnaire (e.g., age and sex) and the Grit Scale (Duckworth et al., 2007; see Tedesqui & Young, 2017, for information on indices of structural validity tested with exploratory and confirmatory factor analyses). Coaches independently rated their athletes' engagement in training at that point in the season (Tl), then again around two months later (T2), and again four months after Tl (T3). The average interval between coaches' survey responses was 50.5 days (range = 39-62) between Tl and T2, and 56.5 days (range = 53-60) between T2 and T3.
Participants were 13 canoe/kayak athletes (ages 13-23, [M.sub.age] = 17.40, SD = 2.99), including three females and 10 maleswho competed at provincial (n = 1),national (n = 5), and international (n = 7) levels. All trained at one of two Eastern Canadian canoe/kayak clubs that had structured and competitive training programs. Eight athletes were from Club A and five were from Club B. On average, they reported 10.46 (SD = 6.34) weekly hr of sport-specific training.
Two coaches participated. Coach A (Club A) had worked for 10 years as a coach, had a master's degree in physical training, an MBA in sport management, and had been coaching the participants for 6 months for approximately 20 hr per week. Coach B (Club B) had been coaching for 8 years, was a kinesiologist, and had been coaching the participants for 12-18 months for 8 hours per week. Thus, the coaches were qualified and had been involved with athletes long enough to be able to rate their engagement in training. Adult participants gave informed consent; minors gave assent and had parental informed consent. The protocols were submitted to, and approved by, an institutional review board for testing of human subjects. Surveys
The Grit Scale. At T1, athletes completed the Grit Scale (Duckworth et al., 2007), which measures: (a) perseverance of effort (4 items; [alpha] = .75), e.g., "1 have achieved a goal that took years of work'"; and (b) consistency of interests (6 items; a = .81), e.g., "My interests change from year to year." The structural validity of these 10 items was established by Tedesqui and Young (2017). Athletes7 scores were computed for the perseverance of effort and consistency of interests subscales, and for overall grit (i.e., composite of perseverance of effort and consistency of interests).
Coach Rating of Practice Engagement. At each time point, coaches rated each of their athletes (who had completed the survey at T1) on a Likert scale ranging from I (not at all true) to 5 (very true) on six items indicating practice engagement and one item indicating attendance. Based on iterative discussions among the authors, the items were intuitively designed to tap athletes' engagement with their practice activities and to conceptually represent indicators of practice quality. These indices have been supported as key indicators of quality athletic practice in the coaching science literature, for example, in studies where coaches have described instances of an effectively self-regulated athlete during practice sessions (Young & Starkes, 2006a, 2006b). Their coach rating form read, "At this point in the season, please use the scale below to rate each of your athletes on the degree to which he/she ...": (a) perseveres through setbacks/difficulties, (b) is diligent with respect to training responsibilities, (c) invests extra time in training, (d) has a positive attitude toward training, (e) takes on new challenges, (f) consistently works hard, and (g) misses practice. We calculated athletes' overall level of practice engagement as an average of the first six items (a through f); the final item (g) was reverse-scored to represent attendance. Notably, coaches' and athletes' responses were independent--the athletes did not know how their practice engagement was rated by coaches, nor did coaches know athletes' grit scores.
Planned Data Analysis
We calculated Pearson correlations between each of the three grit scores and coach ratings (i.e., the seven items separately, and overall level of practice engagement) at each time point (see Table 1). To assess whether athletes' grit scores at T1 were related to stability of practice engagement over time, we computed an index of stability of practice engagement for each athlete. As recommended by Ram and Gerstoff (2009), we employed a commonly used index called the intraindividual standard deviation (iSD); this index is the standard deviation based on the distribution of scores obtained across repeated measurements for a single individual and represents the extent to which the individual's scores vary around the mean score for that individual. Thus, a large iSD indicates low stability in one's practice engagement relative to the person's mean whereas a small iSD indicates high stability. Finally, we inspected the Pearson correlations between athletes' grit scores and their iSD of practice engagement.
Correlations and Longitudinal Trends
Table I shows correlations between grit and ratings of practice engagement. Table 2 shows mean scores on grit, ratings of practice engagement, and index of stability.
Across the three time points, perseverance of effort (PE) was positively related to overall practice engagement. Additionally, PE was positively related to diligence with responsibilities (T2 and T3), investment of extra time (T3), positive attitude (T3), and consistent hard work (T1 and T2), with trends pointing to an increase in these ratings over time (except for consistent hard work). PE was also positively related to attendance at T3. Consistency of interests (CI) was inversely related to positive attitude (T2) and openness to new challenges (T1). No significant associations were found between overall grit and ratings of practice engagement at any time point.
Neither CI (r = .40, p = .17) nor overall grit (r= -.15,p = .63) were significantly associated with indices of stability (iSD) over time. The association between PE and stability approached significance (r = -.54, p = .06). To follow up on this trend with PE and facilitate the visualization of variability around each sub-element of practice engagement, we plotted the coach ratings of practice engagement over time using performance profile cases or radar charts (Butler & Hardy, 1992). In particular, we were interested in visually depicting whether the association between PE and coach ratings over time were different for certain aspects of quality engagement than others.
To illustrate, we show the performance profiles for the three highest and the three lowest scorers on PE (see Figure I).The performance profiles show that athletes with the highest PE scores (IDs 2, 10, 3) cover a wider area of the graph, indicating greater practice engagement. Further, the overlap in T1, T2, and T3 lines suggest that high PE athletes also have stable practice engagement over time. Notably, athlete ID 3 showed the maximum level of practice engagement with no variability over time. Conversely, the profiles of athletes with the lowest PE (IDs 13, 12, 11) cover a smaller area and show little overlap between the lines indicating poorer quality and more variable practice engagement over time. The profiles also depict where athletes stand in each sub-element of practice engagement, allowing coaches to track progress (or the lack thereof) in specific indicators of practice engagement. For example, athlete ID 12 (lowest PE) showed a notable decrease in its diligence, time investment, and positive attitude over time.
We investigated the associations between athletes' grit scores (perseverance of effort, consistency of interests, and overall grit) and their practice engagement as rated by their coaches over time. Overall, we found that athletes' self-reported PE was the grit variable most related to coach ratings of athletes' practice engagement with large effect sizes (Cohen, 1992). Athletes who scored the highest on PE, thus who reported a tendency to sustain effort despite adversity, were also independently reported by their coaches to have shown a high level of practice engagement and to have sustained such high engagement longitudinally. The same could not be said of athletes who scored high on CI or overall grit, with both variables showing respectively little to no associations with practice engagement. These findings suggest that the higher an athlete's level of self-reported PE, the higher and the more stable their practice engagement (as assessed by their coach) is over time. Conversely, the lower an athlete's PE, the lower and the more variable their practice engagement is over time.
Our findings may have practical implications for coaches. Coaches could benefit from knowing whether a measure of an athlete's PE level at one point can predict the level and the fluctuation of their practice engagement as well as their attendance later in a season. From a developmental perspective, an assessment of athletes' PE levels may help coaches identify which athletes are likely to later show low and/or unstable practice engagement. Equipped with grit information, coaches would be in a better position to target preventive interventions to those athletes "at risk" (i.e., those low on PE) for showing less than optimal practice engagement or inconsistent practice habits. Early season assessment using athletes' self-report of grit, however, should be but one strategy among a repertoire of activities that coaches use to better understand their athletes. Additionally, our intent in showing the performance profiles was to demonstrate how coaches may apply the grit (especially PE) ratings with their athletes to help them reflect about changing aspects of their practice engagement over time. As recommended in applied interventions (e.g., Gill, Williams, & Reifsteck. 2017), coaches may engage their athletes in profiling, perhaps using multiple personal (within-athlete) profile instances over a season to track the impact of interventions designed to optimize their practice activities.
From a talent selection perspective, knowing athletes' level of PE might help coaches decide on which athletes to invest limited resources (e.g., spots in a team, access to quality training) to maximize the "returns on their investment" by choosing those who are more likely to get the most out of their training. This said, we are very cautious to make this recommendation based on this one study. Future research replicating the current trends in different contexts, especially with in situ measures of quality practice, would help continue this discussion.
Our results showed that only PE was related to overall practice engagement and that overall grit was not related to any measure of practice at any time point. Had we assessed grit only at the single scale level, we could have mistakenly concluded that grit is not associated with practice engagement. From a methodological angle, therefore, the current findings echo the recommendation that researchers inspect the predictive validity of the grit facets separately (e.g., Tedesqui & Young, 2017). Our findings are also consistent with Crede et al.'s (2017) meta-analytic interpretation outside sport that PE mostly predicts outcomes, not CI. It is important to note that correlations between CI and practice engagement variables were larger than criterion associations portrayed in Crede et al.'s meta-analysis. That said, our small sample size constrained the capability to detect statistical significance. Thus, we have elected to remain conservative and careful in interpreting any non-statistically significant findings; future research using similar procedures and larger longitudinal samples may afford stronger conclusions on these trends. Therefore, the current findings suggest that PE is related to the quality of athletes' efforts in practice engagement over time, but that CI is not necessarily related to practice engagement. Notably, prior cross-sectional work (Tedesqui & Young, 2017) suggested that athletes' self-reported CI scores do not associate with the volume/quantity of athletes' practice, but rather with athletes' efforts to commit to practice. Specifically, athletes" higher CI was associated with the propensity to have fewer thoughts of switching out of, or quitting, a sport. Future research is needed to test a preliminary postulate, that is, CI associates with consistent attraction toward a sport and PE relates to quality practice engagement within sport practice. From a practical standpoint, coaches could benefit from knowing the grit measures they can rely upon to understand different athletic outcomes.
Strengths of this study include its prospective design and the collection of coaches' ratings independent of athletes' self-reported grit scores, which enhanced the validity of our findings. A limitation was the small sample size which prevented us from making stronger inferences. Future studies with a larger sample and a longer assessment period (e.g., across seasons) may confirm our findings and provide stronger evidence for the association between athletes' PE scores and the stability/variability of training.
This article is, to our knowledge, the first longitudinal investigation of the role of grit within sport. It sensitizes sport researchers and practitioners to the importance of considering grit at the facet level in order to appropriately distinguish the contribution of each grit facet on sport outcomes. Notably, we identified perseverance of effort as the only grit variable associated with athletes' quality practice engagement. Finally, the associations between coach-reported practice engagement and self-reported grit scores may constitute evidence of external validity for the grit scale as it is currently applied in the sport domain. Specifically, this study was the first to find evidence that an early self-report measure of athletes' perseverance of effort might be considered a proxy indicator for their subsequent level and consistency of practice engagement over the course of the season.
Rafael A. B. Tedesqui
Bradley W. Young
University of Ottawa
Address correspondence to: Rafael A. B. Tedesqui, 2600 College Street, Mailbox 31, Department of Sports Studies, Bishop's University, Sherbrooke, Quebec, JIM 1Z7, Canada; Telephone: 819-822-9600 ext. 2188; E-mail: email@example.com.
Butler, R. J., & Hardy, L. (1992). The performance profile: Theory and application. The Sport Psychologist, 6, 253-264.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.
Crede, M., Tynan, M. C., & Harms, P. D. (2017). Much ado about grit: A meta-analytic synthesis of the grit literature. Journal of Personality & Social Psychology, 113, 492-511.doi: 10.1037/pspp0000102
Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly, D. R. (2007). Grit: Perseverance and passion for long-term goals. Journal of Personality & Social Psychology, 92, 1087-1101. doi: 10.1037/0022-35188.8.131.527
Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100, 363-406.
Gill, D. L., Williams, L., & Reifsteck, E. J. (2017). Personality. In: Psychological dynamics of sport and exercise (4th ed., pp.37-55). Champaign, IL: Human Kinetics.
Hardy, L., Barlow, M., Evans, L., Rees, T., Woodman, T., & Warr, C. (2017). Great British medalists: Psychosocial biographies of Super-Elite and Elite athletes from Olympic sports. Progress in Brain Research. 232, 1-119. doi:10.1016/bs.pbr.2017.03.004
Hendry, D. T., Williams, M. A., & Hodges, N. J. (2018). Coach ratings of skills and their relations to practice, play and successful transitions from youth-elite to adult-professional status in soccer. Journal of Sports Sciences. Advance online publication, doi: 10.1080/02640414.2018.1432236
Larkin, P., O'Connor. D., & Williams, A. M. (2015). Does grit influence sport-specific engagement and perceptual-cognitive expertise in elite youth soccer? Journal of Applied Sport Psychology. 28, 129-138. doi: 10.1080/10413200.2015.1085922
Martin, J. L., Byrd, B., Watts, M. L., & Dent, M. (2015). Gritty, hardy, and resilient: Predictors of sport engagement and life satisfaction in wheelchair basketball players. Journal of Clinical Sport Psychology. 9, 345-359. doi: 10.1123/jcsp.2015-0008
Meyer, B. M., Markgraf, K. M., & Gnacinski, S. L. (2016). Examining the merit of grit in women's soccer: Questions of theory, measurement, and application. Journal of Applied Sport Psychology, 29, 353-366. doi: 10.1080/10413200.2016.1255277
Paunonen, S. V., & Ashton, M. C. (2001). Big Five factors and facets and the prediction of behavior. Journal of Personality & Social Psychology. 81, 524-539.
Ram, N., & Gerstorf, D. (2009). Time-structured and net intraindividual variability: Tools for examining the development of dynamic characteristics and processes. Psychology and Aging, 24, 778-791. doi: 10.1037/a0017915
Starkes, J. L., Deakin, J. M., Allard, F., Hodges, N. J., & Hayes, A. (1996). Deliberate practice in sports: What is it anyway? In K. A. Ericsson (Ed.), The road to excellence: The acquisition of expert performance in the arts, sciences, sports and games (pp. 81-106). Mahwah, NJ: Erlbaum.
Tedesqui, R. A. B., & Young, B. W. (2017). Investigating grit variables and their relations with practice and skill groups in developing sport experts. High Ability Studies, 28, 167-180. doi: 10.1080/13598139.2017.1340262.
Tedesqui, R. A. B., & Young, B. W. (2018). Comparing the contribution of conscientiousness, self-control, and grit to key criteria of sport expertise development. Psychology of Sport & Exercise, 34, 110-118. doi:10.1016/j.psychsport.2017.10.002
Williams, A. M., & Ford, P. R. (2008). Expertise and expert performance in sport. Internationa! Review of Sport and Exercise Psychology. 7,4-18, doi: 10.1080/17509840701836867
Young, B. W., & Starkes, J. L. (2006a). Coaches' perceptions of non-regulated practice behaviors in competitive swimmers. International Journal of Sports Science and Coaching, 7(1), 53-68.
Young, B. W., & Starkes, J. L. (2006b). Measuring outcomes of swimmers' non-regulation during practice: Relationships between self-report, coaches' judgments, and video-observation. International Journal of Sports Science and Coaching, 7(2), 131-148.
This work was supported by the Social Sciences and Humanities Research Council of Canada (SSHRC).
Caption: Figure 1. Practice engagement of athletes with highest (top) and lowest (bottom) scores on PE.
Table 1 Pearson Correlations Between Athletes' Grit Scores (Perseverance of Effort, Consistency of Interests, and Composite) and Coach Ratings of Practice Engagement Over Time Perseverance of Effort Indicators of Practice Engagement T1 T2 T3 (a) Perseverance .36 .29 .25 through setbacks (b) Diligence with .41 .58 * .59 * responsibilities (c) Investment of extra .33 .53 .83 *** time (d) Positive attitude .55 .55 .79 ** (e) Openness to new .28 -.05 -.24 challenges (f) Consistent hard .60 * .73 ** .45 work Overall practice .60 * .58 * .65 * engagement (a) (g) Attendance .25 -.30 .57 * Consistency of Interests Indicators of Practice Engagement T1 T2 T3 (a) Perseverance -.55 -.37 .02 through setbacks (b) Diligence with -.07 -.44 -.38 responsibilities (c) Investment of extra -.11 -.04 -.38 time (d) Positive attitude -.30 -.62 * -.45 (e) Openness to new -.57 * -.22 .08 challenges (f) Consistent hard -.20 -.40 -.45 work Overall practice -.39 -.46 -.36 engagement (a) (g) Attendance -.08 .25 -.36 Grit Indicators of Practice Engagement T1 T2 T3 (a) Perseverance -.14 -.05 .25 through setbacks (b) Diligence with .32 .16 .22 responsibilities (c) Investment of extra .21 .45 .44 time (d) Positive attitude .25 -.03 .34 (e) Openness to new -.23 -.24 -.15 challenges (f) Consistent hard .39 .33 .03 work Overall practice .22 .14 .29 engagement (a) (g) Attendance .16 -.07 .21 Note. Where r [greater than or equal to] [absolute value of .56], * p < .05. Where r [greater than or equal to] [absolute value of .69], ** p < .01. Where r [greater than or equal to] [absolute value of .83], *** p < .001. Small, medium, and large effcet sizes are respectively r = .10, r = .30, r = .50 (Cohen, 1992). (a) Calculated as an average of items (a) through (f). Table 2 Athletes' Mean Scores on Grit, Perseverance of Effort (PE), Consistency of Interests (CI), Coach Ratings of Practice Engagement, and Index of Stability (iSD) Coach Ratings of Practice Engagement Athlete Grit PE CI (a) Tl T2 T3 iSD (b) 1 3.54 3.78 3.33 3.50 4.00 4.17 .35 2 4.58 5.00 4.17 4.50 4.33 4.67 .17 3 3.38 4.25 2.50 5.00 5.00 5.00 0 4 3.63 3.25 4.00 3.67 4.17 4.33 .35 5 3.88 4.25 3.50 3.17 3.67 3.50 .26 6 3.04 3.75 2.33 3.50 4.17 3.50 .39 7 3.21 3.75 2.67 3.33 3.17 3.33 .10 8 4.13 3.75 4.50 2.67 3.33 3.33 .39 9 3.17 3.00 3.33 3.50 4.17 4.17 .39 10 4.04 4.75 3.33 4.67 4.50 4.67 .10 11 3.38 2.75 4.00 3.50 3.50 3.83 .19 12 3.50 2.50 4.50 3.33 3.67 2.83 .42 13 3.42 2.50 4.33 3.00 2.50 2.50 .29 Note. Both the Grit Scale and Coach Ratings were assessed on a 5-point Likert Scale. (a) Reverse-scored. (b) Intraindividual standard deviation representing the index of stability of practice engagement over time. A lower iSD indicates lower variability (or greater stability) of their overall ratings of practice engagement across the three time points.