Investigation of the interaction between coach transformational leadership and coaching competency change over time

Abstract

Drawing on social cognitive theory, we examined a cross-level interaction between coach transformational leadership and the within-individual change in evaluations of perceived coaching competency over a season. We applied linear growth modeling to analyze 246 collegiate competitive basketball players (147 males and 99 females from 24 teams). Our results revealed that coaches with high transformational leadership were evaluated as having a decrease in perceived coaching competency over time. In contrast, coaches with low transformational leadership were evaluated as having an increase in perceived coaching competency over time. Implications for coaching education in terms of theory and research are discussed.

Keywords

Basketball,motivation,social cognitive theory

Introduction

To serve the needs of athletes, coaches are expected to possess certain knowledge, skills, and abilities that provide athletes with task and psychological resources to enhance their performance.¹,² As coach John Wooden stated:

What it comes down to, I believe, is that mentoring often involves telling people what they need to hear, rather than what they want to hear. When you are able to be humbly honest with someone about a situation with which you have personal experience—even if you risk angering or hurting that person—you are offering the most valuable gift of all.^3:p.61

To capture athletes’ evaluations of coach’s abilities, Myers et al.⁴ presented the concept of coaching competency to illustrate athletes’ perceptions of confidence in their coaches’ abilities. Research suggests that coaching competency is associated with athletes’ satisfaction with their coaches,⁵,⁶ self-perceived antisocial behaviors,⁷ perceptions of role ambiguity,⁸ and perceptions of coach support.⁹ These empirical findings imply that coaching competency serves as a means to predict many important athlete behavioral and attitudinal outcomes.

Given the extant research on coaching competency was conducted by cross-sectional designs (e.g., Bosselut et al.;⁸ Coussens et al.;⁹ Kao and Tsai;⁶ Malete et al.⁷), we have limited understanding of the dynamic nature of coaching competency. Social cognitive theory¹⁰,¹¹ predicts that coaching competency may change over time because it represents efficacy ratings of the proxy agent (i.e., coach). When athletes perceive they are learning from coaches with high proxy efficacy, they may increase greater efforts and motivation to engage managing arduous task demands and the risk of failure. In that manner, athlete abilities will receive better development from their coaches over time. As such, athletes will be capable to handle problems or challenges independently and adjust their perceptions of coaching competency. Given the dynamic nature of coaching competency is assumed by social cognitive theory,¹⁰,¹¹ there is a need to empirically test this theoretical argument.

Further, due to the critical role of the coach leadership style in athlete development,^12–14 we assert that, over time, coach transformational leadership may moderate the within-individual change in perceived coaching competency. Transformational leadership refers to a leader (coach) uses inspirational strategies to develop followers’ (athletes’) full potential.¹³,^15–17 Both the leadership^18–20 and coach literatures²¹,²² indicate that transformational leadership advances follower personal growth due to its developmental nature. In a sense, it is possible that the developed athlete abilities increase feelings of self-esteem and independence from their coaches. Thus, we asserted that athlete self-esteem and independence cultivated by coach leadership behaviors may be associated with a within-individual change in perceived coaching competency. This assertion suggests that coach transformational leadership may interact with the within-individual change in perceived coaching competency.

In the current study, we measured pre- and post-season perceived coaching competency and coach transformational leadership in the context of a competitive sport season. With this design, we were able to examine the interaction between coach transformational leadership and the within-individual change in perceived coaching competency over a season. From a theoretical perspective, we contribute to the understanding of the dynamic nature of coaching competency based on Bandura’s social cognitive theory¹⁰,¹¹ and the developing function of the coach leadership style in this social learning process.

Coaching competency

Coaching competency is defined as athletes’ evaluations of their coaches’ abilities to affect their learning and performance.⁴ The concept of coaching competency was developed from Feltz et al.’s¹ concept model of coaching efficacy (for a review, see Feltz et al.²). Myers et al.⁴ advanced the construct of coaching competency by denoting four types of competency: (a) motivation (i.e., ability to affect the emotions and skill of athletes), (b) game-strategy (i.e., ability to lead during competition), (c) technique (i.e., instructional and diagnostic abilities), and (d) character-building (i.e., ability to influence character development and positive attitude toward sports).

Coaching competency can be understood through Bandura’s¹⁰,¹¹ social cognitive theory. According to Bandura,¹⁰ individuals’ personal efficacy, defined as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (p. 3), guides their lives. Social cognitive theory¹⁰,¹¹ asserts that efficacy beliefs are involved in three models of human agency: personal agency, collective agency, and proxy agency, and translate to self-efficacy, collective efficacy, and proxy efficacy. Of these three models, proxy efficacy (or proxy agency) is proposed to understand the mechanism of coaching competency.⁶ Proxy efficacy refers to how people seek their well-being, security, and valued outcomes through an external proxy agency. People turn to proxy control because they have not developed the means to achieve certain goals. As such, people believe other agents can do this better, or do not want to saddle themselves with the burdensome aspects of direct control.¹⁰,¹¹ In this regard, when an athlete enters a team, he or she begins learning from the coach, and the coach exerts proxy control over an athlete’s improvement and development by providing knowledge and expertise.²³ Thus, coaching competency can be conceptualized as a form of proxy efficacy (e.g., Bray et al.²⁴; Feltz et al.²; Jackson et al.²⁵; Kao and Tsai⁶), and athletes’ evaluations of coaching competency reveal the extent to which athletes can learn from and rely on their coaches to reach their goals in terms of athletic performance and psychological development.

Transformational leadership and coaching competency

A transformational leader transforms followers’ motivation to perform beyond expectations by making followers aware of the importance and value of superordinate goals.¹⁵,²⁶ Bass¹⁵,²⁶ conceptualized four subdimensions of transformational leadership: idealized influence, inspiration motivation, intellectual stimulation, and individualized consideration. Idealized influence refers to the leader behaving in ways that cause followers to identify with the leader. Inspiration motivation refers to the leader articulating a vision that is appealing and inspiring to followers. Intellectual stimulation involves a transformational leader challenging assumptions and soliciting followers’ ideas. Individualized consideration involves attending to each follower’s needs, whereby a transformational leader acts as a mentor to the follower and listens to the follower’s concerns and needs. In sport settings, transformational leadership have been identified as an important construct by which to understand athlete outcomes (e.g., Kao et al.;²⁷ Price and Weiss;²⁸ Rowold²⁹).

According to Bass,¹⁵,¹⁶ transformational leaders serve as mentors to their followers. Transformational leadership involves broadening and elevating followers’ goals and providing them with the values, enhanced skills, and confidence to go beyond minimally acceptable expectations of performance.³⁰ Transformational leaders value learning, empowerment, and development of followers by inspiring and intellectually stimulating followers and serving as coach, teacher, and mentor.¹⁹,³¹ Similarly, the mentor-protégé literature indicates that mentors value personal learning, as seen in their offering opportunities, such as challenging job assignments that help to advance the followers’ career, and coaching to achieve the goals of recognition and success within the organization.³² Transformational coaches likely serve as mentors, and coaches as mentors likely exhibit certain degrees of transformational leadership behavior. In a sense, the mentoring component is embedded in the dynamic transformational leadership coaching process.³³

When considering the developing and nurturing nature of transformational leadership, we argue that a transformational coach is a positive influence on athletes’ development and growth in terms of their competency and motivation. Transformational coaches facilitate athletes’ social learning by empowering athletes to build their own competency and approach their goals during a competitive season. For athletes, a transformational coach is an effective proxy agent to navigate the uncertainties and challenges throughout a competitive season. A transformational coach is concerned with developing meaning, instilling pride, analyzing game strategy, and encouraging athletes to go beyond self-interest for the good of the team. Given athletes’ personal skill and motivation levels have been enhanced by coach transformational leadership, it is possible for athletes to adjust their perceptions of coaches and their evaluations of coaching competency over time. As such, we propose that there is an interaction between coach transformational leadership and the within-individual change in perceived coaching competency over time.

The present study

The purpose of this study was to examine the role of coach transformational leadership in the change of perceived coaching competency over time. To systematically approach our argument, we tested the cross-level interaction between coach transformational leadership and the within-individual change in athlete evaluations of coaching competency over time. Drawing on social cognitive theory¹⁰,¹¹ we argued that the within-individual change in athlete evaluations of coaching competency over time (i.e., a sport season) is contingent on levels of coach transformational leadership. Coaches with high transformational leadership facilitate athletes’ social learning in building competence and enhancing motivation. Over time, the within-individual developmental components in athletes may enhance their self-esteem and independence from coaches and may be associated with a decrease in their perceived coaching competency. In contrast, coaches with low transformational leadership may be not able to effectively develop athlete competence and motivation. In that, athletes relay more on their coaches and may be associated with an increase in within-individual perceived coaching competency over time. As such, we propose:

Hypothesis. There is an interaction between coach transformational leadership and the within-individual change in athlete perceived coaching competency over time. High coach transformational leadership may be associated with a decrease in coaching competency. Whereas low coach transformational leadership may be associated with an increase in coaching competency.

Method

Participants

We recruited 246 basketball players (147 males, 99 females) from 24 teams (14 male teams, 10 female teams) to participate in the current study. The players participated in Division I men’s and women’s college basketball in Taiwan. Division I basketball players were selected based on their competitive performance and regularly practiced and interacted with their coach. The participants’ mean age was 20.2 years (SD = 1.34); they played competitive basketball for a mean of 7.52 years (SD = 2.66) and played for the present coach for a mean of 2.54 years (SD = 1.61).

Procedure

The study was approved by the university’s ethics board. Informed consent was obtained from all participants. Questionnaires were distributed before or after a training session in the absence of coaches. Participants were informed that their involvement in this study was voluntary and that their responses would remain confidential. Participants were administered measures of coaching competency and transformational leadership at two times, at pre-season (November, Time 1) and post-season (March, Time 2). The use of two time points enables us to monitor the within-individual change in evaluations of coaching competency over a season.

Measures

Transformational leadership

Transformational leadership was assessed, using the Multifactor Leadership Questionnaire (MLQ-5X³⁴). We applied the translation and back-translation technique³⁵ to all items of the MLQ-5X. Transformational leadership in the MLQ-5X comprises four subscales: idealized influence (eight items; e.g., “acts in ways that build my respect”), inspirational motivation (four items; e.g., “talks optimistically about the future”), intellectual stimulation (four items; e.g., “seeks differing perspectives when solving problems”), and individualized consideration (four items; e.g., “helps me to develop my strengths”). Athletes were asked to consider how often their respective coach displayed the identified behavior on a 5-point scale (1 = not at all, 5 = frequently). Our results, using growth modeling (as detailed in the Statistical Analyses section), suggest that coach transformational leadership remains stable and does not change over time. Therefore, we used transformational leadership at Time 1 in the following analysis. The results of a confirmatory factor analysis revealed a good fit for Time 1 transformational leadership (χ² = 518.38, df = 164, NFI = .94, CFI = .96, NNFI = .95, RMSEA = .10). These four subdimensions of transformational leadership, as measured by the MLQ-5X,³⁴ were developed across different contexts and should treat all of them as a global construct. That is, the four subdimensions were combined to create an overall measure of transformational leadership (e.g., Kao and Tsai;⁶ Kao et al.;²⁷ Stenling and Tafvelin³⁶). The Cronbach’s alphas for this overall measure was .94.

Coaching competency

Evaluations of coaching competency were assessed, using the Coaching Competency Scale (CCS⁸). Again, we applied the translation and back-translation technique³⁵ to all items of the CCS and measured coaching competency from athletes at two different times (i.e., pre- and post-season, Time 1 and Time 2). The CCS items were preceded by the stem, “How competent is your head coach in his or her ability to …?” The CCS consists of a motivation competency subscale (seven items; e.g., “motivate his/her athletes”), game-strategy competency subscale (seven items; e.g., “understand competitive strategies”), technique-competency subscale (six items, e.g., “coach individual athletes on technique”), and character-building competency subscale (four items; e.g., “instill an attitude of respect for others”). Athletes were asked to consider how competent they perceive their coach at exhibiting the identified ability on a 5-point Likert-type scale (1 = complete incompetence, 5 = complete competence). The results of a confirmatory factor analysis revealed a good fit for coaching competency at Time 1 (χ² = 774.90, df = 246, NFI = .96, CFI = .98, NNFI = .97, RMSEA = .09) and at Time 2 (χ² = 641.66, df = 246, NFI = .97, CFI = .98, NNFI = .98, RMSEA = .08). Previous analysis has shown that the four subscales of coaching competency are highly correlated.⁵ Therefore, we used an overall measure of coaching competency. The Cronbach’s alphas for overall coaching competency measure at Time 1 and Time 2 were .97 and .97, respectively (measurement invariance tests as detailed in the confirmatory factor analysis section).

Statistical analyses

We followed Bliese and Ployhart³⁷ to conduct a three-level linear growth model (i.e., within-individual, between-individual, and group levels of analysis) to test the hypothesis. The three-level growth linear model reflected the hierarchical structuring of repeated measure over time, nested within the individual respondent under different team settings. Our hypothesis asserts an interaction between coach transformational leadership and the within-individual change in coaching competency over time. To test this hypothesis, we regressed evaluations of coaching competency on Time at Level 1, coach transformational leadership at Level 2, and different team settings at Level 3. The R program and the multilevel package³⁸ were utilized to conduct this analysis.

The Level-1 equation reflected the within-individual change in evaluations of coaching competency, which captured the change in evaluations of coaching competency within an athlete over a season. Thus, the Level-1 equation was presented for a three-level linear growth model as

Y_{ijk} = β_{0 j k} + β_{1 j k} {(Time)}_{ijk} + R_{ijk},

where

Y_{ijk}

is coaching competency at Time i for individual respondent j on team k;

β_{0 j k}

and

β_{1 j k}

are the intercept and slope that define the linear trajectory unique to the individual respondent j on team k; and

R_{ijk}

is an error term. The intercept (

β_{0 j k}

) reflects the initial assessment of evaluations of coaching competency, and the slope reflects the change in evaluations of coaching competency between Time t and Time t+1.

The Level-2 equations reflect the role of coach transformational leadership (TFL) in explaining the intercept and slope variation of the Level-1 equation. The slope variability indicates the cross-level interaction between coach transformational leadership and the within-individual change in evaluations of coaching competency. The Level-2 equations for the model in this study are shown below:

β_{0 j k} = γ_{00 k} + γ_{01 k} {(TFL)}_{j k} + U_{0 j k}

β_{1 j k} = γ_{10 k} + γ_{11 k} {(TFL)}_{j k} + U_{1 j k}

The first Level-2 equation states that respondent j’s mean level of coaching competency on team k. The second Level-2 equation determines whether transformational leadership can explain some of the variation in the time-coaching competency slope. With this nesting nature (i.e., multilevel framework), the coefficient ( $γ_{11 k})$ reflects the cross-level interaction between coach transformational leadership and the within-individual change in coaching competency over time (i.e., hypothesis).

Finally, the Level-3 model presents the team-level varying parameters that define the slope, and the intercept of the Level-2 equations consists of a random variable. At the team level, we control team winning rate (TWR) at the end of the season. Thus, we can further define the Level-3 equations for these terms as:

γ_{00 k} = δ_{000} + δ_{001} (TWR) + V_{0 k}

γ_{10 k} = δ_{100} + V_{1 k}

γ_{01 k} = δ_{010} + V_{2 k}

γ_{11 k} = δ_{110} + V_{3 k}

Results

Descriptive statistics

To depict the relationship between evaluations of coaching competency and time, we plotted coaching competence at Time 1 (pre-season) and Time 2 (post-season) for each participant (Figure 1). Additionally, we reported means, standard deviations, and the intercorrelations among variables in Table 1. The results demonstrate that transformational leadership at Time 1 was positively related to coaching competency at Time 1 (r = .83, p = .00) and coaching competency at Time 2 (r = .53, p = .00). Additionally, coaching competency at Time 1 was positively related to coaching competency at Time 2 (r = .59, p = .00).

Figure 1.

Relationship between time and athlete coaching competency. TI: Time 1, T2: Time 2; 1 = complete incompetence, 5 = complete competence.

Table 1.

Means, standard deviations, and correlations between the study variables (N = 246).

Variable	Mean	SD	1	2	3	4	5	6
1. Age	20.20	1.34	–
2. Gender	1.40	0.49	−.02	–
3. Sport year	7.52	2.66	.36**	.33**	–
4. TFL at Time 1	3.82	0.58	.08	−.09	−.02	(.94)
5. CC at Time 1	3.92	0.61	.04	−.16*	−.09	.83**	(.97)
6. CC at Time 2	3.84	0.63	.05	−.27**	−.09	.53**	.59**	(.97)

Note. Time 1: pre-season; Time 2: postseason; TFL: transformational leadership; CC: coaching competency. Player gender is coded as 1 = male and 2 = female. Cronbach’s alpha is indicated in parentheses.

*p < .05, **p < .01.

Confirmatory factor analysis

We conducted confirmatory factor analysis (CFA) to examine the measurement invariance of coaching competency (Table 2) and construct distinctiveness of transformational leadership and coaching competency (Table 3). The indicators CFI and RMSEA were reported, based on the recommendation of Williams et al.³⁹ Table 2 shows the fit results of measurement invariance models. The results indicate that the metric model fit slightly better than did the configural model and that the difference in the chi-square statistics did not achieve significance. We also test the strong and strict models, which restrict factor means and factor covariance matrices variance across times (χ² difference = 26.99, 11.46; p = .00, .04, respectively). The results of the measurement invariance test indicate that metric invariance was satisfied, supporting that the factor loadings of coaching competency are the same across times.

Table 2.

Longitudinal measurement invariance tests of coaching competency.

	df	χ²	df difference	χ² difference	NFI	CFI	NNFI	RMSEA
Configural	4	35.44			.98	.98	.95	.18
Metric	7	35.64	3	0.20	.98	.99	.98	.13
Strong	14	62.63	7	26.99**	.97	.98	.98	.12
Strict	19	74.09	5	11.46*	.96	.97	.98	.11

Note. Measurement model at time 1: df = 246, χ² = 774.90, NFI = .96, CFI = .98, NNFI = .97, RMSEA = .09; measurement model at time 2: df = 246, χ² = 641.66, NFI = .97, CFI = .98, NNFI = .98, RMSEA = .08.

*p < .05, **p < .01.

Table 3

Comparison of measurement models for transformational leadership at time 1 and coaching competency at time 1 and time 2 in the study.

Model	χ²	df	χ²/df	Δχ²	CFI	RMSEA
Time 1
Null model (all indicators are independent)	2725.09	902	3.02	–	.97	.09
Baseline model (two factors)	2656.79	901	2.80	–	.97	.09
Model 1 (one factor)	3255.99	902	3.61	599.20**	.97	.10
Time 2
Null model (all indicators are independent)	2482.80	902	2.75	–	.97	.09
Baseline model (two factors)	2416.78	901	2.68	–	.98	.08
Model 1 (one factor)	3612.76	902	4.01	1195.20**	.97	.11

Note. Model 1 has one factor; transformational leadership and coaching competency were combined into one factor.

*p < .05, **p < .01

Table 3 provides a comparison of measurement models of the main variables in the present study. The results indicate that the baseline two-factor model fit the data well at Time 1 (χ² = 2656.79, df = 901, CFI = .97, RMSEA = .09) and Time 2 (χ² = 2416.78, df = 901, CFI = .98, RMSEA = .08). We also tested the alternative model, a one-factor model that combines transformational leadership and coaching competency into one factor at Time 1 (χ² = 3255.99, df = 902, CFI = .97, RMSEA = .10) and Time 2 (χ² = 3612.76, df = 902, CFI = .97, RMSEA = .11). The CFA results show that the baseline model fit the data significantly better than did the alternative models at Time 1 (Δχ² = 599.20, p = .00) and Time 2 (Δχ² = .119520, p = .00). The CFA results indicate that the baseline model fits the data significantly better than did the alternative models at Time 1 and Time 2, supporting the construct distinctiveness of these variables.

Hypothesis testing

Table 4 provides a summary of the results that we used to test the hypothesis. We posited that coach transformational leadership interacts with the within-individual change in perceived coaching competency. This hypothesis constitutes a cross-level interaction relationship. The three-level linear growth model results revealed that the cross-level interaction was significant (γ = –0.33, p = .00). We plotted the interaction patterns and conducted simple slope tests, following the procedure recommended by Preacher et al.⁴⁰ As presented in Figure 2, coaches with high transformational leadership were associated with a decrease in perceived coaching competency from Time 1 to Time 2 (simple slope r = –.27, p = .00). In addition, coaches with low transformational leadership were associated with an increase in perceived coaching competency from Time 1 to Time 2 (simple slope r = .12, p = .05). In sum, the results for the interaction and simple slope tests support the hypothesis.

Table 4.

Three-level linear growth model results for coaching competency, time, and transformational leadership.

	DV = coaching competency at time 2
Fixed effect	df	γ
Level 1
Intercept	244	3.91** (0.03)
Time	244	−0.08 (0.05)
Level 2
Age	217	0.01 (0.02)
Athlete’s tenure with coach	217	−0.01 (0.02)
Athlete’s year of playing	217	−0.01 (0.01)
Athlete’s time on the court	217	0.00 (0.00)
TFL at Time 1	217	0.83 (0.04)**
Cross-level
TFL at Time 1 × Time	244	−0.33 (0.06)**
Level 3
Team’s winning rate	22	0.11 (0.13)

Note. Standard error is indicated in parentheses; DV: dependent variable; TFL: transformational leadership.

*p < .05, **p < .01

Figure 2.

Interaction between transformational leadership (TFL) and coaching competency.

Discussion

Our result revealed a cross-level interaction between coach transformational leadership and the within-individual change in perceived coaching competency over time. The interaction pattern indicated that coaches with high transformational leadership were evaluated as having a decrease in coaching competency from pre-season to post-season. In contrast, coaches with low transformational leadership were evaluated as having an increase in coaching competency from pre-season to post-season. Taken together, these results reflect the dynamic nature between coach transformational leadership and within-individual evaluations of perceived coaching competency over time.

This study may contribute to the understanding of proxy agency assumed by social cognitive theory¹⁰,¹¹ in sport settings. Proxy agency, or proxy efficacy, is an important concept and has been studied in many sports settings (e.g., exercise context and rehabilitation settings, see Bray et al.;⁴¹ Bray et al.;⁴² athletic dyadic settings, see Beauchamp et al.;⁴³ Jackson et al.;²⁵ Jackson et al.;⁴⁴ Jackson et al.⁴⁵). Using coaching competency along with coach transformational leadership, we studied the mechanism of proxy agency over a season. Our findings confirmed the developmental nature of coach transformational leadership and suggested different levels of coach transformational leadership had opposite directions in within-individual change of perceived coaching competency over time. We posited that these two reverse change patterns capture the dynamics of athlete competence/skill development over time and reflect athlete independence/dependence from the proxy agency (i.e., coach).

In addition, our cross-level model and empirical findings extend the work of Kao and Tsai,⁶ who found that the positive relationship between coach transformational leadership and coaching competency at a single level. Given the cross-level framework has been proposed, we were able to test the within-individual change in coaching competency. We further illustrated that a transformational coach averts athlete dependency and supports athletes toward autonomy, growth, and development. The growth of athletes over time may make them independent from their coaches and is associated with a decrease in perceived coaching competency change over time.

Academic and practical implications

In regard to the implications for coaching education,⁴⁶ we encourage coaches to understand the interaction between transformational leadership and the within-individual change in perceived coaching competency over time. Transformational coaches may enhance athletes’ development and growth (i.e., competence, confidence, connection, and character) by inspiring athletes with vision and value, working as a role model, providing opportunities for athletes’ engagements into the coaching process, and caring for each athlete with individualized support and needs. That is, transformational coaches invest extra efforts in cultivating athlete development. In return, the developed athlete competence/skills promote their independency from their coaches and is associated with a decrease in perceived coaching competency over time. Although there is a decrease in perceived coaching competency over time, coaches should attribute this decrease as a positive indicator of levels of athlete development and improvement.

In the current coach education, coaches are being encouraged to pay careful attention to maximize athlete personal development by nurturing athlete psychological qualities associated with performance enhancement.^46,47 To fulfill this developmental function, coaches with transformational leadership make extra efforts to serve as mentors who navigate athletes’ journey through care, guidance, as well as support and eventually promote personal independence of athletes. Therefore, instead of only focusing on coaching competency at a specific static high or low value, coaches could evaluate the direction of change in coaching competency over time to assess their leadership and mentorship in athlete personal development.

Limitations and recommendations for future research

Several limitations are associated with the present study. First, only athletes’ evaluations of coaching competency (i.e., proxy efficacy) were assessed; neither athletes’ self-efficacy nor collective efficacy was measured. Thus, we cannot provide a comprehensive understanding of athletes’ regulation of personal agencies. That is, the interactions among those three different agencies (i.e., self-efficacy, proxy efficacy, and collective efficacy) cannot be fully depicted. For instance, we did not determine whether a decrease in proxy efficacy leads to better development of self-efficacy. To fully understand the various relationships among different human agencies in social cognitive theory,¹⁰,¹¹ researchers are encouraged to study both athletes’ self-efficacy and collective efficacy along with their evaluations of proxy efficacy in the process of their skill development and acquisition over a season.

Second, athletes’ dependence/independence is a complicated issue (Côté et al.;⁴⁷ Fletcher and Ragins;⁴⁸ Vallée and Bloom⁴⁹), and there have been calls for further investigation. Our results specify the coach leadership/mentoring process over the phases of a game season. It is possible that athletes may become independent in regard to their performance but still be dependent on their coaches for other supports or resources (e.g., social resources). Furthermore, it is possible that athletes’ trust in their coach is an important psychological factor that influences athletes’ dependence/independence.^50,51 Future research should clarify these relationships by controlling other possible factors related to athlete dependence/independence.

Third, we measured only two time points over the game season, and this, in turn, may have limited the understanding of the dynamic changes in coach-athlete interaction (e.g., non-linear changes of evaluations of coaching competency). Although two time points changes and three-level hierarchical linear modeling has been conducted in this study (e.g., Leiter et al.⁵²), we suggest that researchers, using more data points over the game season, test the relationships in the present research and whether these relationships are found.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Ministry of Science and Technology (MOST 107-2410-H007-084) and Sue Tien Educational Fund, College of Education, National Tsing Hua University.

ORCID iD

San-Fu Kao

References

Feltz

Chase

Moritz

, et al. A conceptual model of coaching efficacy: preliminary investigation and instrument development. J Educ Psychol 1999; 91: 765–776.

Feltz

Short

Sullivan

PJ.

Self-efficacy in sport. Champaign: Human Kinetics, 2008.

Wooden

Yaeger

A game plan for life: the power of mentoring. New York: Bloomsbury, 2009.

Myers

Feltz

Maier

, et al. Athletes’ evaluations of their head coach’s coaching competency. Research Quart Exerc Sport 2006; 77: 111–121.

Myers

Beauchamp

Chase

MA.

Coaching competency and satisfaction with the coach: a multi-level structural equation model.

J Sports Scis 2011; 29: 411–422.

Kao

Tsai

CY.

Transformational leadership and athlete satisfaction: the mediating role of coaching competency. J Appl Sport Psychol 2016; 28: 469–482.

Malete

Chow

Feltz

DL.

Influence of coaching efficacy and coaching competency on athlete‐level moral variables in Botswana youth soccer. J Appl Social Psychol 2013; 43: 2107–2119.

Bosselut

Heuzé

Eys

, et al. Athletes’ perceptions of role ambiguity and coaching competency in sport teams: a multilevel analysis. J Sport Exerc Psychol 2012; 34: 345–36.

Coussens

Rees

Freeman

Applying generalizability theory to examine the antecedents of perceived coach support.

J Sport Exerc Psychol 2015; 37: 51–62.

10.

Bandura

Self-efficacy: the exercise of control. New York: Freeman, 1997.

11.

Bandura

Social cognitive theory: an agentic perspective.

Annu Rev Psychol 2001; 52: 1–26.

12.

Bormann

Schulte-Coerne

Diebig

, et al. Athlete characteristics and team competitive performance as moderators for the relationship between coach transformational leadership and athlete performance. J Sport Exerc Psychol 2016; 38: 268–281.

13.

Hodge

Henry

Smith

A case study of excellence in elite sport: motivational climate in a world champion team. Sport Psychol 2014; 28: 60–74.

14.

Turnnidge

Côté

Applying transformational leadership theory to coaching research in youth sport: a systematic literature review. Int J Sport Exerc Psych 2018; 16: 327–342.

15.

Bass

BM.

Leadership and performance beyond expectations. New York: Free Press, 1985.

16.

Bass

BM.

From transactional to transformational leadership: learning to share the vision. Organ Dyn 1990; 18: 19–32.

17.

Smith

Young

Figgins

, et al. Transformational leadership in elite sport: a qualitative analysis of effective leadership behaviors in cricket. Sport Psychol 2017; 31: 1–15.

18.

Scandura

Williams

EA.

Mentoring and transformational leadership: the role of supervisory career mentoring. J Vocational Behav 2004; 65: 448–468.

19.

Sosik

Godshalk

VM.

Leadership styles, mentoring functions received, and job-related stress: a conceptual model and preliminary study. J Organ Behav 2000; 21: 365–390.

20.

Sosik

Godshalk

Yammarino

FJ.

Transformational leadership, learning goal orientation, and expectations for career success in mentor-protégé relationships: a multiple levels of analysis perspective. Leadership Quart 2004; 15: 241–261.

21.

Kao

Watson

JC.

A multilevel study of transformational leadership and motivational climates in university basketball teams. Int J Sport Psychol 2017; 48: 50–69.

22.

Lefebvre

Turnnidge

Côté

A systematic observation of coach leadership behaviors in youth sport. J Appl Sport Psychol. Epub ahead of print 2019. DOI: 10.1080/10413200.2019.1609620.

23.

International Council for Coaching Excellence and the Association of Summer Olympic International Federations. International sport coaching framework version 1.1. Champaign: Human Kinetics, 2012.

24.

Bray

Shields

Jackson

, et al. Proxy agency and other-efficacy in physical activity contexts. In: Beauchamp

Eys

(eds) Group dynamics in exercise and sport psychology. 2nd ed. New York: Routledge, 2014, pp. 91–109.

25.

Jackson

Bray

Beauchamp

, et al. The tripartite efficacy framework in physical activity contexts. In: Mellalieu

Hanton

(eds) Contemporary advances in sport psychology: a review. London: Taylor & Francis Group, 2015, pp. 120–147.

26.

Bass

Riggio

RE.

Transformational leadership (2nd ed.). Mahwah: Lawrence Erlbaum, 2006.

27.

Kao

Tsai

Schinke

, et al. A cross-level moderating effect of team trust on the relationship between transformational leadership and cohesion. J Sport Sci 2019; 37: 2844–2852.

28.

Price

Weiss

MR.

Relationships among coach leadership, peer leadership, and adolescent athletes’ psychosocial and team outcomes: a test of transformational leadership theory. J Appl Sport Psych 2013; 25: 265–279.

29.

Rowold

Transformational and transactional leadership in martial arts. J Appl Sport Psychol 2006; 18: 312–325.

30.

Bass

Avolio

BJ.

Full range of leadership: manual for the multi-factor leadership questionnaire. Palto Alto: Mind Garden, 1997.

31.

Scandura

Schriesheim

CA.

Leader-member exchange and supervisor career mentoring as complementary constructs in leadership research. Acad Manage J 1994; 37: 1588–1602.

32.

Kram

KE.

Mentoring at work: developmental relationships in organizational life. Glenview: Scott Foresman, 1985.

33.

Chelladurai

Leadership in sports. In: Tenenbaum

Eklund

(eds) Handbook of sport psychology. 3rd ed. Hoboken: Wiley, 2007, pp. 113–135.

34.

Bass

Avolio

BJ.

Manual for the multifactor leadership questionnaire (MLQ-Form 5X). Redwood City: Mind Garden, 2004.

35.

Brislin

RW.

Translation and content analysis of oral and written materials. In: Triandis

Berry

(eds) Handbook of cross-cultural psychology: vol. 2. Methodology. Boston: Allyn and Bacon, 1980, pp. 389–444.

36.

Stenling

Tafvelin

Transformational leadership and well-being in sports: the mediating role of need satisfaction. J Appl Sport Psychol 2014; 26: 182–196.

37.

Bliese

Ployhart

RE.

Growth modeling using random coefficient models: model building, testing, and illustrations. Organ Res Methods 2002; 5: 362–387.

38.

Bliese

Multilevel: multilevel functions (R package version 2.6), https://CRAN.R-project.org/package=multilevel (2016, accessed 13 August 2020).

39.

Williams

Vandenberg

Edwards

JR.

12 structural equation modeling in management research: a guide for improved analysis. Acad Manage Annals 2009; 3: 543–604.

40.

Preacher

Curran

Bauer

DJ.

Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. J Educ Behaval Stat 2006; 31: 437–448.

41.

Bray

Gyurcsik

Culos-Reed

, et al. An exploratory investigation of the relationship between proxy efficacy, self-efficacy and exercise attendance. J Health Psychol 2001; 6: 425–434.

42.

Bray

Brawley

Millen

JA.

Relationship of proxy efficacy and reliance to home-based physical activity after cardiac rehabilitation. Rehabil Psychol 2006; 51: 224–231.

43.

Beauchamp

Jackson

Loughead

TM.

Leadership in physical activity contexts. In: Horn

Smith

(eds) Advances in sport and exercise psychology. 4th ed. Champaign: Human Kinetics, 2019, pp. 151–170.

44.

Jackson

Gucciardi

Lonsdale

, et al. “I think they believe in me”: the predictive effects of teammate- and classmate-focused relation-inferred self-efficacy in sport and physical activity settings. J Sport Exerc Psychol 2014; 36: 486–505.

45.

Jackson

Knapp

Beauchamp

MR.

The coach-athlete relationship: a tripartite efficacy perspective. Sport Psychol 2009; 23: 203–232.

46.

Turnnidge

Côté

Transformational coaching workshop: applying a person-centred approach to coach development programs. Int Sport Coaching J 2017; 4: 314–325.

47.

Côté

Young

North

, et al. Towards a definition of excellence in sport coaching. Int J Coaching Sci 2007; 1: 3–17.

48.

Fletcher

Ragins

BR.

Stone center relational cultural theory. In: Bagins

Kram

(eds) The handbook of mentoring at work: theory, research, and practice. London: Sage, 2007, pp. 373–399.

49.

Vallée

Bloom

GA.

Building a successful university program: key and common elements of expert coaches. J Appl Sport Psychol 2005; 17: 179–196.

50.

Zhang

Chelladurai

Antecedents and consequences of athlete's trust in the coach. J Sport Health Sci 2013; 2:115–121.

51.

Jowett

At the heart of effective sport leadership lies the dyadic coach-athlete relationship. Sport Exerc Psychol Rev 2017; 13: 62–64.

52.

Leiter

Laschinger

HKS

Day

, et al. The impact of civility interventions on employee social behavior, distress, and attitudes. J Appl Psychol 2011; 96: 1258–1274.