Motivation is an important topic for adherence towards sport and consequently for health, in young adults. Thus, knowing more about the antecedents of adaptive and maladaptive forms of motivation is important. Specifically, the possible nesting effect of the motivational climate at the team level is of interest and becomes the major contribution of this paper, as it has not been studied in previous literature. Concretely, this study aimed to test a multilevel mediation model which examined the relationships between the perceived motivational climate created by coaches at team level and motivational regulations toward sport at individual level, as mediated by individual goal orientations.
Within the theoretical frameworks of achievement goal theory (AGT; Ames, Reference Ames1992; Nicholls, Reference Nicholls, Ames and Ames1984; Reference Nicholls1989) and self-determination theory (SDT; Deci & Ryan, Reference Deci and Ryan1987; Reference Deci and Ryan2000), it has been established that both social (e.g., motivational climate) and personal (e.g., goal orientations or dispositional goals) variables act as determinants of intrinsic motivation. On one hand the AGT (Ames, Reference Ames1992; Nicholls, Reference Nicholls, Ames and Ames1984; Reference Nicholls1989) posits that in achievement contexts such as sport, individuals develop a series of behavioral and cognitive processes to achieve their goals and demonstrate competence; moreover, this competence or skill is defined differently depending on developmental and individual differences that are influenced by dispositional and situational factors. Thus, two of AGT’s main constructs are goal orientation and motivational climate.
Goal orientation – the manner in which individuals define competence – can be self-referenced (task orientation) or referenced by others (ego orientation). In this sense, individuals who are predominantly task-oriented feel successful by learning something new or improving their skills, whereas ego-oriented individuals feel successful by attaining superiority over others (Nicholls, Reference Nicholls1989).
Biddle (Reference Biddle and Roberts2001) posited that individual differences regarding goal orientation might be influenced by the environment. In studying the environment, AGT incorporates the motivational climate, which includes those situational goals that are set in an achievement context (Ames, Reference Ames1992) and that are created by others who are significant to the person. Several authors have considered that the coach of a sport team is one of the most influential people in athletes’ sporting experience based in the environment that he or she creates within the team (e.g., Bartholomew, Ntoumanis, & Thogersen-Ntoumani, Reference Bartholomew, Ntoumanis and Thøgersen-Ntoumani2010).
According to AGT, there are two types of motivational climates known as task-involving and ego-involving climates. A task-involving climate is distinguished by private recognition based on one’s own progress and an assessment based on individual improvement. This approach considers that mistakes are a natural part of the improvement process. An ego-involving climate includes public recognition that depends on social comparisons as the basis on which to judge success and considers mistakes as something to avoid (Ames, Reference Ames and Roberts1995).
Previous studies have tested the relationship between motivational climates and goal orientations in sport, considering motivational climates at the individual level. Empirical evidence has shown that task orientation is positively associated with a task-involving climate (e.g., Moreno, Cervelló, & González-Cutré, Reference Moreno, Cervelló and González-Cutré2007; Reference Moreno, Cervelló and González-Cutré2008; Pineda, López-Walle, & Medina, Reference Pineda, López-Walle and Medina2011; Reyes, Reference Reyes2009; Sánchez et al., Reference Sánchez, Leo, Gómez, Sánchez, Cruz and García2009) and that an ego-involving climate is positively related to ego orientation (e.g., Balaguer, Castillo, Duda, & García-Merita, Reference Balaguer, Castillo, Duda and García-Merita2011; Moreno et al., Reference Moreno, Cervelló and González-Cutré2007; Reference Moreno, Cervelló and González-Cutré2008). Although other studies have not supported that last association (e.g., Pineda et al., Reference Pineda, López-Walle and Medina2011; Reyes, Reference Reyes2009; Sánchez et al., Reference Sánchez, Leo, Gómez, Sánchez, Cruz and García2009).
On the one hand, AGT examines how perceptions of motivational climate created by significant others (e.g., coach) interact with the dispositional goals to influence behavior in achievement contexts, so the goals are developmentally acquired through the social learning and identification process; on the other hand, the SDT analyzes how social conditions address the inherent self-actualization tendencies that either support or thwart behavioral internalization. Both theories have very different philosophical starting points (e.g., social-cognitive vs. organismic), however Ryan and Deci (Reference Ryan and Deci1989) argued that each theory focuses on distinct bodies of ideas and insights that can be considered complementary rather than contradictory. So the frameworks can be used together to extract additional information regarding motivational processes, although they cannot in themselves be integrated for the same ends. Following this recommendation, recent empirical research has combined AGT and SDT to examine motivational environments in relation to athlete motivation (Smith et al., Reference Smith, Tessier, Tzioumakis, Fabra, Quested, Appleton and Duda.2016) and physical activity (Fenton, Duda, Appleton, & Barrett, Reference Fenton, Duda, Appleton and Barrett2016).
With regard to motivation, SDT considers that the basic dichotomy of intrinsic/extrinsic motivation is insufficient, based on the level of self-determined behaviour. Moreover, SDT suggests that intrinsic motivation is not the only case of self-determined activity and that extrinsically motivated behavior can be either autonomous (self-determined) or controlled (Deci & Ryan, Reference Deci and Ryan1987). Deci and Ryan (Reference Deci and Ryan2000) suggested a continuum of four extrinsic motivational regulations with a varying degree of determination that would involve the following (from highest to lowest self-determination): 1) integrated regulation, when the behavior is consistent with other values and the needs of the individual; 2) identified regulation, where the person identifies with an activity and gives value to it; 3) introjected regulation, a regulatory process within the person that is experienced as a demand or pressure; and finally, 4) external regulation, which refers to those behaviors controlled by external sources. Based on this reasoning, Deci and Ryan (Reference Deci, Ryan and Dienstbier1991) suggested that integrated and identified regulations can be considered as autonomous forms of motivation; meanwhile introjected and external regulations can be considered as controlled ones.
Autonomous motivation is perceived as intentional behavior that is based on the values and interests of a particular individual who experiences volition for actions, selects the results to pursue and chooses how to achieve them; it is also characterized by integration and the absence of pressure and conflict (Deci & Ryan, Reference Deci, Ryan and Dienstbier1991). Therefore, autonomous motivation is composed by the combination of intrinsic motivation, integrated regulation, and identified regulation. By contrast, in controlled motivation, internal and/or external pressure and external control guide behavior (Deci & Ryan, Reference Deci, Ryan and Dienstbier1991); thus, it is composed by the combination of introjected regulation and extrinsic motivation. Finally, amotivation refers to the lack of intention to act.
According to Duda (Reference Duda and Roberts2001), motivational responses in sport are shaped by the coach’s creation of a particular motivational climate. In addition, dispositional factors are also important to the study of self-motivation. Personal processes such as goal orientations are included within dispositional factors (Deci & Ryan, Reference Deci and Ryan1987). A specific line of research in AGT has examined the relationship of goal orientation and/or perceptions of the motivational climate, with motivation (Standage, Duda, & Ntoumanis, Reference Standage, Duda and Ntoumanis2003).
In the physical education context, Brunel (Reference Brunel1999) found that task orientation is positively related with intrinsic motivation – whereas ego orientation is positively related with external and introjected regulations (forms of controlled motivation) – and that a task-involving climate is positively correlated with autonomous forms of motivation. Other studies have shown that task orientation positively predicts autonomous forms of motivation (intrinsic motivation and identified regulation) (e.g., Ntoumanis, Reference Ntoumanis2001; Standage et al., Reference Standage, Duda and Ntoumanis2003) and negatively predicts amotivation (e.g., Standage et al., Reference Standage, Duda and Ntoumanis2003), whereas ego orientation is positively related with intrinsic motivation (e.g., Standage et al., Reference Standage, Duda and Ntoumanis2003). In the sport context, positive relationship has been found between ego-involving climate and identified regulation (autonomous form of motivation) (Moreno et al., Reference Moreno, Cervelló and González-Cutré2007), and proxy variables of autonomous motivation as dedication (Curran, Hill, Hall, & Jowett, Reference Curran, Hill, Hall and Jowett2015). Meanwhile, the task-involving climate was positively related to autonomous forms of motivation (Smith, Cumming, & Smoll, Reference Smith, Cumming and Smoll2008), and proxy variables as dedication, enthusiasm (Curran et al., Reference Curran, Hill, Hall and Jowett2015), and enjoyment (Jaakkola, Ntoumanis, & Liukkonen, Reference Jaakkola, Ntoumanis and Liukkonen2016).
Newton and Duda (Reference Newton and Duda1999) reported that sport research has traditionally used regression analysis to examine the interaction between goal orientations and motivational climates (at the individual level) and their further effects on behavioral variables. In this line, previous research has examined the independent contribution of motivational climate – or the interaction between goal orientation and motivational climate – to explain autonomous motivation, operationalizing motivational climate at the individual level (the level of the athlete’s perception of the climate created by the coach), but not at the team (the level of the perception of the team). The individual approach helps to understand how the athletes’ individual perception of interpersonal coaching styles influences motivational regulations toward sport. However, we cannot directly assume that the relationship that has been found at the individual level will also take place at the team level (Duda, Reference Duda and Roberts2001). It is therefore necessary to study the nesting effect of the motivational climate at the team level, as it has not been addressed in the previous literature. The team level approach attempts to understand how a specific climate created in a team and perceived and shared by its different players influences motivational regulations. In this regard, it is important to remark that climate can be defined as the shared perceptions of members of a team or group (Anderson & West, Reference Anderson and West1998). Testing motivational climate at the team level implies considering the fact that athletes belonging to the same team have the same coach, so they will have similar situational experiences that will be distinct from the experiences of other teams. Thus, testing motivational climate at the team level allows capturing how belonging to a team influences social and cognitive factors. To address these issues, Raudenbush and Bryk (Reference Raudenbush and Bryk2002) proposed the use of hierarchical linear models to help assess the effects of both individual and team variables.
Therefore, the present study aimed to test a multilevel mediation model that analyzed the effects of contextual (perceived motivational climate created by coaches at team level) and dispositional (goal orientations at individual level) factors on motivational regulations towards sport at individual level. Concretely, the model tested the mediator role of individual goal orientations (task and ego) in the relationship between team motivational climate (task-involving and ego-involving) and individual motivational regulations towards sport (autonomous, controlled and amotivation). According to previous literature that has analyzed relationship at the individual level, we expected to find empirical support for the hypothesized relationships in the proposed model (see Figure 1).
Figure 1. Hypothesized multilevel mediation model.
Note: The solid lines express positive relationships and the dashed lines express negative relationships.
Method
Participants
The study sample included 211 Mexican college athletes of both genders (115 women and 96 men), belonging to 20 teams training in different sporting disciplines (e.g., athletics, basketball, handball, baseball, football, weightlifting, swimming). Participants aged between 17 and 28 years old (M = 19.97 years, SD = 2.01), with an average time spent practicing the sport of 6.3 years (SD = 3.02), and an average seniority with the team coach of 2.1 years (SD = 1.01).
Instruments
The perceived motivational climate created by the coach was measured using the Perceived Motivational Climate in Sport Questionnaire (PMCSQ-2), as adapted to the Mexican context (López-Walle, Balaguer, Castillo, & Tristán, Reference López-Walle, Balaguer, Castillo and Tristán2011). This questionnaire consists of 24 items, 11 of which measure the perceived task-involving climate (e.g., “Athletes help one another learn”), and the other 13 measure the perceived ego-involving climate (e.g., “The coach only congratulates athletes when they stand out from others”). The questions were answered using a five-point Likert-type scale ranging from never (1) to always (5).
Individual trends on goal orientations in sport were measured with the Task and Ego Orientation in Sport Questionnaire (TEOSQ), as adapted to the Mexican context (López-Walle, Balaguer, Meliá, Castillo, & Tristán, Reference López-Walle, Balaguer, Castillo and Tristán2011). This questionnaire consists of 13 items and is divided into two scales that measure task orientation using seven items (e.g., “I am more successful in my sport when I learn a new exercise, and it makes me want to practice more”) and ego orientation using six items (e.g., “I am more successful in my sport when I am the only one who is able to do the play or who has that skill”). The questions were answered using a five-point Likert-type scale ranging from strongly disagree (1) to strongly agree (5).
Motivational regulations were measured with the Spanish version (Balaguer, Castillo, & Duda, Reference Balaguer, Castillo and Duda2007) of the Sport Motivation Scale (SMS), which consists of 28 items that can be grouped into three factors, consistent with SDT (Deci & Ryan, Reference Deci and Ryan2000) and with previous studies (e.g., Vansteenkiste, Zhou, Lens, & Soenens, Reference Vansteenkiste, Zhou, Lens and Soenens2005). Thus, autonomous motivation was assessed by asking athletes to indicate whether they practiced sport for intrinsic (e.g., “For the pleasure of learning new training techniques”) or identified reasons (e.g., “Because it is one of the best ways of developing other aspects of myself”). To assess controlled motivation, athletes were asked to indicate whether they practiced sport for introjected (e.g., “Because I need to engage in sport to feel good about myself”) or external reasons (e.g., “Because of the prestige of being an athlete”). A similar procedure was followed to assess amotivation (e.g., “I’m not sure, but I feel that I am unable be successful in this sport”). The questions were answered using a seven-point Likert-type scale ranging from does not correspond at all (1) to corresponds exactly (7). In this investigation, autonomous motivation was the average resulting from intrinsic motivation and identified regulation, and controlled motivation was the average resulting from external and introjected regulation.
Procedure
The present research was conducted in accordance with international ethical guidelines that are consistent with American Psychological Association (APA) guidelines. Ethical approval for the study was obtained from a university ethics review committee. Personal contact was first made with team coaches to inform them about the project. Instruments were applied during a training session in which the main researcher was present to give instructions and answer athletes’ questions. Emphasis was placed on the confidentiality of athletes’ individual responses as well as on the need to answer honestly. In addition, the researcher explained to them that responding to the questionnaires implied voluntarily accepting to participate in the research.
Data analysis
Descriptive statistics (mean and standard deviation) and scale reliability (Cronbach’s α coefficient) were also calculated for all variables employed in the study using SPSS 19.0. The hypothesized relationships in the proposed multilevel mediation model were estimated using hierarchical linear modeling (HLM; Raudenbush & Bryk, Reference Raudenbush and Bryk2002) with the PRELIS application of LISREL 8.80. This analysis strategy assumes a hierarchical structure for the data, it allows for the consideration of the influence of variables operationalized at the team-level (level 2) on variables at the individual level (level 1). In the sport context, athletes belong to different teams, and a different coach leads each team. This grouping tends to promote uniformity within the team, which naturally violates the assumption that data corresponding to athletes constitute independent observations. The multilevel hierarchical analysis allows for consideration of nested structure in connection with the data collected from athletes from multiple teams.
Before testing the hypothesized multilevel mediation model, aggregation of team members’ scores for task-involving climate and ego-involving climate was justified in SPSS. Therefore, it was necessary to assess within-team agreement, and between-team discrimination (Chan, Reference Chan1998). Within-team agreement was assessed using the average deviation index (ADI; Dunlap, Burke, & Smith-Crowe, Reference Dunlap, Burke and Smith-Crowe2003), taking the criterion established by Dunlap and colleagues (2003) of ADI < c/6 (where c is the number of categories in the response scale) as a reference for interpretation. In addition, analysis of variance (ANOVA) was used to test between-team discrimination on perceived task-involving and ego-involving climates.
Following the recommendations discussed in the previous literature (e.g., Zhang, Zyphur, & Preacher, Reference Zhang, Zyphur and Preacher2009), the independent variables were centered. Individual variables (task orientation and ego orientation) were centered at their group mean; team-level variables (task-involving climate and ego-involving climate) were centered at their grand mean.
According to Zhang and colleagues (Reference Zhang, Zyphur and Preacher2009), the model tested in the present research (see Figure 1) would be a 2–1–1 multilevel mediation model, in which a team-level variable influences an individual-level variable, which in turn has an effect on another variable at the individual level. The sequence of hierarchical linear models used to estimate the relationships hypothesized in the model is detailed below: (1) one-way ANOVA models (baseline model with random intercepts) to estimate intra-team and inter-team variance for the dependent variables at the individual level (goal orientations and motivational regulations), calculating at the same time the intraclass correlation coefficient (ICC). The presence of inter-team differences validates the inference that the data present a hierarchical structure, thus pointing to the logic in developing hierarchical linear models (Heck & Thomas, Reference Heck and Thomas2000). (2) Intercepts-as-outcome models to test the cross-level effects of team-level predictors (task-involving climate and ego-involving climate) on mediator variables (task orientation and ego orientation) at the individual level. (3) Random regression coefficients models to estimate the effects of individual level predictors (task orientation and ego orientation) on outcome variables at the individual level (autonomous motivation, controlled motivation and amotivation). This model allows to estimate variances for intercepts and slopes of regression across teams and to estimate the proportion of variance explained by predictors at the individual level. Finally, (4) when the results of the previous analysis indicated a multilevel mediation effect, an additional intercepts-as-outcome model was tested. In this model, the effect of the team-level predictor (X) on the outcome variable (Y) was included, in addition to the effect of the individual predictor (the mediating variable, M) on the corresponding outcome variable (Y). This last effect represents the direct or nonmediated effect (τ) of X on Y, thus offering information regarding whether there was total or partial mediation.
Finally, the product-of-coefficients test proposed by Taylor, Mackinnon, and Tein (Reference Taylor, MacKinnon and Tein2008) was used to confirm the detected effects of mediation. The effect of mediation is estimated using the (αβ) product, where (α) represents the effect of the team-level predictor (X) on the mediating variable (M) or individual-level predictor, and (β) represents the effect of the mediating variable (M) on the outcome variable at the individual level (Y), controlling for the effect of X on Y. A 95% confidence interval for the mediation effect (αβ) can be estimated by adding or subtracting 1.96 times the standard error (where 1.96 is the critical value for the normal distribution, and the standard error was the Sobel (Reference Sobel1982) multivariate delta standard error). If the estimated confidence interval does not include zero, the mediation effect is confirmed.
Results
Descriptive statistics
The descriptive statistics, reliability (Cronbach’s α), and the correlations between the study variables are shown in Table 1. Internal consistency in all the scales was acceptable as Cronbach’s alpha was above the criterion of .70 determined for psychological scales (Nunnally, Reference Nunnally1978). Motivational climates were positively related to goal orientations in their respective dimensions. Autonomous motivation was positively correlated with task-involving climate and task orientation. Amotivation was negatively related to task-involving climate and goal orientation. Additionally, autonomous motivation was positively related to controlled motivation.
Table 1. Descriptive statistics, Pearson product-moment correlation coefficient between study variables and the reliability of measurement scales
Note: *p < .05; **p < .01. The value in parentheses represents the Cronbach’s alpha coefficient for each scale.
Multilevel analysis
Before aggregating team members’ scores for the team level variables, within-team agreement and between team discrimination was tested. The average ADI values for task-involving and ego-involving climates were .60 (SD = 0.15) and .72 (SD = 0.16), respectively. Both values were below the criterion of .83 for Likert scales with 5 response categories (5/6 = .83), which suggested the existence of shared perceptions within the team regarding task-involving and ego-involving climates. Additionally, ANOVA results indicated adequate discrimination between teams for both task-involving (F(19, 191) = 2.69, p < .01) and ego-involving climates (F(19, 191) = 2.24, p < .01). Based on these results, we concluded that aggregating team members’ scores for task-involving and ego-involving climates was justified.
Regarding HLM results, one-way ANOVA models carried out showed no significant between-groups differences for task orientation (τ00 = .44, p = .36; ICC = .02) and ego orientation (τ00 = .67, p = .30; ICC = .04). For this reason, additional analyses with SPSS were run to explore variance among teams for the goal orientations measures. The ANOVA results showed statistically significant differences between teams for ego orientation (F(19, 191) = 1.66, p = .04), and marginally significant differences for task orientation (F(19, 191) = 1.50, p = .08). Based on these results, we decided to keep both goal orientation dimensions and continue with the next multilevel model. The ICC values for task orientation and ego orientation indicated that 2% and 4% of the variance in these variables was due to differences between teams, respectively.
The intercepts-as-outcome models introduced the perceived motivational climate (task-involving and ego-involving climate) as predictor of the corresponding individual-level goal orientation (task orientation and ego orientation). Results showed that only task-involving climate at team-level was significantly associated with individual task orientation (γ01 = .77, p < .001).
One-way ANOVA models for motivational regulations did not exhibit significant differences for autonomous motivation (τ00 = .97, p = .32; ICC = .05), controlled motivation (τ00 = 1.67, p = .29; ICC =.02) or amotivation (τ00 = 1.99, p = .35; ICC = .02). The random regression coefficients models showed that task orientation positively predicted autonomous motivation (γ10 =.51, p < .01) and negatively predicted amotivation (γ10 = –.76, p < .001). The intercepts-as-outcome models – nested in the previous models after controlling for the team-level effects of motivational climate on motivational regulations – showed that task-involving climate had a positive effect on autonomous motivation (γ01 = .68, p = .03). Figure 2 displays the relationships and parameters that were shown to be significant. Table 2 displays all parameters for the different tested models.
Figure 2. Multilevel mediation model of the relationships between motivational climate (team level), goal orientations and motivational regulations (individual level).
Note: ***p < .001; **p < .01; *p < .05.
Table 2. Multilevel hierarchical analysis: motivational climate, goal orientations and motivational regulations
Note: ***p < .001; **p < .01; *p < .05. SE = standard error. Variables in bold letters in the first column represent de dependent variable for each model.
Based on the previous results, two multilevel mediation effects were detected. Task orientation partially mediated the relationship between team perceptions of task-involving climate and individual autonomous motivation (b1b2 = .39; 95% CI = [.11, .68]; τ = .68, p < .05), and totally mediated the relationship between team perceptions of task-involving climate and individual amotivation (b1b2 = –.58; 95% CI = [–.92, –.25]; τ = –.62, p > .05).
Discussion
In this study, we tested the effects of team perceptions of motivational climates (task-involving and ego-involving) on athletes’ individual goal orientations (task and ego), and in turn on their individual motivational regulations (autonomous, controlled and amotivation) using hierarchical linear modeling analysis. The adding value of the study was testing motivational climate at the team level which allowed us to analyze the top-down influence of team-level constructs (perceived motivational climate created by coaches) on individual-level constructs (goal orientations and motivational regulations towards sport) and also examine multilevel relationships.
Results partially confirm the relationships hypothesized in the proposed multilevel mediation model (see Figure 1). Team perceived task-involving climate created by the coach had a positive direct effect on autonomous motivation for sport practice, as well as a positive indirect effect through task orientation. In other words, task orientation partially mediated the relationship between the team perceived task-involving climate created by the coach and autonomous motivation towards sport. Also, task orientation completely mediated the relationship between task-involving climate and amotivation towards sport.
According to the theoretical assumptions of the AGT (Ames, Reference Ames1992; Nicholls, Reference Nicholls, Ames and Ames1984; Reference Nicholls1989), the above results suggest that when members of a sport team perceive that the coach emphasizes learning, emphasizes improving and mastering of exercises, considers mistakes to be part of the learning process, and appreciates their effort, they will feel successful (competent or skilled) in their sport when learning or mastering an exercise. All the aforementioned is in line with results of previous studies that analyzed the relationship between climate at the individual level and goal orientations (e.g., Moreno et al., Reference Moreno, Cervelló and González-Cutré2007; Reference Moreno, Cervelló and González-Cutré2008; Pineda et al., Reference Pineda, López-Walle and Medina2011; Reyes, Reference Reyes2009; Sánchez et al., Reference Sánchez, Leo, Gómez, Sánchez, Cruz and García2009).
According to the principles of the SDT (Deci & Ryan, Reference Deci and Ryan1987; Reference Deci and Ryan2000), the results of this study suggest that when athletes judge their competence using self-referenced criteria, they participate in their sport because they have freely chosen to do so – based on their interest in the activity itself, the enjoyment triggered by training, or because they identify with the sport –by giving it value or importance. This is consistent with the empirical results of Deci and Ryan (Reference Deci and Ryan2000) and Reeve (Reference Reeve1989), which found that task orientation is positively correlated with autonomous motivation. The explanation for this is that task orientation involves a less external or evaluative perspective that allows individuals to focus on an activity to improve tasks and acquire mastery (Deci & Ryan, Reference Deci and Ryan2000; Nicholls, Reference Nicholls, Ames and Ames1984); furthermore task orientation is experienced as an end in itself, therefore, it is more likely to be regulated by self-determined reasons (Deci & Ryan, Reference Deci and Ryan1987). All that supports the suggestion made by Duda (Reference Duda and Roberts2001) that subjects’ usage of criteria under their personal control contributes to developing a sense of autonomy.
Additionally, results of this study indicate that when the team perceives a task-involving climate generated by the coach, that facilitates athletes defining themselves as competent individuals when they are learning or mastering an exercise, and additionally, that will prevent them from attending training with no motivation. It should be noted that the effect of task-involving climate on amotivation is in accordance with results of previous studies in sport and physical education contexts in which motivational climate has been considered at the individual level (e.g., Balaguer et al., Reference Balaguer, Castillo, Duda and García-Merita2011; Moreno et al., Reference Moreno, Cervelló and González-Cutré2007; Standage et al, Reference Standage, Duda and Ntoumanis2003).
Also, when the team perceive that the coach appreciates the athletes’ efforts and focuses on learning the exercises, this directly facilitates that athletes perceive themselves as initiators of their own conduct and responsible for it; in this manner, athletes experience the volition to act. Thus, sport participants find the activity interesting and enjoy it, which eliminates the need to work out based on extrinsic reasons. This result ensues because the task-involving climate is less coercive and favors athletes’ active involvement in the training process, and awareness of their own performance. This finding offers empirical support in the sport context to the notion proposed by Deci and Ryan (Reference Deci and Ryan1987) that autonomous motivation must come from oneself and is therefore facilitated only by contextual events (situational variables). Also reinforces the results of approximations of previous studies at the sport and physical education contexts that measured motivational climate at the individual level (e.g., Balaguer et al., Reference Balaguer, Castillo, Duda and García-Merita2011; Brunel, Reference Brunel1999; Curran et al., Reference Curran, Hill, Hall and Jowett2015; Jaakkola et al., Reference Jaakkola, Ntoumanis and Liukkonen2016; Smith et al., Reference Smith, Cumming and Smoll2008; Standage et al., Reference Standage, Duda and Ntoumanis2003).
However, our results do not support the relationship between team perceptions regarding the ego-involving climate created by a coach and athletes’ ego orientation. This result contradicts empirical evidence from previous studies that have analyzed the motivational climate perceived by athletes at the individual level (e.g., Balaguer et al., Reference Balaguer, Castillo, Duda and García-Merita2011; Curran et al., Reference Curran, Hill, Hall and Jowett2015; Moreno et al., Reference Moreno, Cervelló and González-Cutré2007; Reference Moreno, Cervelló and González-Cutré2008), but is consistent with other studies that have also evaluated the climate at the individual level (e.g., Pineda et al., Reference Pineda, López-Walle and Medina2011; Reyes, Reference Reyes2009; Sánchez et al., Reference Sánchez, Leo, Gómez, Sánchez, Cruz and García2009). This finding also suggests that relationships built at the individual level do not automatically emerge at the team-level.
The multilevel mediation approach employed in this study enabled to consider the group nature of the motivational climate created by a coach on a sport team, which has led to a new understanding of the predictors of motivational regulations in sport and to the effects of situational predictors and dispositional factors, in particular. Although past studies had evaluated the effects of the social context on motivational regulations through motivational climate (e.g., Brunel, Reference Brunel1999; Moreno et al., Reference Moreno, Cervelló and González-Cutré2007; Ntoumanis, Reference Ntoumanis2001; Standage et al., Reference Standage, Duda and Ntoumanis2003), these evaluations were undertaken on the basis of the perceptions of each athlete separately. The present work replicates these results, but considering climate as shared perceptions of the members of a team.
Among the findings of this study, it is showed that autonomous motivation and controlled motivation are positively related. In this regard, Judge, Bono, Erez, and Locke (Reference Judge, Bono, Erez and Locke2005) indicated that both motivations do not seem to be negatively related. Moreover, recent research has highlighted that individuals can report high levels of both motivation types (e.g., Healy, Ntoumanis, & Duda, Reference Healy, Ntoumanis and Duda2016; Langan et al., 2015), which is in line with the results of the present study.
As theoretical implications, this paper gives another methodological approach in favor of better understanding of the reality of the phenomenon, bearing in mind that athletes belong to teams and are trained by the same coach, and that teams may provide athletes with unique experiences linked to schemes, training approaches and social norms. From the practical point of view, the results suggest that for coaches to encourage training for volitional reasons – without resorting to imputing feelings of obligation and incompetence – it is necessary to plan a variety of exercises or tasks for training sessions that encourage teamwork during training sessions and that consider mistakes as part of the learning process. In addition, when evaluating athletes, their learning, effort and mastery of technical skills and tactics should be considered and this assessment should be based on the improvement of personal marks and on relying on objectives that respect the principle of individuality of training.
Simultaneously, there are some limitations concerning the size and specific characteristics of the sample used in this study. First, we are aware about the small sample size (20 teams). Furthermore, we also assume that the results of the study can be influenced by the competitive level of the sample, and thus we must be cautious about its generalizability to other samples of athletes. However, these limitations indicate possible directions for future research. There is no doubt that continuous efforts are required to engage in research that adopts the SDT approach to deepen base knowledge regarding athletes’ social experiences, so that we encourage developing future research on this topic using study samples with different competition levels and greater number of teams.
Another important limitation of the study points to its design, as the use of a cross-sectional design does not enable to arrive to strong conclusions on the temporal ordering of the variables. With this in mind, it is important that future research adopt a longitudinal data approach, in light of the possibility that athlete motivation influences coach behaviors. Longitudinal designs are presumed to offer good opportunities to add further to our understanding for the causal process in the phenomena of interest.
In conclusion, this study extends the existing sport-scientific literature and takes into consideration both the individual and the team perspectives. The present research has provided a new perspective to better understand the contextual and dispositional factors that predict motivational regulations from a different methodological approach. This study adds to previous studies that have employed a traditional methodological approach, deepening the understanding of the influence of team level situational and individual dispositional factors on motivational regulations in sport. Therefore, when analyzing the motivational climate from a multilevel perspective, the relationship between the motivational climate and goal orientations is replicated in task dimensions, but not in ego dimensions. Furthermore, empirical evidence has been provided which indicates that task orientation has a partial mediating role between task-involving climate and autonomous motivation towards sport practice, and a total mediating role in the relationship between task-involving climate and amotivation.
We thank the Universitat de Valencia, particularly the Unidad de Investigación de Psicología del Deporte [Research Unit on Sport Psychology], directed by Dr. Isabel Balaguer, who is co-responsible for the project that supports this article, and we also thank the Instituto Universitario de Investigación en Psicología de los Recursos Humanos, del Desarrollo Organizacional y de la Calidad de Vida Laboral [University Research Institute on Psychology of Human Resources, Organizational Development and Quality of Working Life] (IDOCAL), led by Dr. José María Peiró, for lending the facilities used in the data collection and data analysis.
Open access
