Effects of a multifaceted team goal-setting intervention for youth volleyball teams

Abstract The current study examined the effects of a season-long multifaceted team goal-setting intervention (with emphasis on both individual and team level goal) on perceptions of team cohesion and collective efficacy. Using a non-randomized controlled design with 81 female volleyball players (Mage = 16.57, SD = .25) from six teams, three teams (n = 3) were assigned as intervention condition while the remaining teams (n = 3) represented no-treatment control condition. Teams in the intervention condition participated in a three-stage team goal-setting protocol with an extension of the individual goal-setting phase throughout a season. All participants completed questionnaires measuring perceptions of team cohesion and collective efficacy at three time-points throughout the season (i.e., beginning, midseason, end-season). In addition, participants in the intervention condition completed performance profiles at the beginning and the end of the season. Compared to the control group, the task cohesion perceptions of the intervention group were significantly higher in the midseason, the social cohesion perceptions were significantly higher at the end of the season, and their collective efficacy perceptions were significantly higher both in the midseason and at the end of the season. These results revealed the effectiveness of the multifaceted team goal-setting intervention on team cohesion and collective efficacy. Lay summary: This study extends the team goal-setting literature by conducting a season-long multifaceted team goal-setting intervention in youth volleyball teams. The intervention involved both individual and team goals and aimed to improve coherence between those goals through task interdependence.

clarify the paths to these goals, include all team members in the process, and reward the team's progress. Similarly, Eys et al. (2006) developed a three-stage protocol for implementing team goal-setting programs for practitioners. The protocol stages involve collective athlete selection of team goals, establishing performance targets for the team goals, and evaluating the progress toward team goals throughout the season.
The advanced protocols and guidelines are also helpful for researchers who aim to test the effectiveness of team goal-setting programs on various team outcomes. For example, using the three-stage team goal-setting protocol, Sen ecal et al. (2008) conducted an intervention with high-school basketball teams. Their results indicated that cohesion levels were maintained in intervention teams throughout the season but decreased in control group teams. Further, Rovio et al. (2012) implemented a season-long multifaceted team building program with a junior ice-hockey team. The program included team goal-setting followed by performance profiling, role clarification, and individual goal-setting. The performance profiles were used as a facilitator of the program and included identifying the areas needed for development for the team, setting individual and team goals, and progress toward them. Indeed, such features provide a practical advantage for using performance profiles in team-building interventions and team goal-setting programs (e.g., Rovio et al., 2010;Wikman et al., 2014). Although perceptions of task cohesion remained constant in the study, social cohesion perceptions increased throughout the season. These examples provide support the effectiveness of team goal-setting programs on improving team functioning and maintaining team cohesion. Nevertheless, several limitations are worth noting pertaining to team goal-setting programs in sport to date.
One limitation is that team goal-setting programs have predominantly focused on team goals. Typically, the programs focus on team-level goals and involve setting different team goals (i.e., performance, process, and outcome) at different time points (i.e., long-and short-term goals). However, teams are multilevel goal environments where individuals strive for both team goals and individual goals (DeShon et al., 2004;Widmeyer & Ducharme, 1997;Zander, 1971). Considering most athletes set individual goals, the relationship between the team and individual goals seems critical for the programs' effectiveness. Specifically, research in work settings showed that certain types of individual goals might hinder team performance in highly interdependent teams due to their focus on maximizing individual performance (Crown & Rosse, 1995;Kleingeld et al., 2011). Crown and Rosse (1995) defined these goals as egocentric individual goals which direct the efforts and attention toward individual outcomes rather than team performance. In contrast, group-centric individual goals focus on increasing individual effort on team performance and cooperation on team tasks. Given the highly interdependent nature of sports teams, these different types of goals may affect outcomes targeted by team goal-setting programs. Moreover, team goals must be relevant to individual goals and meet individual requirements (Johnson & Johnson, 1987). Thus, team goal-setting programs should include activities or phases to ensure the coherence between team goals and individual goals.
Another limitation of team goal-setting programs in sport pertains to task interdependence. As a task-oriented strategy, team goal-setting is related to team tasks as the team goals set the threshold for task outcomes (Weldon & Weingart, 1993). Tasks in sport teams vary according to their interdependence, which refers to the degree that team members must work together to perform a task (Hinsz, 1995;Johnson & Johnson, 2005). For example, whereas a particular defensive strategy in basketball could be a dependent task that requires the coordination and cooperation of team members, a free throw is an independent task performed individually with minimal coordination (e.g., Eccles & Tenenbaum, 2004). Since sport teams have specialized labor to perform these different types of team tasks (Benson et al., 2014), the commitment levels to the team goals may differ among team members with different roles according to the interdependence level of tasks. Precisely, a team goal related to an interdependent task might direct the coordination and cooperation between team members in specific roles (e.g., attackers in soccer). On the other hand, the same team goal might be perceived as not relevant to team members in other roles (e.g., defenders), and as a result, their commitment to that team goal can be lessened. For independent team tasks, team goals are attained with the athletes' combined individual contributions. For instance, team members need to achieve higher individual free-throw percentages to reach the 75% free-throw rate as a team goal. The importance of task interdependence levels in team goal-setting has been shown in organizational research (e.g., Aube & Rousseau, 2005). However, task-related variables such as interdependence, complexity and knowledge were rarely considered in goal-setting interventions in sport (Healy et al., 2018). Thus, team goal-setting in sport must consider of nature of team tasks and specific roles in sport teams for the effectiveness of the programs.
The purpose of this study was to examine the effects of a season-long multifaceted (i.e., team and individual components) team goal-setting intervention on team cohesion and collective efficacy. Both team cohesion and collective efficacy are considered important team outcomes in team building and team goal-setting research (Eys & Brawley, 2018). Based on the limitations mentioned above, the current program sought to extend the three-stage team goal-setting protocol (Eys et al., 2006) by including an individual goal-setting phase that aimed to provide coherence between individual and team goals. In addition, paying attention to the recommendations in team-building research (Bruner et al., 2013;Martin et al., 2009), the goal-setting program in this study was implemented throughout a season, used a control group and the effects of the program on team cohesion as well as on collective efficacy were evaluated. It was hypothesized that the season-long multifaceted team goal-setting intervention would enhance team cohesion and collective efficacy to a greater extent than those in a control condition. Also, individual goal levels would increase throughout the season.

Participants
Participants were 81 female athletes (Mage ¼ 16.57, SD ¼ .25) from six teams competing in regional an elite youth volleyball league. The athletes had been on their current teams for 2.48 (SD ¼ .82) years and had played volleyball for 6.77 (SD ¼ .82) years. The league consisted of 24 teams, with ten agreeing to participate in the study. For the goalsetting program to be implemented effectively, a team statistician was needed to capture performance statistics during contests. Given that this was a youth league, only seven teams had this type of member on staff. Further, four of these teams lacked available resources (e.g., no meeting rooms, rental fees for training) and felt that intervention implementation would be too difficult. As a result, three teams were assigned to the intervention condition, whereas three of the remaining teams represents the control group. As a result, the current study involved a non-randomized controlled trial with three teams in the goal-setting (Mage ¼ 16.76, SD ¼.43) and three teams in no-treatment control (Mage ¼ 16.38, SD ¼ .26) conditions. Of note, team statisticians of intervention teams were professional members of the coaching staff who trained to record team statistics in a computer-based program.

Procedures
Upon university ethical board approval, the researchers contacted the 10 head coaches who agreed to participate in the study. The researchers explained the study's aim and necessary conditions/resources. Accordingly, it was stated to the coaches that the study's primary purpose was to investigate the effects of multifaceted team goal-setting intervention on team cohesion, collective competence, and individual goal levels. Six teams were subsequently assigned to either an intervention (n ¼ 3) or control (n ¼ 3) conditions. The primary researcher separately met with all coaches and provided detailed information about the study. Specifically, they discussed the procedure as well as the questionnaires and their collection periods (i.e., beginning, midseason, and end-season). Once assent was provided by the athletes and consent forms from athletes' guardians and coaches were obtained, a meeting was held with coaches to discuss goal-setting implementation strategies and procedures. The first author remained involved with the teams to implement the intervention program, organize sessions, and answer possible questions throughout the season. Teams in the control condition completed the questionnaire packages at the beginning, middle, and end of the season, but otherwise had no interaction with the research team. Manipulation test results confirmed that control condition teams did not receive any systematic intervention.
For each of the intervention teams, the goal-setting program involved the entire season. Specifically, the program was introduced to intervention teams one week after the start of their pre-season workouts and continued until one week after the end of the intervention teams' official games (see Figure 1). Intervention sessions were held in the team facilities either before or after a team practice. The mean number of team sessions ($45 min in duration) per team was 19.0 sessions (SD ¼ 3.60), with 4.66 (SD ¼ 1.52) meetings per coach ($25 min in duration). Details of the program and the sessions can be found in the supplementary file. Athletes responded to questionnaires three times throughout the season (beginning, midseason, end-season) at training facilities (i.e., meeting room) before a team practice. For sample context, the league consists of three stages and the teams that are successful in the first two stages progress to the third stage. The first two mandatory stages consist of a five-month competition period with a two-week midseason break, while the third stage consists of two months. Except for one of the control group teams, all teams reached the third stage. In addition, all teams played one official match alongside training five days per week throughout the season, and no previous exposure to team goal-setting intervention.

Experimental design
Team goal-setting The team goal-setting intervention consisted of four stages in which team and individual goals were determined, evaluated, and revisited. Team goal-setting activities followed the three-stage team goal-setting protocol advanced by Eys et al. (2006) and aligned with a direct intervention delivery (Martin et al., 2009). Accordingly, the first author worked directly with the athletes in the sessions, and the coaches only attended the meetings about the team goals to convey their opinions on the determined goals to the team. In the first stage of the intervention, the first author provided athletes with general information about the intervention and an individualized workbook (see supplementary file) with related activities for each player. These activities aimed to enable athletes to understand the goal-setting technique, and in this context, they included determining the different types of goals (i.e., outcome, performance and process goals), establishing short and long-term goals, and setting individual goals that are compatible with team goals. He also asked players to think about their teams' long-term goals (e.g., finishing the season in the top three) and short-term objectives that would help them achieve those goals (e.g., promotion from the final stage). In a meeting, he had members independently set these goals, write them in their workbooks, and discuss them with team members. A series of sharing sessions ensued with the purpose of reaching a consensus among members for the top objectives as a team.
A subsequent meeting was held with coaches to discuss the long-term and short-term goals set by athletes. With some minor ordering changes and revisions to wording, the long-term and short-term goals were presented to the players as a group who were asked to now advance game indices (e.g., block percentage, middle attack number, sideout percentage) needed to attain their goals. Athletes independently selected four indices and discussed their opinions within one of four randomly assigned subgroups (three athletes per group). Each group then presented their four indices to the team. Eventually, the team agreed on the four indices (e.g., point-winning serve, attack percentage, errors, and block points) and these were positioned as team goals. Athletes were then provided with statistics from their previous season and exhibition games to aid with establishing ideal levels of performance in these categories. The long-(e.g., obtain 50% service reception percentage at the end of the season) and short-term target levels were determined in a similar process. Accordingly, athletes independently determined the target levels, discussed them in subgroups, and collectively agreed upon goal levels. Finally, athletes noted all of the goals and target levels in their workbooks according to the SMART goal guidelines (i.e., specific, measurable, adjustable, realistic, timebased). An example of one of the intervention teams' short-term goal was "increasing percentage of block points from 3.5 to 6 for the following three games." In line with the purpose of the current study, the second stage of the intervention involved the inclusion of individual goal-setting to the three-stage team goal-setting protocol of Eys et al. (2006). Specifically, athletes determined individual goals in accordance with the team goals through a series of activities. Since volleyball is a highly interdependent team sport, the first author provided general information about sport-specific tasks (e.g., middle attack, block) and their degree of interdependence with teammates (i.e., highly independent and interdependent tasks). Then, athletes were asked to identify the interdependence types of their teams' goal-related tasks. For example, increasing middle attacks as a team goal involve an interdependent task requiring team members to receive, pass, and attack in a coordinated way. Serving points, however, are an example of an independent task. During this individual goal process, the researcher had athletes identify the task components required to accomplish their goals (see Figure 2). Athletes established their individual goal contents (i.e.,  "what?") for each task component according to their team roles (i.e., "who is responsible"). There are five fundamental positional roles in volleyball: setter, outside hitter, middle blocker, libero, and opposite hitter. Athletes specialize in these roles, and each role has specific demands. For instance, outside hitters are primarily responsible for attacks while also responsible for making certain blocks. Finally, using performance profiling techniques, each athlete rated her current performance levels (i.e., at the beginning of the season) of identified goal contents and set four individual goals for the end of the season.

Team Goal
As suggested by previous team goal-setting programs (e.g., Sen ecal et al., 2008), the third stage of the intervention included reminders for coaches about team goals. Specifically, coaches were tasked to remind players about the team and individual goals through online platforms (e.g., team websites or social media platforms) and verbal feedback before and during practices and games. Individual goals were tracked by athletes using performance charts and performance profiles. Each player evaluated their current level of individual goals on a chart on a weekly basis (i.e., 1-10). These charts helped the athletes to remember their goals and track their progress. Similarly, athletes provided performance profiles for each goal before and at the end of the season. To obtain third party perspectives, coaches were also asked to evaluate athlete individual goals on the same scale. The coaches wrote the reasons for a possible difference between scores under the performance profile of each athlete.
In the fourth stage of the intervention, the researcher held sessions with teams in a direct fashion after each block of three games to review the team goals and target levels and make modifications if necessary. These modifications involved target level adjustments (i.e., raising or lowering the target level), changing a team goal, and possible solutions to encountered problems while striving for a team goal. With the assistance of the team statistician, the researcher provided statistics of the team goals for a threegame period in these meetings. While the accomplished goal levels were replaced with more difficult ones, the target levels were kept the same or lowered for goals that were not achieved in that period. All of these modifications were carried out according to the steps previously described. Only one goal was changed across all teams during the season. In that case, as close sets and matches were lost in a row, the team shifted their focus to points in critical moments of the sets and replaced their goal with a new one. The primary researcher and team then discussed the ways to make this new goal measurable and specific. One solution was to keep recording points taken and loss (i.e., þ/À) in critical moments and set a goal level for each block of three games. After the consensus, the primary researcher informed the coach and asked his/her approval about the modification.

Measures
Cohesion Cohesion was measured using the Turkish version of the Youth Sport Environment Questionnaire (YSEQ; Eys et al., 2009). The Turkish version was adapted and validated by Sezer and Kocaekşi (2018) with soccer, basketball, and volleyball players. Like the English version, the Turkish YSEQ (Sezer & Kocaekşi, 2018) is an 18-item questionnaire that assesses task and social dimensions of cohesion. Each dimension has eight items (e.g., task cohesion: "We all share the same commitment to our team goals," social cohesion: "I invite teammates to do things with me"). Items are scored on a 9-point Likert-type scale from 1 (strongly disagree) to 9 (strongly agree), and higher scores reflect greater perceptions of cohesion. The Turkish version has demonstrated adequate reliability scores (task ¼ .87, social ¼ .87) and internal consistency scores (i.e., Cronbach alpha) for present study were also satisfactory (task ¼ .84, social ¼ .91).

Collective efficacy
We used the adapted and validated version of the Collective Efficacy Scale (Riggs et al., 1994) advanced by € Ocel (2002). As with the original version, the Turkish version comprises seven items (e.g., "The members of this team have excellent game skills") with a 5 -point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The scale has no subdimensions and the collective efficacy assessed on a total score ranging from 0 to 35. The internal consistency score of the Turkish version was acceptable, with a Cronbach alpha score of .70. In the present sample, internal reliability was also acceptable (a ¼ .79).

Performance profiles
Performance profiling (PP) is a technique that allows athletes to identify and assess essential aspects of their performance (Butler & Hardy, 1992;Gucciardi & Gordon, 2009). The PP technique can also be used as a feedback tool in which coaches rate athletes' structures, thereby focusing on possible discrepancies between athlete and coach scores (Butler et al., 1993). In line with these purposes, PP was used for establishing and monitoring individual goals in this study. PP was established by athletes in the team goal-setting group (N ¼ 37) in line with their team goals as a part of the intervention (see details above). Thus, only four individual goals per athlete were taken into account for the present study. Athletes rated their individual goals two-time points at the beginning and end of the season.

Postintervention-manipulation check
The postintervention questionnaire comprised five open-ended items that aimed to test the effectiveness of the team goal-setting intervention in the experimental group. The questionnaire was based on sport psychology assessment forms (Bloom & Stevens, 2002) and adapted from the team goal-setting intervention by Sen ecal et al. (2008). Examples of questions included: "Did the team goal-setting program helped your team to play better together?," "How the team goal-setting program helped your team to play better together?" A similar questionnaire administrated to the control group aimed to investigate whether there was a team-building intervention applied by the coach in these teams (e.g., "Throughout your regular season, were there any activities implemented by your coach designed to affect the closeness of your team either on or off the court?").

Data analysis
Data analyses were performed using SPSS version 22.0 software (IBM SPSS). Missing data were estimated through maximum likelihood estimation. A multivariate analysis of variance was used to test whether there was a significant difference between perceptions of cohesion and collective efficacy before the intervention (Time 1). Results showed that there was a significant difference between groups at the beginning of the season, Wilk's k ¼ .870, F(3,73) ¼ 2.766, p ¼ .034, g 2 ¼ .130. Follow-up univariate analysis of variance (ANOVAs) showed significant differences between groups for task cohesion [F(1,75) Considering that the control group had higher scores for all three variables (collective efficacy, task cohesion, and social cohesion), a multivariate analysis of covariance was used to compare scores at midseason (Time 2) and the end of the season (Time 3), with pretest scores serving as covariates. Dependent t-tests were used to analyze the experimental groups' differences between the performance profile scores before and after the season at 95% confidence interval.

Descriptive
A summary of descriptive statistics can be found in Table 1. Notably, intervention group teams had higher perceptions of task cohesion, social cohesion and collective efficacy levels after completing the season-long intervention (i.e., Time 1 vs Time 3). Also, the mean winning percentage of the control group teams was 42.60 (SD ¼ 6.33) in comparison to 45.77 (SD ¼ 9.41) for the intervention teams.

Main analysis
A multivariate analysis of covariance (MANCOVA) was used to identify group differences in the dimensions of cohesion and collective efficacy through the season, with pretest scores being included as covariates. Results (Table 2) revealed that there was a significant difference between groups (g p 2 ¼ .28). Follow-up univariate ANOVAs indicated that the groups differed significantly in task cohesion in midseason (g p 2 ¼.06) but not at the end of the season (g p 2 ¼ .00). While groups' social cohesion perceptions did not significantly differ in midseason (g p 2 ¼ .01), there was a significant difference at the end of the season (g p 2 ¼ .12). For collective efficacy perceptions, groups significantly differed in midseason (g p 2 ¼ 09) and at the end of the season (g p 2 ¼ 07).

Individual goals
A total of 37 athletes in the team goal-setting group completed performance profiles at the beginning and the end of the season. In addition to other performance indices, athletes identified four group-centric individual goals compatible with their team goals and roles (e.g., setter, outside hitter, libero). For example, a setter who had a team goal to increase mid-attacks had set her individual goal to "increase the percentage of passes to the middle attackers." In total, athletes set 128 technical (e.g., better reception) and tactical goals (e.g., diversifying attack options). Table 3 shows the mean scores of athletes and coaches performance profile ratings at the beginning and the end of the season. Accordingly, while the scores of the athletes decreased, the scores of the coaches increased.
According to the normality test (Kolmogorov-Smirnov), the statistical significance between athletes' mean scores at two-time points were tested with the Wilcoxon sign rank test. The results showed no significant differences between athletes' performance profile scores at the beginning and the end of the season (z ¼ À1.040, p ¼ .298, r ¼ À.23). For the scores of coaches, dependent sample t-test results showed that there were no significant differences between T1 and T3 [t(36) ¼ À.560, p ¼ .582]. Finally, the difference between coaches' and athletes' scores at two time points (T1 vs. T3) were also nonsignificant [t(36) ¼ 1.16, p ¼ .256].

Postintervention manipulation-check
A total of 59% of the athletes in the intervention group stated that the team goal-setting program improved overall team cohesion. While 35% of these athletes stated that the reason for this increase in team cohesion was open discussions in the meetings held within the scope of the program, 27% stated that the increase occurred due to the program's focus on team goals. A total of 41% of the participants who stated that the program did not affect team functioning noted that they could not follow the goals during the games (35%) and highlighted organizational problems during the season (e.g., uncertainties in the training and match schedules). Results also showed that 60% of the control condition participants participated in activities (e.g., team dinners), and 85% reported that these activities improved cohesion.

Discussion
The purpose of this study was to examine the effects of a season-long multifaceted goalsetting intervention on perceptions of cohesion, collective efficacy and individual goal levels of elite youth volleyball players. Compared to the control group, the task cohesion perceptions of the intervention group were significantly higher in the midseason (g p 2 ¼.06), the social cohesion perceptions were significantly higher at the end of the season (g p 2 ¼ .12), and their collective efficacy perceptions were significantly higher both in the midseason (g p 2 ¼ 09) and at the end of the season (g p 2 ¼ 07). In contrast, a decrease was observed in the control group team cohesion and collective efficacy perceptions. In addition, individual goal levels of team goal-setting athletes remained constant throughout the season. These findings and related implications are discussed below.
Our findings support those from previous research pertaining to the importance of team goal-setting for enhancing cohesion. In line with previous studies (e.g., Rovio et al., 2012), our results showed that cohesion perceptions increased in team goal-setting group after completing the season-long intervention (i.e., Time 1 vs Time 3). Notably, those perceptions were significantly higher than the control group in the midseason (g p 2 ¼.06). One prominent feature of team goal-setting interventions is the ability to direct team members to focus on team goals and tasks (Widmeyer & Ducharme, 1997). To ensure this, it is necessary to clearly define team goals with team members, discuss required strategies to achieve the goals, and systematically follow the team progress. Our intervention program employed the team goal-setting strategy of Eys et al. (2006), which included clarification of team goals through team discussion and tracking identified goals throughout a season in the experimental group teams. Therefore, increased social and task cohesion perceptions of experimental group teams may stem from the team goal-setting program.
Given the task-oriented nature of the team goal-setting programs, it is plausible to expect positive effects on task cohesion. Indeed, studies in sport and organizational settings have shown that team goal-setting interventions enhance or maintain team cohesion levels (O'Leary-Kelly et al., 1994;Stevens & Bloom, 2003). For instance, with 14-18 years old basketball players, Sen ecal et al. (2008) conducted a team goal-setting intervention and found that experimental group teams' task cohesion remained stable throughout the season, whereas the control group perceptions significantly decreased. The authors surmised that a ceiling effect might have been responsible for this result since experimental group teams' cohesion levels were already high before the intervention. The increase of task cohesion in the midseason in the current study may stem from the moderate task cohesion levels before the intervention. Thus, our findings align with previous research suggesting the effectiveness of team goal-setting interventions for increasing task cohesion. Indeed, the results of Martin et al.'s (2009) meta-analysis support this argument by revealing that goal-setting was the most effective type of team building intervention. Nevertheless, task cohesion perceptions of experimental group teams decreased at the end of the season and resulted in a non-significant difference between the two groups. One of the reasons for the decrease may be the decrease in the performance of the experimental group teams in the third stage of the season. Specifically, while two of the three teams moved away from the outcome goals they set for the end of the season (e.g., finishing top 3), only one of them was able to reach the outcome goal. Studies in the literature have demonstrated the reciprocal relationship between cohesion and performance (e.g., Benson et al., 2016). Mainly, this relationship is more robust in the dimension of task cohesion (Filho et al., 2014). In this respect, the decrease in performance may have caused the decrease in task cohesion perceptions in experimental group teams. Another possible explanation of increased task cohesion perceptions in the midseason (g p 2 ¼.06) could be related to the current program's multifaceted design (i.e., individual þ team goal-setting). As an additional phase of Eys et al.'s (2006) three-stage team goalsetting protocol, the current program included group-centric individual goals, defined as individual goals set to maximize contribution to a group task (Crown & Rosse, 1995). These goals promote individual contribution to task performance and provide a clear path for identifying that contribution for interdependent tasks. When implemented with group goals (i.e., multifaceted), the group-centric individual goals enhance team productivity and individual commitment to team goals (Kleingeld et al., 2011).
Similarly, team tasks in the current sample require high interdependence due to the nature of volleyball (i.e., a maximum of three consecutive touches of the ball return to ball). Like most team sports, each team member in volleyball plays a role to perform a team task. For instance, to achieve a team goal, such as increasing the opposite attack rate from 20% to 30%, the team needed a better reception percentage from receivers, more passes to the opposite attackers from passers, and more accurate attacks from opposite attackers. Therefore, with the addition of group-centric individual goal-setting, athletes may have a better understanding of not only the team goals, but also how they can contribute to achieving these team goals as an individual, which may have led to more commitment to team tasks. Notably, having individual goals that aligned with team goals (i.e., group-centric individual goals) could promote reflection about individual roles and direct individual attention toward team goals.
It is important to note the increase in perceptions of social cohesion for the intervention teams at the end of the season (g p 2 ¼ .12), consistent with the previous literature (e.g., Martin et al., 2009). Despite their task-oriented approach, team goal-setting programs in sport enhance social cohesion perceptions due to implemented with team building principles. Widmeyer and Ducharme (1997) listed principles of team building through team goal-setting and emphasized the participation of athletes in team goal-setting processes and other principles (e.g., monitoring team progress). As mentioned above, the current team goal-setting program followed the three-stage team goal-setting protocol (Eys et al., 2006) based on these principles (Eys & Kim, 2017). Specifically, each athlete identified team goals and levels first, then discussed them goals in team discussion. This strategy, which provides open communication and discussion about team goals, might positively affect the team goal-setting group social cohesion perceptions. Indeed, post-intervention tests results supported this argument, and 35% of athletes indicated "open communication about team goals" as a reason for the increased cohesion. On the other hand, there was a non-significant difference in social cohesion levels between the experimental and control groups in the midseason. One of the main reasons for this interesting finding may be the already high social perceptions in the control group in pre-season. Specifically, while social perceptions in the control group tended to decrease throughout the time points, the increase in the experimental group in midseason brought social cohesion levels closer to each other. Considering that social cohesion displays variations in time (Leeson & Fletcher, 2005), the program's features mentioned above may support maintaining social cohesion levels in the experimental group.
Another finding worthy to note involves collective efficacy perceptions in midseason and at the end of the season. Of interest, perceptions of collective efficacy were higher for the intervention group than the control group both in-season and at the end of the season. As an important construct for team performance, strategies that develop collective efficacy have been studied with practical and theoretical interests. Team goal-setting is considered an ideal strategy for developing collective efficacy in sport teams, as it includes steps such as clarifying collective goals and necessary actions (Carron et al., 2005;Heuz e et al., 2007). The current team goal-setting program also included these steps with providing collaborative discussions about team goals as a group. Also, the program involved goal modifications which increased the likelihood that teams would achieve certain levels of performance. Specifically, if a team could not reach the predetermined goal in a threegame period, the goal was adapted to ensure it could be achieved. This approach may have contributed to improved efficacy perceptions in the team.
Another feature of the program that may have led to increased collective efficacy perceptions in midseason and at the end of the season is the alignment of individual goals with team goals. In interdependent groups, collective efficacy perceptions stem from members' judgements about others' capabilities and efficacies (Bandura, 1997). Thus, knowing the individual goals of others and the alignment of these goals with the team's objectives might have increased athlete confidence that their teammates could achieve individual goals. Consequently, this confidence in others may have positively affected the athletes' collective efficacy beliefs about their teams. Another possible explanation for this finding is the established association between collective efficacy and cohesion (Heuz e et al., 2006;Kozub & McDonnell, 2000;Paskevich et al., 1999). The current study's findings supported this relationship with parallel changes in collective efficacy and cohesion levels in the goal-setting teams throughout the season.
The current study included the individual goal-setting phase as a novel inclusion to the three-stage team goal-setting program (Eys et al., 2006). The purpose of this addition was to implement the team goal-setting intervention at both individual and team levels (i.e., multifaceted) in line with the recommendations in organizational psychology (e.g., Johnson & Johnson, 1987;Zander, 1971). To fulfill this purpose, the intervention group athletes first analyzed the team tasks associated with their team goals. Then, they determined four individual performance indices related to their team roles (e.g., spiker). Finally, the athletes scored these four performance indices on the performance profile at the beginning of the season and determined their individual goals (i.e., scores they wanted to reach) at the end of the season. The intervention team coaches similarly scored the individual goals before and after the season. According to results, it appears that the multifaceted team goal-setting intervention failed to improve athletes' and coaches' perceptions of individual goal levels. Specifically, there was no significant relationship between the athletes and coaches' performance profile scores before and at the end of the season. This finding was unexpected given the support for the effects of individual goal-setting programs on target behaviors and performance in sport settings (e.g., Burton & Weiss, 2008;Kyllo & Landers, 1995). The primary cause of this unexpected finding may be attributed to the use of performance profiles in assessing individual goal levels. As a subjective assessment tool, performance profile scores are affected by different social and cognitive structures (D'urso et al., 2002;Doyle & Parfitt, 1996). In addition, Doyle and Parfitt (1996) pointed out that their experience or inexperience can shape the performance profile scores of the athletes with the technique and the evaluations of the coaches. Specifically, in this study, while the pre-season low scores given by the athletes regarding their individual goals might express a perception of the aspects they need to develop, the scores they give at the end of the season may have been influenced by the feedback they received from the coaches and other social variables.
Another possible explanation could be related to systematic feedback on team-level goals. In multifaceted goal-setting studies conducted in organizational settings, it has been observed that giving feedback at the team level leads to the development of attitudes toward the team. In contrast, the individual level feedback causes the individuals in the team to focus on their performances (DeShon et al., 2004). In this study, feedback was given by the coaches and the primary researcher only for the team goals. Also, these team-level feedbacks were supported by objective data on team outputs, while a subjective tracking method was followed for individual goals (other than scoring and tracking these goals), and no systematic feedback mechanisms (e.g., weekly feedback of coaches) was included in the program by the researchers. These strategies may have enabled the individuals to focus on team goals rather than their own goals by providing feedback at the team level. This issue will be addressed by involving effective feedback mechanisms for individual goals in future interventions.
It is also essential to note that the current intervention program is methodologically different from previous goal-setting studies in sport settings. The current study is the first report of a multifaceted team goal-setting intervention in sport settings to our best of knowledge. While most goal-setting programs in sport settings have focused on individual or team goals on performance, the individual goals were mainly unrelated to the team goals in several omnibus intervention programs. For example, Rovio et al. (2012) implemented a seasonlong team building program with high school ice hockey players. The intervention program involved team goal-setting, role clarification, and individual goal-setting strategies (i.e., omnibus program). Upon a closer inspection, the team goals focused on several team norms (e.g., preparation for training and matches; no absences) while the individual goals targeted task-related outcomes such as stick control, carrying the puck. As emphasized in organizational settings (Deshon et al., 2004), the lack of coherence between individual and team goals limit the programs generalizability to other settings. We believe such issues could be overcome by establishing intervention programs implemented on both individual and team levels. Finally, these null findings at the individual level could be attributed to a focus on a specific goal-setting strategy (i.e., SMART). The effectiveness of these strategies has recently been called into question because they do not take into account individual differences in goal-setting practices (Jeong et al., 2021). Future studies may implement individual goal-setting programs that take individual differences into account.
Despite the strengths of this research, our findings must be considered in light of several limitations. First, full-randomization was not possible because teams did not have a team statistician in their coaching staff. Moreover, it was impossible to implement this program in several teams due to the lack of team resources (e.g., club facilities). Consequently, the results of the current study employed a non-randomized sampling. Although randomized controlled trials are frequently applied in medical sciences, they are not always feasible in social sciences (Bickman & Reich, 2008). These constraints also resulted in a decrease in the sample size. Although such interventions are highly demanding, increasing sample size does provide benefits for generalizability and statistical power. Given the nested structure of the data (i.e., players within teams), increasing the sample size will also enable multilevel analyzes that can provide more precise insights into the effects of interventions for future studies. In addition, the inability to implement the intervention due to lack of statistician or club facilities in some teams should be considered a practical limitation. Since some clubs do not have these opportunities for various reasons (e.g., operate at the recreational level), it will be difficult to implement the intervention program effectively. Another limitation pertaining to the methodology is that tracking individual goal levels objectively (i.e., individual game statistics) has not been possible because of data analysts increasing workload. Since the individual goals required separate adjustments from team statistics, team statisticians had to re-watch the game videos to follow these levels objectively and could not achieve this due to workload. As a result, the individual goals were followed with performance profiles. Future studies may work with more statisticians to eliminate this limitation. It is also worth noting that the use of the Collective Efficacy Questionnaire for Sports (CEQS; Short et al., 2005) may provide detailed knowledge about the effect of the intervention on collective efficacy. Since the Turkish version of CEQS (20 items) has recently published ( € Onc€ u et al., 2018) and showed significant correlations with the Collective Efficacy Scale (Riggs et al., 1994), we preferred to use the shorter, one-dimensional Collective Efficacy Scale (7 items) to reduce the burden of filling out too many items on athletes. Finally, it is important to note that team performance is not a priori focus of the study. Although the season performance of the intervention and control group teams was not significantly different, and five out of six teams reached the third part of the season, the change in the performance during the season may have affected the teams' perceptions of collective efficacy and cohesion. This is relevant considering the findings of the reciprocal relationships of cohesion (e.g., Benson et al., 2016) and collective efficacy (e.g., Myers et al., 2004) with team performance. Since the winning percentage was calculated only at the end of the season, it could not be included in the statistical analysis as a covariate. Future studies could consider including more specific (i.e., attack percentage) and longitudinal performance measurements throughout the season to depict the effects of team performance on team outcomes.
In conclusion, this research has shown that a multifaceted team goal-setting intervention can significantly increase team cohesion and collective efficacy levels of youth volleyball teams at different time points throughout a season. These findings are in line with previous team goal-setting studies in sport settings (e.g., Sen ecal et al., 2008). Another contribution of this study is that, according to our knowledge, it is the first example of a multifaceted goal-setting strategy in sport settings. This implementation strategy can make an important contribution to goal-setting literature by enabling more comprehensive interventions in sport psychology. Future studies could test the effectiveness of multifaceted goal-setting strategies on the various team and individual outcomes. Researchers can also compare the effects of goal-setting programs that include performance-only, process-only, and outcome-only goals on the team and individual outcomes, thereby improving our understanding of goal-setting. Another area for future consideration concerns the delivery strategy. In line with the direct delivery strategy first author directly worked with athletes to implement the goal-setting program. Training the coaches to apply the goal-setting program to their teams (i.e., indirect delivery strategy) may improve the program's effectiveness. Finally, practitioners must acknowledge the advantages and disadvantages of season-long team goal-setting programs regarding delivery and content. The use of technology (e.g., mobile applications) would be beneficial in reducing program workload and ensuring effective delivery.