In recent years, social development through sport is a highly topical issue. At the mention of sport, there is a significant amount of news and reports support that sport is an effective instrument that can help improve the quality and development of human, families and countries in general. There are also many sports organizations strongly promoted the sport as a tool for social change. The UN (2008) recognizes that sports are incredibly valuable for improving social development and economic progress. The International Olympic Committee (IOC) has a project, which uses sport as a tool for development and advances the Sustainable Development Goals (SDGs). Many academics have conducted many studies and they have come to the view that sport has both positive and negative effects on social development.

The impact of sport on social development has been largely debated. Sport is a context that provides participants with the opportunity to control themselves, resolve conflicts, and learn to work with others (Boardley & Kavussanu, 2011). An extensive literature records the impacts of sport and physical activity in improving character since the early 1900s. According to one point of perspectives, sport provides an effective tool through personal and social development and can have a positive influence on young people (Morris, Sallybanks, Willis & Makkai, 2003). Many coaches, sport enthusiasts, and advocates for sport firmly convinced that sport can build positive character values like honesty, responsibility, fairness, and respect, and they suppose them to be one of the ultimate objectives of physical participation (Doty, 2006).

Previous studies have provided evidences in which more intense sport participation such as elite sport increase diverse socio-psychological development. Different levels of sport participation show different social behaviors between elite sport participants and recreational sport participant (Côté & Fraser-Thomas, 2007). Spink (1992) found that both individual attractions to a group task and social integration of a group were different significantly between the elite and the recreational in volleyball, and that cohesiveness was positively related to the elite volleyball teams but not for recreational volleyball teams. Serrao, Martens, Martin and Rocha (2008) found that elite sport participants in collegiate level had higher competitiveness than recreational sport participants did. Also, Šmela, Pačesová, Kraček, and Hájovský (2017) reported that there were significant differences between elite athletes and recreational sport participants in performance motives, anxiety inhibiting performance, and anxiety supporting performance. Nonetheless, it is hard to find an evidence that different participation intensities are associated with prosocial and antisocial behaviors.

Much literature in sport have focused on the negative effect of social behaviors. Sport attracts everyone to do some illegal acts such as rule-breaking and deception (Boardley & Kavussanu, 2011). Sage (1998) revealed that sport did not build character in general, team sport athletes come out more negatively influence by the competitive experience compared to individual athletes. The social nature of sport providing positive educational opportunities, in the meantime quite often leads to the possibility of cheating, lying, bullying and hurting other participants, leading to antisocial behavior (Kavussanu, Seal, & Phillips, 2006). Kroll and Petersen (1965) presented evidence indicates that such principles were less, rather than more, commonly endorsed by regular sport participants.

Generally, social behavior includes prosocial and antisocial behaviors. Prosocial behavior required some sacrifices for the actor and should be done without any external reward (Underwood & Moore, 1982). Antisocial behaviors are the action that harms or lack consideration for the happiness of others (Horowitz, 1987). It has also been defined as any type of conduct that violates the basic rights of another person (Calkins & Keane, 2009) and any behavior that is considered to be disruptive to others in society.

The previous research of prosocial and antisocial behavior have not been sufficient. In the late twentieth century, researchers focus on Rest’s four-component model of social behavior action: awareness, judgment, commitment and action (Rest, 1984), but neglected social behavior reasoning, athlete’s attitudes and orientations, and the judgment of the legality of harmful activities, and so on (Vallerand, Brière, Blanchard, & Provencher, 1997). Bandura (1999) proposed that social behavior had two aspects: proactive and inhibitive. According to the different stages of human growth, Haan (1978) divided the social behavior theories into five levels: assimilative phase (level 1 and 2), accommodative phase (level 3 and 4), equilibration phase (level 5).

The term prosocial behavior has always been associated with developmental characteristics in children (Eisenberg & Mussen, 1989), the literature on the topic has occurred to include adult behaviors since recently as well. As the part of sport studies have primarily considered the inhibitive morality by a survey that has anti-sociality effects on others (Kavussanu, 2008). Studies of the effects of sport competition have revealed that it reduces prosocial tendencies, such as helping and sharing (Barnett & Bryan, 1974; McGuire & Thomas, 1975) and increases antisocial tendencies (Berkowitz, 1972) and all these effects are exacerbated by losing (Kleiber & Roberts, 1981). Al-Yaaribi and Kavussanu (2018) found that prosocial behavior was positively related to effort, commitment and perceived performance in a team sport.

In the studies of sport, researchers have explored different social behaviors. Most sport researches have focused on the related concepts of prosocial and antisocial behavior (Graupensperger, Jensen, & Evans, 2018; Shields, Funk, & Bredemeier, 2018). Sage, Kavussanu, and Duda (2006) utilized prosocial and antisocial behaviors and the dimensions of social behavior. At the same time, Sage and Dual (2006) also addressed some problems of social behavior in sport by using the Prosocial and Antisocial Behavior in Sport Scale (PABSS) which is developed by Kavussanu and Boardley (2009). The PABSS is 20 items and includes 4 factors: individual prosocial and antisocial behavior, directed at either teammates or opponents (Kavussanu & Boardley, 2009). However, most of the participants in the studies were team sport players, and there was a lack of individual sports cases (Funk, 2017). Moreover, scant literature identified the factors constructing prosocial and antisocial behaviors.

Therefore, the purpose of the present study was to find prosocial and antisocial behaviors using an existing social behavior questionnaire and to examine the effect of sport on prosocial and antisocial behaviors. To fulfill the purpose of the study, a research question was set for this study if there were mean differences on social behavior factors between elite sport participation group and recreational sport participation group.

Table 3 shows the mean difference between the elite sport participant and the recreational sport participant on each dependent variable. On most of the social behavior factors, the mean scores of the recreational sport participant group were higher than those of the elite sport participant group, except for the arbitration and aggressiveness.

The present study conducted a one-way MANOVA as followed. Prior to the main test, the multivariate test for homogeneity of dispersion matrices, Box’s Test was examined, and it was significant, which rejected the homogeneity assumption, Box’s M=65.47, F(28, 569673.63)=2.304, p<.001. Given that, Levene’s Test for the homogeneity of variance was also examined, and the results were also significant except for social facilitation and innergroup conflict (Table 4). However, if the sample size is bigger than 30, MANOVA is robust against the violation of homogeneity of variance-​covariance matrices assumption, Box’s Test, and a stricter α level than .05 should also be considered for univariate F test such as .001 to compensate a significant Levene’s Test result (Allen & Bennett, 2008).

The MANOVA result shows a significant difference across the two groups on the set of dependent variables (Table 5). Hotelling’s T-Square indicated a statistical significance between the elite sport participation group and the recreational sport participation group, Hotelling’s T²=.094, F(7, 581)=7.833, p<.001. 8.6% of the generalized variance for the set of DVs was explained by the group difference.

ANOVAs on dependent variables were conducted as follow-up tests to the MANOVA (Table 5). Because of the significant result of the Levene’s Test as explained the above, each ANOVA was tested at .001 level. ANOVAs were significant on the leadership, F(1,587)=28.591, p<.001, η²=.046; the social facilitation, F(1, 587)=19.086, p<.001, η²=.031; and, the group cohesiveness, F(1,587)=12.403, p<.001, η²=.021. However, other ANOVAs were nonsignificant on the emotional support, F(1,587)=9.732, p=.002, η²=.016; the arbitration, F(1,587)=.121, p=.728, η²<.001; the aggressiveness, F(1,587)=.068, p=.794, η²<.001; and the innergroup conflict, F(1,587)=11.314, p=.001, η²=.019.

In sum, the social behavior of the elite sport participant group and that of the recreational sport participant group was statistically different. Especially, the recreational group scored higher than the elite group on 3 of the prosocial behavior factors including the leadership, the social facilitation, and the group cohesiveness. However, there was no difference on the antisocial behaviors.

In sum, the present study founds as the followings. First, utilizing an existing questionnaire, this study identified diverse facets of prosocial and antisocial behaviors in sport, which could be comparable to the previous studies. There were 5 prosocial behaviors (emotional support, leadership, social facilitation, arbitration, and group cohesiveness) and 2 antisocial behaviors (aggressiveness, and innergroup conflict) in sport settings. Second, an elite sport participation that was more intensive and serious participation in sport did not increase social behavior; furthermore, less intensive participation forms could provide better social behavior. Third, the recreational sport participation and the elite sport participation differ on prosocial behavior. Especially in leadership, social facilitation, and group cohesion, the recreational sport showed higher than the elite sport. Fourth, the elite sport participant and the recreational sport participant have no significant difference on antisocial behaviors including aggressiveness and innergroup conflict.

Given the above, overly serious participations do not guarantee social behaviors positively. According to this, those organizations that want to promote social development through sports, then they should pay more attention to the promotion of the recreational sport. Those who claim that sport brings social instability also should be noted that antisocial behavior of different level of sport participants is no different. Generally, however, this research should be regarded as preliminary. Consideration of other team sports or individual sports, various age or gender, and different regions, and more sophisticated longitudinal research designs will all be needed to adequately determine the role of sport in social development. It would be interesting for future research to find out these possibilities.