Group-Based Emotions: Evidence for Emotion-Performance Relationships in Team Sports

ABSTRACT

Purpose: In team sports, players have to manage personal interests and group goals, emphasizing intricacies between personal and social identities. The focus of this article was to examine the effect of identity mechanisms on appraisal processes, based on the following research question: Does the level of self-abstraction (low [personal identity] versus high [social identity]) lead to group-based emotions and influence performances? Method: An experimental design was used in which the level of self-abstraction was manipulated through the induction of a self- versus a team-oriented goal. Thirty elite male rugby players (M_age = 19.06, SD = 0.78, randomly split) participated in a match reproducing conditions similar to those of official games. Individual and perceived team-level emotions and performance were measured 17 times during the match. Results: Linear Mixed Effects models showed that a high level of self-abstraction: (a) led to more positive and less negative individual (variances explained: 52% and 46%) and perceived team-referent (variances explained: 57% and 40%) emotions; (b) reduced the correlation of team-referent emotions with individual ones; and (c) positively influenced team and individual performances (variances explained: 50% and 19%). Moreover, after controlling for potential effects of the level of self-abstraction, only positive team-referent emotions influenced performance. Conclusions: This study is the first to experimentally manipulate athletes’ social identity to examine group-based emotions in sport. Challenging the usual intrapersonal approaches, these findings suggested that social identity and its association with team-referent emotions could be one of the key dimensions of emotion-performance relationships in team sports.

KEYWORDS:

• Social identity
• group dynamics
• social appraisal
• intergroup context

Notes

¹ Following Cohen, a qualitative lecture grid for interpreting δ effect size (respectively r effect size) could be small (0.2, resp. 0.1), medium(0.5, resp. 0.3) and large (0.8, resp. 0.5).