Big Data Analysis of Twitter-Based Sports Fandom: Celebrating Our Achievements Together During the 2019 Fifa Women World Cup

Abstract

ABSTRACT Social identification theory suggests that individuals try to categorize themselves into one group with people who shared similarities with them. While this premise applies to sports fans, nations, teams, and gender could be all sharing traits for sports fans to build group identities. Meanwhile, scholarship has uncovered that sports fans prefer to associate themselves with successful teams (i.e., Basking in Reflected Glory, or BIRGing) and to disassociate themselves from unsuccessful teams (i.e., Cutting Off Reflected Failure, or CORFing). Therefore, the purpose of the current study devoted to exploring whether sports fans identified with teams and nations, and how likely national identification and team identification lead to BIRG or CORF on Twitter about England’s matches against Norway, the United States of America (USA), and Scotland and USA’s matches against England, Spain, Chile, and Thailand during the 2019 Fédération Internationale de Football Association Women’s World Cup (FIFA WWC). Additionally, previous literature has confirmed that as a result of media consumption of sporting events, sports in media are male-dominated and may result in perceiving sports through male-gaze and categorizing sports in terms of sports gender typing. This study tried to observe whether the national and team identification outplay the gender differences in the audience. In Studies 1 and 2, statistical analysis and machine learning results both revealed that English fans tended to BIRG when England was leading or victorious. U.S. fans demonstrated the same behavior, BIRGing when the United States Women’s National Team (USWNT) was winning. However, English fans and U.S. fans were both less likely to CORF even though their team was losing or trailing. Rather, the identification with the nation or the team only brings fans together to celebrate. Thus, it is proposed that COATing was more accurate to describe sports fans’ reactions on Twitter. In Study 3, the unsupervised topic modeling analysis revealed that some English fans still BIRFed when USWNT finally beat the Lionesses. This finding challenged an essential assumption about social identification theory, which the inter-group difference does not necessarily result in ridicules or negative attitudes toward the out-group members. Further, by adopting a gender-guessing library from the Python dictionary, the names provided by Twitter users were analyzed for analyzing gender differences among the audience. Findings show that men were more likely to discuss FIFA WWC than women on Twitter. Meanwhile, it offered evidence that women cared less than men about female athletes and female sports. This implied gender differences were not minimized because of the national identity in mega-sporting events. Methodological contributions, theoretical and practical implications are discussed.