Abstract
An increasing number of athletes are speaking out on different social issues. There is also a growing acceptance and expectations of athletes utilizing their influence to push for social change. In recent years, some athletes have started disclosing their mental illness and raising awareness of mental health. This study investigates Twitter user data on Naomi Osaka’s withdrawal from the French Open due to her mental health concerns, as well as examines how different Twitter sources or handles (e.g., athlete vs. news media) further shape public responses. Both thematic analysis and sentiment analysis were conducted on Twitter data to identify emergent themes and audience sentiment towards the case. Findings reveal growing positive reception of athletes’ mental health disclosures, athletes’ behavior as a form of advocacy, and utilization of social media to amplify activism efforts. The results also extend the athlete activism literature by showing how message sources shape public reactions differently. Findings provide implications for athletes and relevant stakeholders to effectively navigate through the situation by creating and delivering messages to their target audiences.
Introduction
Professional athletes can have a significant influence on social changes when they speak publicly (Gill, 2016; McCauley & Yang, 2020). Over the past few years, more athletes have taken advantage of their influence and utilized sports as a platform to advocate for a variety of social issues. For example, Colin Kaepernick took a knee during the national anthem to protest racial injustice and police brutality (Carpenter, 2016). The U.S. Women’s National Soccer Team has continued to fight for equal pay and has garnered huge media attention and support, which eventually made the United States Soccer Federation (USSF) settle for $24 million in back pay to female athletes (Das, 2022). A large body of literature has examined the media coverage of athlete activism (e.g., Coombs et al., 2019) and fan responses to player protests (e.g., Frederick et al., 2017; Gill Jr., 2016). These studies have contributed to a better understanding of how athletes use their platform to create conversations around various social issues, communication strategies, the experience of athlete activists, and possible consequences of their behaviors. It is important to note that when activism involves controversial behaviors or polarizing issues that have competing values, it can create strong emotions and even alienate certain groups of consumers depending on their predisposed beliefs (Moorman, 2020; Schmidt et al., 2018a).
In the current study, we focus on the case of Naomi Osaka’s withdrawal from the French Open in 2021 due to her mental health concerns. During the first week of the tournament in 2021, then No. 2 ranked Naomi Osaka announced that she would not participate in press conferences due to her mental health concerns. Soon afterwards, the organizing body of the tournament issued the athlete a $15,000 fine for her violations of contractual media obligations (Roland Garros, 2021). Naomi Osaka also faced potential suspension from future Grand Slams with her refusal to fulfill media obligations (Goel, 2021). On May 31st, Naomi Osaka posted a statement on Twitter and Instagram to announce her decision of formally withdrawing from the tournament and reveal more details on her mental health struggles (Osaka, 2021a). Her tweets received support from other athletes as well as her sponsors such as Nike and Nissan (Gregory, 2021b). Naomi Osaka also received support and praise from the public and mainstream media (e.g., Wallace, 2021). At the same time, Naomi Osaka’s tweet drew some criticisms from groups of fans and sport journalists. Previous studies have found positive and supportive narratives surrounding mental health disclosures from athletes (Cassilo, 2022; Cassilo & Kluch, 2021; Parrott et al., 2020, 2021). News media and social media users also framed these athletes as mental health advocates (e.g., Cassilo & Kluch, 2021). Athletes who speak out on mental health create a snowball movement (Parrott et al., 2021) and their experiences may help change the cultural norms of mental illness (Cassilo & Sanderson, 2019).
It is critical for sport scholars to study mental health disclosures, which have an important influence on larger discussions about mental health (Lavelle, 2021). Researchers have called for more research in this area to continue examining the frames of mental illness in sport (Cassilo, 2022), intensify previous findings (Cassilo & Kluch, 2021), and determine more generalizable trends (Parrott et al., 2021). Researchers also suggest that future studies should consider athletes from different identity groups and sports, as well as factors such as communication strategies and context (Cassilo, 2022; Cassilo & Kluch, 2021). Thus, the purpose of this study is to further extend the body of work by investigating Twitter user responses to Naomi Osaka’s decision to withdraw from the French Open due to her mental health concerns. The case presents a different setting from previous studies in that Naomi Osaka is a mixed-race (Asian and Black) female athlete competing in a predominantly White sport (cf. Cassilo, 2022; Parrott et al., 2020). Moreover, she not only disclosed her mental struggles on social media, but also decided to retire from a major tournament, which drew a lot of media backlash and even threats from the governing body (Goel, 2021). This study advances our understanding of audience perceptions of mental health advocacy when an athlete’s action draws both criticism and support from the audience. Moreover, the current study extends the athlete activism literature by considering message sources (athletes vs. sport media) as an important predictor of audience response and exploring the potential effect of parasocial interaction in mental health communication. We collected replies from various sources (e.g., Osaka’s and other athletes’ tweets, ESPN Sports Center, Bleacher Report) that featured Osaka’s withdrawal announcement. This allowed us to compare responses across different media sources and further delineate the source effect in mental health disclosure.
Athletes as an Important Voice for Social Change and Mental Health
Sport serves as a key platform for activating awareness and promoting social change (Nielson, 2020). Traditionally, professional athletes prefer to remain silent regarding politics and social issues because it may risk their employment status and diminish their endorsement potential (Gill, 2016). However, growing numbers of athletes are willing to take a stand and participate in social activism, including racial injustice (Schmittel & Sanderson, 2015), gender inequality (Manney, 2021), and LGBTQ rights (Williams, 2020). As more professional athletes are participating in social activism, fans also have higher acceptance and expectations of athletes using their influence to speak out about important social issues compared to the past (ESPN, 2020; MarketCast, 2021; Nielson, 2020).
Mental health issues present an important global public health and social problem (Roberts, 2018). The COVID-19 pandemic has led to increased stress and anxiety levels, as well as more mental health concerns (Reardon et al., 2021). Mental illness is usually linked to weakness (Love, 2018), and professional athletes are more likely to feel stigmatized (Lavelle, 2021) and reluctant to talk about their struggles. However, we witnessed a growing number of athletes speaking out on mental health, shedding the stigma of mental health, and calling for systematic changes for mental health protection. For instance, Kevin Love, a professional basketball player, shared his experiences with mental illness on social media and in a news article (Love, 2018). Royce White, another professional basketball player, spoke publicly about his mental health concerns and asked for changes on how the National Basketball Association (NBA) handles mental health (MacKay, 2018). Although these players may receive negative responses and risk their image as a professional athlete, researchers found the stories about their struggles created a widespread conversation around the health issue and reduced the stigma of mental health (Cassilo, 2022; Mitchell et al., 2018; Parrott et al., 2020).
Our study investigated the case of Naomi Osaka, who posted a statement on her Twitter and Instagram account to announce her decision to withdraw from the French Open due to her mental health concerns. Her decision to exercise self-care over fulfilling her media obligations created a wide range of conversations around mental health and its importance (Wallace, 2021). While many athletes hesitate to reveal their mental health issues, Naomi Osaka advocated that athletes suffering from mental illness does not mean they are weak. She expressed her willingness to work with the Tour and “make things better for the players, press and fans” (Osaka, 2021b, para. 1). Media reported that her behavior can be seen as a form of advocacy (Gregory, 2021a) to reduce the stigma of mental health issues (Gavin, 2021; Wallace, 2021) and raise awareness of mental health (Van Raalte, 2021). Inevitably, Naomi Osaka received criticism for her actions. As a professional tennis player, Naomi Osaka refused to talk to the press, which was considered a contractual obligation, and eventually withdrew from the tournament. Some journalists criticized Naomi Osaka for using mental health as an excuse and attacking the media (Suliman, 2021). As such, media coverage of the incident also revealed a conflict between work responsibility and mental health, which drew discussions around handling mental illness in the workplace.
Reactions to Athletes’ Social Activism and Mental Health Disclosures
A growing literature has investigated how the public perceived and responded to athletes advocating for a variety of social issues. As social media has become an important platform for athletes to speak directly to their fans and create conversations (Hoffarth, 2019; Hsu, 2021), previous studies collected user comments online or from social media to explore themes around various athlete activism contexts. For instance, Gill (2016) collected fan-generated media comments and discovered divisive opinions on the five St Louis Rams players who staged a pregame protest “hands up don’t shoot.” Some fans sent out harmful comments towards the athletes and thought they should be punished, while others supported athletes’ protest (Gill, 2016). Frederick et al. (2017) examined how people reacted to the University of Missouri football players’ activism regarding racial injustices. They discovered divisive opinions throughout the emergent themes. While many praised the athletes and showed support for their actions, some suggested advocacy was incompatible with sport and student-athletes should not engage in such activities. Miller and Behm-Morawitz (2017) looked into LGBTQ issues and found that the public’s response to the coming out interview of Diane Sawyer and Caitlyn Jenner were mostly positive. Mitchell et al. (2018) examined the conversations about athletes revealing their eating-disorder. The authors found that comments were supportive towards the athletes and suggested the athletes brought awareness to the issue.
To understand the narratives surrounding mental health in sport, a growing body of research has examined media frames (e.g., Cassilo & Kluch, 2021; Lavelle, 2021; Parrott et al., 2021) and social media user responses (Parrott et al., 2020) to mental health disclosures by professional and collegiate athletes. Existing studies discovered positive news frames, large support, and praise for the athletes within media coverage (e.g., Cassilo & Kluch, 2021; Parrott et al., 2020). Press articles also humanized the athletes and portrayed these athletes as mental health advocates (Cassilo & Kluch, 2021). These studies revealed the potential role of athletes in mental health promotion and stigma reduction. Researchers suggest that other forms of data, including audience reaction on social media to mental health, should be examined to provide different insights about the issue (Cassilo, 2022). One previous study included social media user comments and found encouragement, personal experience, thanks, advice, and strength in the conversations around the mental health disclosure tweets posted by athletes (Parrott et al., 2020). Based on the extant literature and the various themes that may exist, we propose the following research question:
What are the main themes that emerged from Twitter users’ comments regarding Naomi Osaka’s decision to withdraw?
Sentiment Analysis in Sports
Sentiment analysis, also called opinion mining, analyzes people’ opinions, sentiments, and attitudes from written languages towards subjects such as events, topics and issues (Liu, 2012). The tool has been used to understand how the public react to a variety of social events such as the Me-Too Movement (e.g., Bandi & Fellah, 2019; Kaur & Sharma, 2020), school shooting (Sailunaz & Alhajj, 2019), and COVID-19 (e.g., Oyebode, et al., 2021). Findings from sentiment analysis can help policy makers and related organizations to document public understanding of the issues and develop strategies to effectively address them.
In sport-related contexts, several researchers have utilized sentiment analysis to facilitate understanding of public opinions (e.g., Monaco, 2020; Wang & Sant, 2022; Workewych et al., 2017). For instance, Wang and Sant (2022) used sentiment analysis to analyze Instagram posts of athlete protests. They found that media coverage of the protests was sensitive to social media user interest and sentiment. Monaco (2020) conducted a sentiment analysis of Twitter user posts to assess whether transgender female student athletes should not take part in female competitions. The results showed divisive opinions on female transgender athletes competing in female sporting activities and suggested there was still a long way for the acceptance of transgender athletes. In another study, Workewych et al. (2017) analyzed tweets related to traumatic brain injuries in sports and their results revealed mixed sentiments towards the issue. Findings of the study demonstrate public misunderstanding of concussions and present a new direction for injury prevention and education. Parrott et al. (2020) explored how social media users responded to athletes’ mental health disclosures and classified user comments into positive, neutral, and negative. By comparing positively-valenced comments and negatively-valenced comments, the authors found the athletes receive overwhelmingly positive responses, which may reduce mental illness stigma and encourage people to open up on their mental struggles. At the same time, the insights on the negative comments show that social media communication can be hostile and mental illness still remains stigmatized. As such, sentiment analysis can offer useful empirical evidence in understanding the public opinion about a contested topic. Therefore, we propose the following research question.
What are the sentiments of Twitter users’ replies and tweets regarding Naomi Osaka’s decision to withdraw?
Message Source as a Predictor of Consumer Response
Sports fans’ reactions to messages may differ depending on the source (Schultz et al., 2011). Research has shown that media sources can evoke different perceptions (e.g., Morris, 2017; Na et al., 2020) and subsequent social media responses (e.g., Na et al., 2020; Schultz et al., 2011). Wang and Sant (2022) explored media coverage of athlete protests on social media and found that media’s decision to cover the protests were sensitive to social media user interest and sentiment. The approaches of media covering or delivering the news can also influence social media user sentiment. In the present study, we considered messages coming directly from athletes as well as from the media coverage of the case (e.g., ESPN’s Sports Center). Audiences may perceive messages coming directly from celebrities’ social media account to be more genuine than messages coming from mass media (Levinson, 2011). Receiving the messages directly from the celebrity could also lead to higher levels of parasocial interactions than receiving them from other sources (Lee & Jang, 2013). Parasocial interaction describes the one-sided interaction between audience and media figures, which is an illusionary experience as if they are talking with each other (Horton & Wohl, 1956). Research has shown that parasocial interaction increases audience’s willingness to perform a supportive action (Lee & Jang, 2013), and willingness to forgive athlete’s misdemeanor (Lee et al., 2018). Therefore, public figures like to be active on social media and talk directly with the public to garner more favorable attitudes (Dunn, 2018; Utz, 2009).
Stronger parasocial interaction with a celebrity or a television persona who has suffered from mental illness was found to be associated with lower mental illness stereotypes, lower social distance from people with depression, and greater willingness to seek treatment for depression (Hoffner & Cohen, 2012; 2018). Previous research in mental health communication highlights the function of parasocial interaction as an avenue for stigma reduction and health promotion (Parrott et al., 2020). Parrott and colleagues (2020) analyzed replies to the mental health disclosure tweets posted by athletes and examined the replies that tagged the athletes, which represented some levels of presumed parasocial interaction. The results revealed largely positive responses on Twitter, which indicated the potential function of parasocial interaction on social media. However, it remains to be examined whether the similar parasocial interaction might occur when the Twitter handle is not the athlete who is struggling (e.g., media outlet). Our study expands on the previous studies by collecting Twitter users’ replies to Naomi Osaka’s tweet and the retweets of her decision, as well as comparing user responses across different message sources. According to previous research (Dunn, 2018; Lee & Jang, 2013; Utz, 2009), we assume that receiving the news directly from the athlete’s Twitter handle creates higher levels of parasocial interaction, which subsequently leads to more positive and supportive responses. Thus, we expect audiences may be more supportive and positive when they receive the news directly from the athlete than from the media. In this way, we not only discover how individuals frame the incident, but also extend previous research by providing further empirical evidence of the message source effect and the potential role of parasocial interaction in social media communication. Therefore, we propose the following research question to investigate whether the themes and sentiments of the replies across different Twitter handles (Naomi Osaka’s personal account, other athletes’ accounts, and sport news outlets) show any different patterns:
Do themes and sentiments differ depending on the source of Naomi Osaka’s statement on Twitter?
Methodology
Data Collection
We collected replies and tweets on Twitter to explore the public’s reactions to Naomi Osaka’s statement. Twitter serves as a forum where people send out their opinions and debate on a variety of topics (Burton, 2019). Twitter data provides informative insight into users’ opinions, which can facilitate implementation of strategies (Workewych et al., 2017). Data collection was conducted over a 10-day period beginning May 31, 2021, the date when Naomi Osaka posted a statement on Twitter to announce her withdrawal from the French Open. To facilitate data collection, we acquired a Twitter Academic Research API to access Twitter data with Python software 3.8. We used tweepy, an open-source Python package that can easily access the Twitter API. We used the Twitter attributes in_reply_to_status_id_str to scrape replies of tweets.
We collected replies from different sources to generate a larger dataset and explore the effect of source factor. First, we focused on the replies to Naomi Osaka’s original tweet that contained her statement (2338). Next, we selected two other athletes, Steph Curry and Lewis Hamilton, who were the first ones to retweet Naomi Osaka’s statement to their followers (700). These two athletes are influential male athletes (with 15 million and 5 million followers respectively on Twitter) in different sports (i.e., basketball and motorsport), and their retweets received considerable replies. For media outlets that retweeted Naomi Osaka’s withdrawal case, we selected two popular and reputable sports news outlets, Bleacher Report and ESPN Sports Center, who neutrally reported the event and cited Naomi Osaka’s statement in their Twitter account (426). We found most of the users left a comment on the same day or the next day when the athlete or a third party (re)tweeted the statement. After completing data collection, we scraped a total of 3464 replies. In addition, we also collected tweets that contained the hashtag #NaomiOsaka during the time of data collection period (May 31, 2021 to June 9, 2021) to broaden our data set. We limited the time frame to 10 days to make sure the tweets collected were relevant to the breaking news associated with Naomi Osaka’s statement to withdraw from the tournament. Previous social media studies have used hashtags to collect data and discover users’ sentiment on different events (e.g., Mohammad & Kiritchenko, 2015; Sailunaz & Alhajj, 2019). We collected a total of 10,000 tweets containing the hashtag #NaomiOsaka during this time period. All tweets were stored in a CSV file for future processing.
Data Cleaning and Preprocessing
Preprocessing data is a key step for any Natural Language Processing and makes automatic sentiment analysis more precise (Bandi & Fellah, 2019; Liu, 2012). We removed duplicate data and replies and tweets that were in another language. Then we deleted content that was completely unrelated such as commercials. After the cleaning process, the complete dataset contained 5563 cases. In addition, we removed URLs, stop words, and unnecessary spaces from the text. We used Regular Expression (re) package and Natural Language Toolkit (NLTK) package to preprocess data and stored the preprocessed data in CSV files.
Data Analysis
Thematic analysis
We conducted an inductive thematic analysis with each tweet or reply serving as the unit of analysis. The dataset was analyzed using Leximancer (version 5.0), which is a software that conducts quantitative content analysis using a machine learning technique. It is an effective tool to analyze user comments and social media posts (Bal et al., 2012), and it has been utilized in sport-related social media research (e.g., Frederick et al., 2020; Schmidt et al., 2018a). One of the strengths of Leximancer is that it not only detects the main concepts in a text but also analyzes how they are interrelated to identify their relative importance. The concept map presents the themes in terms of their relevance and connectivity within the dataset (Leximancer Pty Ltd, 2021). Those at or above 50% are extracted as emergent themes. Leximancer produces Hits, the number of text blocks associated with the theme and the software ranks the themes by their relative importance. Themes on the content map are heat-mapped, where hot colors denote the more important themes, and cool colors denote the less important themes. After utilizing Leximancer as the initial thematic analysis, we reviewed the content in each theme to better understand the data and renamed the themes based on the content and the context of this study (Schmidt et al., 2018a). We read through all of the examples in each theme and the concept seed words generated by the software. For instance, the first theme showed “media”, “athletes”, “job”, and “speak” as seed words, and many of the comments in this theme were about speaking to the press as part of athletes’ job but had mixed views. After reviewing the seed words and examples, we renamed the first theme as Debates on Press Conference.
Sentiment Analysis
We used sentiment analysis to evaluate the overall valence of the tweets and replies to Naomi Osaka’s case. We further analyzed sentiment for each theme we identified. When user comments and replies draw various views and conversations around the topic, it is important to conduct sentiment analysis following thematic analysis to further examine users’ emotional valence about the topic. Based on the size of our dataset, we used a lexicon-based approach, which is an efficient and widely adopted approach dealing with larger data sets (e.g., Wunderlich & Memmert, 2020). In this study, we used TextBlob, a lexicon-based model to classify tweets based on their sentiment polarity. TextBlob is an open-sourced Python library built upon NLTK. It is a rule-based approach using a predefined set of categorized words. The textblob.sentiments module’s default implementation is PatternAnalyzer, which is based on the pattern library (Loria, 2020). Sentiment scores are calculated based on semantic relations and the frequency of each word in a sentence. If the polarity is less than 0, the sentiment of the tweet is considered as negative. If the polarity is larger than 0, the sentiment is considered as positive. Otherwise, the sentiment is neutral. The polarity of each tweet was recorded in the CSV files.
Results
Thematic Analysis Results
The five primary themes emerging from the thematic analysis included (1) Debates on Press Conferences, (2) Mental Health Matters, (3) Hope for a Positive Change, (4) Appreciation, and (5) Support the Decision. As shown in Figure 1, Debates on Press Conferences was the most prominent theme in the dataset, indicated by the reddest color on the map and most hits (1700). Mental Health Matters was the second most prominent theme with 1457 hits, followed by Hope for a Positive Change (923), Appreciation (850), and Support the Decision (553). The numbers of hits of each theme are shown in Table 1. The concept map of themes identified by leximancer. Hits of Themes (Overall).
Debates on Press Conferences
The replies and tweets within the theme Debates on Press Conferences highlighted the public’s different opinions on whether athletes should be obligated to talk to the media or not. On one hand, people believe that athletes’ job is to play the sport but not to answer the press. For example, “An athlete’s job is to play a sport,” and “Athletes have an obligation to play sport but not to talk to the media.” Some users also thought tennis tournaments should not put a clause in player contracts to speak to the media. One user posted, “The CONTRACTUAL OBLIGATION for athletes, artists, etc., to speak with the media is insane.” These users questioned the functions of post-game interviews by asking if anyone watches players’ interviews or if the skills of answering questions influence players’ rankings. One commentator claimed that post-game interviews “are the most cliched, say-nothing, meaningless, formulaic, traditionalist, patriarchal rituals one could find.” Press conferences in sports need changes, for example, “the questions they ask and how long they can hold the player needs to be looked at,” “let athletes choose if they want to speak to the press or not,” or at least “have more control over the questions allowed.”
On the other hand, many people also believed that athletes should fulfill their media obligations. Many users posted “media is part of the job” and “do your job” under this theme. Press conferences were also considered to be an indispensable part of the tennis tournaments and any other sports because the media helped increase the exposure and revenue potential of the athletes and these sporting events. Users tried to highlight the importance of player interviews by asking questions, “do these players know how these massive tournaments get funded” and “if athletes don’t have to talk to the press do events and sponsors not have to pay prizes?” They also claimed that the media helped these athletes to get famous, “You don’t want to talk to the media, don’t be a sports star.” As such, much of the comments under this theme were debates around player interviews at the tournament.
Mental Health Matters
The Mental Health Matters theme emerged as the second most prominent in the dataset. While the first theme was focused on the debate about media obligations, the second theme was focused on the reason that Naomi Osaka decided to withdraw from the French Open. Under this theme, mental health was believed to be a critical health issue. Many users claimed that “mental health is more important,” “mental health is as important as physical health,” and “it is important to be your own advocate for physical and mental health.” Some users also highlighted that it is not easy to speak out on mental illness, “Naomi Osaka’s public admissions of mental health challenges carry important lessons for us all.” A number of users also shared their personal experiences and expressed sympathy to Naomi Osaka. It was not surprising that many commentators were or had experienced mental health issues such as anxiety or depression. For example, “As someone who has suffered from extreme depression and anxiety, I understand what you’re going through,” and “If you don’t have a close connection to someone with mental illness (yourself included), then you can’t truly understand.” Regarding public speaking, one user posted “I’m an introverted lawyer who hates the public speaking part of it” and another user commented “I too am an introvert and I understand how speaking up and out can trigger anxiety.” Meanwhile, some people criticized how the sports authorities treated their athletes. “I am so angry right now that I can’t watch my favorite player in the French Open because they don’t know how to treat their players dealing with mental health issues.” Other users also reflected the importance of mental health issues in a broader work environment, “Naomi Osaka’s plight with mental health issues highlights a thorny area for many companies in dealing with employees.” Overall, findings showed that users were sharing and discussing various issues related to mental health and its importance under this theme.
Hope for a Positive Change
Another theme emerged from the dataset was the Hope for a Positive Change. Under this theme, users expressed their hope to see changes of the current situation – including the athlete’s mental health status and mental health problems in our society and professional sports. Some people wished Naomi Osaka to get better soon, while some users looked for increasing awareness of mental health and acceptance towards people struggling with the issue. For example, one user hoped that Naomi Osaka’s actions could “bring awareness to how vital it is to prioritize our mental health.” A commentator posted “I think (I hope!) we’re going to reach a point that when a person says they’re in mental distress, we will just…listen.” Users also wished this entire discussion could “shine the much-needed light on how much access we have for our mental health.” Furthermore, the public expected the French Open and other sports authorities to take actions and make systemic changes to protect athletes. Users hoped “the executive at the French Open TAKE NOTE” and “the officials of the French Open understand” after Naomi Osaka clearly expressed her feelings and concerns. The organizations needed to “see how much stress they are putting on their people” and the tennis world should start “showing some sensitivity soon.” A user also claimed, “Just saying, journalists and senior leadership at the grand slam tournaments need to do better.” Some fans would like to see the athlete work with the tour to “help them value their athletes’ mental health better when it comes to press interviews”. The mandatory press showings during a tournament should be changed without forcing the players to do interviews. Overall, people were looking forward to positive changes in the athlete’s health status, mental health awareness, and policy concerning mental health in professional sports.
Appreciation
Many users expressed appreciation towards Naomi Osaka for speaking out on an important health and social issue. They were thankful that the player took a stand for mental health and against the tournament. For example, “Thank you for taking care of yourself and being an awesome model of self-care for young folks,” “I applaud you for your courage to stand up for yourself and not be bullied by the tennis establishment,” and “Thank you for taking a stand against a powerful sports organization.” They praised Naomi Osaka’s bravery and courage to reveal her mental health issues and prioritize her self-care. Naomi Osaka’s experience was what many people could resonate with, and her actions encouraged those who were suffering to take care of themselves and access professional help.
Support the Decision
The final theme emerged from the dataset was Support the Decision. Under this theme, users openly expressed their approval and support for Naomi Osaka’s decision to prioritize her mental health. A number of users posted “I support your decision” and showed respect and love to Naomi Osaka at the same time. The majority of the comments within this theme also indicated that users believed Naomi Osaka made the right decision, which was “the best for her peace of mind” and “sparked a controversy and drew attention to the topic.”
Comparison of Replies from Different Sources
Hits of Themes among Different Message Sources.
Note. 0 in the table represents that the theme did not appear to be primary in the dataset. Percentage = Hits/Total replies within the message source.
Sentiment Analysis Results.
Sentiment Analysis
TextBlob classified sentiments of the tweets and replies based on their polarity and labeled each one as positive, neutral, or negative sentiments. Among the total of 5563 tweets and replies, 51.4% were positive, 29.3% were neutral, and 19.3% were negative. These tweets and replies include positive words and phrases to convey their support, appreciation, and best wishes to the athletes or exhibit hope for mental health. The messages that expressed negative sentiments appear to contain negative words and phrases towards Naomi Osaka, the French Open, or mental health issues. For example, some users reacted negatively to Naomi Osaka, saying “you don’t want to talk to the media, don’t be a sports star”, “such a joke rich people are so soft,” and “you are a public figure, and this note was rude to all tennis fans.” Some users criticized the tennis authority as “greedy French Open and sports huckster organizations.”
The sentiments of replies across different message sources showed similar patterns, as shown in Figure 2. While the comments from all message sources generally appear to be more positive sentiments than neutral and negative, sentiments of sports media outlets’ retweets seem to be less positive and more negative than the other sources. To examine if sentiments were significantly different across the message sources, we conducted a Chi-Square Test with Python. A chi-square test of independence showed that there was a significant association between the type of message source and the sentiments of the replies, χ2 (4, N = 3063) = 54.18, p < .001. Sentiment analysis results of replies among different sources.
Discussion
The present study used thematic analysis and sentiment analysis to analyze replies and tweets responding to Naomi Osaka’s withdrawal from the French Open due to mental health concerns. This study advances our understanding of fan reactions to athletes’ mental health disclosures by investigating multiple sources of Twitter handles to examine whether emerging themes and sentiments differ depending on the source. Compared to previous research that explored athletes’ mental health disclosures (e.g., Cassilo, 2022; Parrott et al., 2020), this case presents a unique context that Naomi Osaka is a woman of color competing in a traditionally White sport. Moreover, her decision to withdraw from the tournament was so controversial that it immediately drew media backlash. The tournament’s governing body issued her a fine for not speaking to the press which escalated into heated debate around the issue. Considering how the case has unfolded, our findings contribute to the existing body of work by showing how audience reactions differed based on who was (re)tweeting the breaking news. Specifically, we examined the differences in the themes and sentiments among different message sources (athletes vs. media) which further illuminates different response patterns depending on the source of message. Our findings highlight the potential role of parasocial interactions in health communications narrated by athletes. When athletes were tweeting out the breaking news, significantly more comments were centered on showing empathy and support toward the athlete than when sport media handles retweeted the same breaking news. The findings provide recommendations for athletes and relevant stakeholders of utilizing social media and source factors to reinforce activism efforts.
Our data shows that a significant number of users offered support (15.5%) to Naomi Osaka and appreciated her for speaking up (10.1%). In addition, the overall valence of the replies and tweets were more positive (51.4%) than negative (19.3%). By investigating an athlete in different sport and identity groups, our study augments previous findings (Cassilo & Kluch, 2021; Parrott et al., 2020, 2021) that narratives around athletes’ mental health disclosures are largely positive. The athletes in the previous studies (e.g., Kevin Love and DeMar DeRozan) mainly shared their own experiences and opinion about mental health (e.g., Cassilo, 2022; Parrott et al., 2021). In our study, Naomi Osaka chose to prioritize her mental health over fulfilling her media obligations and work responsibilities as a professional tennis player. The incident created a conflict between mental health and work responsibility. Nevertheless, our results show that Naomi Osaka’s decision still received a lot of support and appreciation on Twitter. Findings further illuminate positive reception of athletes opening up on mental health struggles and triggering more conversations around mental health issues. As such, findings of our study add to the literature by showing how professional athletes’ behavior can be a form of advocacy to push for a positive change (Gregory, 2021a).
While speaking out on health issues is considered a non-controversial and safe activism (Cunningham & Regan, 2011; Schmidt et al., 2018b), Naomi Osaka’s decision to exercise self-care drew heated debates. It is important to note that Naomi Osaka’s statement was intending to explain why she withdrew from the tournament by disclosing her mental illness rather than directly advocating for mental health. Our data showed that Debates on the Press Conferences (31.0%) as the most prominent theme, but at the same time, Mental Health Matters (26.6%) emerged as the second major theme. The conversations had largely focused on the athlete’s controversial action (withdrawal from the tournament) itself and the main cause (mental health) of the action. Athlete activists employ different methods and strategies to advocate for social causes. When the methods are seen as controversial, there might be an overwhelming focus on the method of advocacy over the message (Coombs et al., 2019). For example, Colin Kaepernick protested racial injustice by kneeling during the national anthem, which ignited heated debates around his behavior (e.g., disrespect for flag) and nationalistic narratives (Carpenter, 2016). Compared to protests on the field (e.g., hands up don’t shoot, kneeling during the anthem), Naomi Osaka’s behavior has a stronger link with the cause. Although critics questioned or disagreed with her decision, the heated debates revealed the conflicts between work duty and mental wellness, and encouraged reflection on mental health in the workplace. The conversations moved beyond the micro level of athlete’s mental health to larger social and health issues (Cassilo & Sanderson, 2018). One commentator said, “being allowed to set healthy work boundaries based on mental health issues is something we should all contend for.” Another user applauded Naomi Osaka, “you raised an important issue which needed to be heard.” The TIME magazine depicted Naomi Osaka’s stance at the French Open as “a form of advocacy” (Gregory, 2021a, para. 6).
The present study advances our understanding of source factors on social media user interest and sentiment (e.g., Na et al., 2020). Thematic analysis across different sources showed that a significantly larger number of users expressed support and appreciation to athletes’ tweets and retweets. In contrast, comments on the retweets from the media coverage of the event were more focused on debating whether athletes should be obligated to talk to the press or not. We draw this finding to the notion of parasocial interaction, which respondents feel more empathy towards the celebrity when they receive messages directly from the person’s social media accounts (Lee & Jang, 2013). Hearing directly from the person who struggles (and who shares empathy) might have evoked feelings that they are talking directly to the athlete and are thus more likely to view them as personable. Sanderson and Emmons (2014) also highlighted that humanizing the athlete allows fans to express sympathy and support even when the athlete is involved in a controversy. Therefore, our findings contribute to the literature that the same breaking news on social media can have different ramifications on the public discourse depending on the source of the news. Athletes posting stories in their own social media accounts can be powerful in facilitating parasocial interaction (Lee & Jang, 2013), athlete-fan interactions (Kassing & Sanderson, 2010), and initiating important conversations about social issues directly to their followers (Schmittel & Sanderson, 2015; Utz, 2009).
Rather than releasing a statement to mainstream media or holding a press conference, Naomi Osaka announced her decision to withdraw from the tournament on her Instagram and Twitter accounts. She was the first person to break the news to the public. Our study supports the notion that modern-day athletes wanting to own their story (Hoffarth, 2019; Hsu, 2021). Similar to Mitchell et al.'s (2018) findings, our study shows that sharing athletes’ own struggles to the public was influential in drawing a wide range of conversations about the cause and made strong connections with users. Our findings also add to the health communication literature (Chen & Bell, 2022) that first-person narratives, compared to third-person point of view, facilitate message processing and persuasion by increasing perceived identification with character. People can relate more with the character by vicariously experiencing the health threat described in the message. Therefore, athletes sharing their struggles in fact can create feelings of ‘being the character’ among audience. Public health organizations and athletes can develop strategic collaborations that highlight first-person point of view to create an authentic and persuasive storytelling to emphasize the issues and call to action. Such collaborations can create authentic voice that highlights precise social benefits and focus on substantial activist efforts rather than general and vague social benefit claims (Vredenburg, et al., 2020).
Limitation & Future Studies
There are several limitations in our research. First, we used Leximancer and TextBlob to automatically analyze textual data rather than human coding. The data analysis tools might not capture all the nuances and meaningful themes in the dataset, and automatic sentiment analysis is less accurate than human coding due to misinterpretation. When we compared responses among different sources, no consumer characteristics were considered because we do not have access to the Twitter user’s personal information. It is plausible that differences in the Twitter user characteristics from different sources might have affected the results. Followers’ characteristics of the ESPN Sports Center and Bleacher Report might be different from those of Naomi Osaka’s followers. In addition to individuals’ demographic variables, those who find mental health to be personally relevant could also be more likely to show support. Therefore, these limitations present potential future study opportunity to examine how different demographics and consumer segments respond to athlete controversy or activism by conducting surveys or running experiments. The findings can provide more specific recommendations for athletes and sports organizations on tailoring communication messages when they target different audiences.
In the era of new media and athletes as content creators (Hoffarth, 2019; Hsu, 2021), another interesting avenue for future research is to investigate different types of storytelling on digital platforms in promoting social causes. For example, future studies can examine how audiences react to athlete mental health documentaries (e.g., Quiet Storm: The Ron Artest Story), which could have significant influence on mental health. Future studies can also explore different elements of a story or content that increase effectiveness of advocacy. Furthermore, as athletes are increasingly controlling their narratives (Hoffarth, 2019) using social media, it would be interesting to explore the relationship between athletes and news outlets and the role of news outlets in athlete activism.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
