Fandom Thread:
The Roles of Fan Identification and Social Media in Crisis Communication
Elia (Simiao) Chen 12049514 Master’s Thesis
Graduate School of Communication Master’s programme: Communication Science
Corporate Communication University of Amsterdam
Supervisor: Dr. James Slevin June 25, 2019
Abstract
In the current day and age, in which fans of media products can unite through social media and express their collective attitudes regardless of their geographic locations, entertainment companies are getting increasingly wary when dealing with corporate crises, for there might be serious consequences if certain decision is made. The purpose of this paper is to provide recommended solutions to this problem, by studying the association between fan
identification and account acceptance of certain crisis communication strategies, while also touching upon the role of social media in the process. Two research questions, namely “to what extent is the level of fan identification associated with the level of account acceptance of the rebuild strategies deployed by companies during preventable corporate crises?” and “how is the association between fan identification and account acceptance mediated by social media use, secondary crisis communication, and social media opinion climate?” have been proposed to guide the research. A survey study focusing on the real-life crisis example of Netflix and the Kevin Spacey scandal has been conducted to find out the answers. It has been identified that there is no direct association between fan identification and account
acceptance under the circumstance. However, fan identification does have an indirect influence on account acceptance through secondary crisis communication and social media opinion climate. Taking effect size into consideration, the conclusion that fan identification negatively predicts account acceptance through secondary crisis communication can be made. Consequently, it is recommended that traces of secondary crisis communication and social media opinion climate in certain communities should be closely monitored by the company, to better deal with the situation and prepare for what is to come.
Keywords: fan identification, account acceptance, crisis communication, social media, entertainment industry
Introduction
Corporate crisis is always the worst nightmare for any company, whichever industry it is in. When companies are going through corporate crises, the timing, the message, and even the way of delivering that message are all crucial when announcing the response. It is known that crisis communication has a tight connection with corporate reputation in the mind of the public (Coombs & Holladay, 2008; Van Der Meer & Verhoeven, 2014). Thus, it is always a tough and somehow sensitive decision, considering which aspects to take into account when making the response—especially for entertainment companies, who rely greatly on fan base (Grady, 2014; Liang & Shen, 2016; Harriss, 2017). Fans are quite different from the general audiences, as they are a relatively extreme group of stakeholders who have strong feelings towards certain objects produced by the companies, and/or celebrities participating in the production, and they are likely to perceive certain situations completely differently (Click, Lee, & Holladay, 2013; Brown, Brown, & Billings, 2015; Groene & Hettinger, 2016). As a result, they may react more radically to the decision, compared to the general public (Brown, Brown, & Billings, 2015). They can use the power of their community to express their strong emotions, and pressure the company into doing what they would have wanted if they feel the decision is inappropriate—particularly nowadays, through the use of social media to spread their influences, which can be fairly devastating for the companies (Pearson, 2010; Fillis & Mackay, 2014; Brown, Brown, & Billings, 2015). Most of the time, under these
circumstances, companies are likely to further apply the rebuild strategies, as explained in the Situational Crisis Communication Theory (SCCT) by one of the most important figures in the academic world of crisis communication Timothy Coombs, to acknowledge the crisis
situation and hope for public acceptance, while in the meantime, generate new reputational assets to rebuild its image (1995; Coombs, 2007; Coombs, 2011).
What may be the connections between the perception of whether a rebuild strategy is effective and the level of fan identification then? How, if possible, can companies prevent these “online riots” on social media? Is there anything companies can learn to do to better serve this group of their very loyal customers when these unfortunate events happen? Finding answers to these questions is of great societal and practical relevance, as it will not only provide managerial insights for companies during crises, but also help them to better understand this group of consumers and the role of social media in this process for future business development.
Towards this aim, this paper is going to provide feedback on how companies can be more prepared to combat situations like this, by presenting a survey study, focusing on a real crisis example involving a celebrity scandal—Netflix and its response regarding the Kevin Spacey crisis. There are two research questions guiding this study:
RQ1: To what extent is the level of fan identification associated with the level of
account acceptance of the rebuild strategies deployed by companies during preventable corporate crises?
RQ2: How is the association between fan identification and account acceptance
mediated by social media use, secondary crisis communication, and social media opinion climate?
This study is also of vital scientific relevance. While a wide range of studies have been conducted in the separate realms of fan identification, crisis communication, and social media, very little is known in terms of how they are in connection to each another—even though some articles have looked into the relationship between crisis communication and social media (Roshan, Warren, & Carr, 2016; Kim, Zhang, & Zhang, 2016; Jahng & Hong, 2017; Du Plessis, 2018; Zhu, Anagondahalli, & Zhang, 2017), whereas some of the others have taken on the perspective of fan identification and social media (Lukach, 2012; Click,
Lee, & Holladay, 2013; Mudrick, Miller, & Atkin, 2016; Billings, Qiao, Conlin, & Nie, 2017; Wakefield & Bennett, 2018). This study has not only combined these areas and proved that they are tightly related, but also further expanded existing research done in respective fields. Moreover, although Brown and her colleagues have done several studies integrating the above-mentioned concepts, they have been focused on sports, which has also been the centre of attention for fan culture studies for a fairly long time, and this study has gone beyond the traditional territory and applied these insights onto the context of entertainment fandoms (Brown, 2014; Brown, Brown, & Billings, 2015; Groene & Hettinger, 2016).
Following this section, an overview of theoretical framework and proposed
hypotheses will be provided. Introduction of methodology will come after, including further information on the particular illustration, followed by analysis and results parts. Lastly this paper is going to conclude with discussion and conclusion sections.
Theoretical Framework Fan Identification
Technically speaking, everyone is a fan of something. Fan identification stands for personal commitment, as well as emotional involvement and identification one has with a certain interest (Sutton, McDonald, Milne, & Cimperman, 1997; Click, Lee, & Holladay, 2013; Groene & Hettinger, 2016). This interest can be anything, ranging from sports teams and books, to celebrities and hobbies, but as far as this study is concerned, media objects as that fan interest is the key in the setting of entertainment industry (Reysen & Branscombe, 2010; Click, Lee, & Holladay, 2013; Harriss, 2017). People who have the identification are fans, and fans form fandoms—the sub-cultured communities based on shared love, support, and devotion to that interest (Bird, 2002; Thorne & Bruner, 2006). Media or entertainment fandoms are still relatively under-studied by the academic world, compared to sports and
celebrity fandoms (Consalvo, 2003; Maltby, Day, McCutcheon, Houran, & Ashe, 2006; Schimmel, Harrington, & Bielby, 2007; Groene & Hettinger, 2016). Although it seems that on the level of personal emotions and social interactions, sports and media fandoms are quite similar to each other, there are still some significant differences between them (Schimmel, Harrington, & Bielby, 2007). While sports fans rely more on social identification, media fans are more predominant when it comes to personal identification (Thorne & Bruner, 2006; Fillis & Mackay, 2014; Markman, 2017).
According to Sutton and his colleagues, there are three different levels of fan
identification: low identification (social fans), medium identification (focused fans), and high identification (vested fans) (Sutton, McDonald, Milne, & Cimperman, 1997). They vary mainly in terms of devotion, commitment, loyalty, and the level of active involvement (Sutton, McDonald, Milne, & Cimperman, 1997). These categories were first proposed by Sutton and his colleagues and applied in the context of sports team fans (1997), and later on, Milne and McDonald expanded the horizon to the general concept of fan identification (1999). Since then, other media and entertainment fandom scholars have also taken on this classification, and put into use in other contexts (Gwinner & Swanson, 2003; Liang & Shen, 2016). Yet, one can also argue that one’s status is never static, and this scale should be seen as a continuum instead of rigid segments. For instance, one can transfer from a social fan into a focused fan by becoming more devoted throughout the process, and at the same time, vested fans may also lose interest as time goes by and become social fans.
Besides the money, time, and energy they are willing to devote to their interest, what else may be different for fans with various levels of identification?
Corporate Crises
Nobody likes the sound of a “crisis”. Corporate crises are unexpected and abrupt events that threaten the organisation’s daily operations, reputation, and financial performance
(Coombs, 2007). During the crises, a wide range of stakeholders are affected, whether “physically, emotionally, and/or financially”, and the stakeholder group members can vary from employees and customers, all the way to stockholders and the general public (Coombs, 2007, p. 164). Based on Coombs’ Situational Crisis Communication Theory (SCCT), there are mainly three types of corporate crises: victim cluster, accidental cluster, and preventable cluster, and they are categorised upon the level of corporate responsibility in a crisis (2007). Victim cluster has the weakest link with the actual crisis and it stands for the fact that
organisation itself is also a victim, for instance, during the time of natural disasters or product tampering; accidental cluster refers to the minimal responsibility that is unintentional and cannot be 100% controlled by the organisation, such as technical error; preventable cluster has the strongest association between the organisation and the crisis, as it involves human error and is considered purposeful (Coombs, 2006; Coombs, 2007).
As a result, whichever type of the corporate crises takes place, post-crisis
communication is essential for companies, especially regarding the management of the likely damaged reputation (Coombs, 1995; Coombs, 2006). Crisis response strategy is the wording and actions that are communicated and taken by the organisation after a crisis, to repair damaged corporate image as well as reputation and combat other possible negative
consequences (Coombs, 2007). As Coombs proposed in his article, there are five categories of crisis response strategies, namely nonexistence, distance, ingratiation, mortification, and suffering (1995). Later on, Coombs developed his own work, and regrouped them into three typologies: deny, diminish, and deal (2006). Deny strategies seek to remove the associations between the corporate and the crisis, while diminish strategies focus on trying to make the public believe that the crisis is not so bad as many people may think (Coombs, 2006). Deal strategies go even further, aiming to provide new materials and relevant aids to the victims with the hope of rebuilding reputation (Coombs, 2006). Deal strategies are also known as the
rebuild strategies at a later stage in SCCT, after further theoretical development achieved by both Coombs himself and other scholars (Coombs, 2007). Conclusively speaking, in SCCT, there are three primary crisis response strategies, which are deny, diminish, and rebuild, and one secondary strategy, known as bolstering (Coombs, 2007).
Account Acceptance
Crisis communication strategy is never an isolated island. As a set of actions to deal with crisis initiated by a single party, it does not have practical meaning, unless when combined with its actual effectiveness. This is when account acceptance enters the stage. Account acceptance signifies how people feel about the crisis response provided by the organisation—in other words, the extent of acceptance from the public regarding the response (Blumstein et al., 1974; Coombs & Holladay, 2008). The higher the account acceptance level is, the more appropriate the crisis communication strategy is perceived by the public (Verčič, Verčič, & Coombs, 2019). The account acceptance has a direct link with corporate reputation during post-crisis period (Coombs & Holladay, 2008).
People have similar reactions to sympathy, compensation, and apology strategies (Coombs & Holladay, 2008). In addition, studies have shown that rebuild strategies, such as apology and compensation, are more effective compared to diminish strategies, such as underplaying the seriousness of the situation (Claeys, Cauberghe, & Vyncke, 2010; Van Der Meer & Verhoeven, 2014). According to SCCT, following the guidelines and applying matching strategies under different circumstances should result in a relatively higher level of account acceptance (Coombs, 2007; Cha, Suh, & Kim, 2015). However, some studies have proposed critical evidence, suggesting that matching strategies in SCCT may not always be the best options (Claeys, Cauberghe, & Vyncke, 2010; Claeys & Cauberghe, 2014). To better explain this, besides possible methodological differences, it has been found that several other factors are also at play when determining the level of account acceptance, even when
matching strategy is applied—for instance, the level of crisis involvement, perception of crisis severity, locus of control, and so on (Choi & Lin, 2009; Claeys & Cauberghe, 2014; Claeys, Cauberghe, & Vyncke, 2010; Laufer, Gillespie, McBride, & Gonzalez, 2005).
What else may be the reason behind the different levels of account acceptance then? What other conditions should be met when this occurs?
Fans in Crises
With the development of technology, especially social media, fans, amongst all the other stakeholder groups, are able to be actively involved in every stage of the unfolding of corporate crises (Du Plessis, 2018; Zhu, Anagondahalli, & Zhang, 2017). Notwithstanding, what fans genuinely care about during sensitive and turbulent times can be completely different from what the general public hold dear (Brown, Brown, & Billings, 2015). For example, following the Penn State sex scandal, one of the football coaches of Penn State then Joe Paterno was fired by the university. Although fans acknowledged the unhuman acts conducted by Jerry Sandusky, the central figure of this crisis, they still felt deeply connected with the university and coach Joe Paterno, and they, in a way, turned against the university to back up Joe Paterno, protesting the firing of him on social media, which caused wild social discussion (Brown, Brown, & Billings, 2015).
Various studies have revealed that the level of identification plays a key role in determining account acceptance, especially when a celebrity scandal is involved. Celebrity scandals are usually associated with celebrities’ personal life, concerning certain
misbehaviour that is either illegal or morally unacceptable (Fong & Wyer, 2012). It is known that higher level of identification with the celebrities leads to higher level of trust in their innocence after the scandal broke out (Johnson, 2005). Moreover, prior attitudes towards celebrities have a direct and strong association with the impact people perceive (Fong & Wyer, 2012; Maiorescu, 2017). Fans, to some extent, are an extreme group of stakeholders
who feel strong connections and identification with their interest, and that interest can be the company, the celebrity involved, and/or the media product with the participation of the controversial celebrity. Most of the time, celebrity scandals belong to the classification of preventable crises, since they are improper human conduct that could have been prevented (Fong & Wyer, 2012; Kozman, 2013; Coombs, 2007). Consequently, according to SCCT, rebuild strategy is the best practice after a preventable crisis (Coombs, 2007).
When rebuild strategies are utilised, it generally concerns distancing the company from the central controversial figure, making apologies, and compensating the victims (Coombs, 2007; Coombs & Holladay, 2008). Nevertheless, sometimes this poses as a problem for fans. They would hate to see corporates do this to people and/or media products they love, as can be seen in the Penn State case, and it is very likely that they would take actions to get their messages and emotions across. Based on the aforementioned arguments, the first hypothesis is proposed as the following:
H1. Regarding the rebuild strategies deployed by companies during preventable
corporate crises, the higher the fan identification level is, the lower the account acceptance level is.
Social Media
Social media has permanently changed the media landscape, together with how different players in the society communicate and interact with each other (Van Dijck & Poell, 2013). Having said this, in the field of crisis communication, scholars have long been
focusing on the corporate side, researching how organisations can better utilise this tool to communicate during crisis time, but neglecting what the public can do with it (Schwarz, 2012; Veil, Buehner, & Palenchar, 2011; Kim, Zhang, & Zhang, 2016). It seems that,
notwithstanding, in recent years, this is no longer the case and more voices are heard with regard to the other end of this game (Chung, & Lee, 2016; Zheng, Liu, & Davison, 2018;
Hong & Kim, 2018). It has been established that, in general, the more one identifies as a fan, the more frequently one uses social media to stay updated, always catching up with new things happening around his/her interest subject (Lukach, 2012). Therefore, the association between fan identification and account acceptance may be mediated by the use of social media, and related hypothesis reads as below:
H2a. The negative relationship between fan identification and account acceptance is mediated by social media use.
With the rise of social networking sites, new forms of crisis communication have come into being as well. Compared to traditional crisis communication, which emphasises the role of firms and their crisis response, secondary crisis communication focuses on the crisis relevant information communicated amongst and by the public (Coombs & Holladay, 2014; Zheng, Liu, & Davison, 2018). This has also changed the traditional view on the public being the passive party, only receiving information disseminated by the organisations during corporate crises (Laufer & Coombs, 2006; Avery, Lariscy, Kim, & Hocke, 2010). Nowadays, with the help of social media, the public have the capability and freedom to initiate
discussion online regarding corporate crises, whether expressing opinions or sharing insights (Zheng, Liu, & Davison, 2018). They are no longer solely recipients on the other end of this process, but have become the active players, engaging in, or sometimes even, dominating the conversation (Roshan, Warren, & Carr, 2016).
When one is more passionate about the topic, the more willing he/she is to participate in the discussion on social media (Lukach, 2012). The online environment in which one feels he/she is surrounded by like-minded people plays a determining role in the tendency of engaging in secondary crisis communication as well, and fandoms on social media is simply the best example of that type of environment (Ho & McLeod, 2008; Bennett, 2013; Lothian, 2013). Furthermore, in turn, with more participation in secondary crisis communication, the
attitudes towards the crisis and misconduct are likely to be strengthened (Zheng, Liu, & Davison, 2018). These help to bring up the next hypothesis:
H2b. The negative relationship between fan identification and account acceptance is
mediated by secondary crisis communication.
What’s more, secondary crisis communication helps to not only shape a more comprehensive picture of the crisis, but also build up the public opinion climate on social media (Zerback, Koch, & Krämer, 2015; Zheng, Liu, & Davison, 2018). Social media opinion climate refers to the general opinion climate on the social media platforms, for instance, whether people in one’s network all share the same view on certain issues or not (Schultz, Utz, & Göritz, 2011; Neubaum & Krämer, 2017). Social media users always follow other people, have different voices on their dashboards, and get exposed to the opinion climate created by their versions of the public (Van Dijck & Poell, 2013; Zheng, Liu, & Davison, 2018).
Studies have shown that, in general, the more one feels that other people on social media share the same opinion as him/her, the more likely it is for one to engage in
conversations, the same logic applied earlier in the context of secondary crisis
communication (Moy, Domke, & Stamm, 2001; Ho & McLeod, 2008). This idea is rooted in the spiral of silence theory from mass communication literature, which implies that people are more likely to voice their opinions when they perceive themselves to have the majority view (Noelle-Neumann, 1974). When like-minded people communicate, it is more likely for them to collectively create a more homogenous opinion climate, at least compared to a more diverse one, and this less diverse opinion climate on social media is also likely to further enhance the existing opinion with respect to the account acceptance of certain matters (Noelle-Neumann, 1974; Nekmat & Gonzenbach, 2013; Williams, Mcmurray, Kurz, & Lambert, 2015). Yet again, fandoms on social media are exactly the communities grouped on
the basis of like-mindedness (Thorne & Bruner, 2006; Kellner & Deraa, 2008; Deluca, 2018). Only when different communities dominated by their own single viewpoint interact, can mixed-attitudes emerge and opinion climate shift (Williams, Mcmurray, Kurz, & Lambert, 2015). Thus, the last hypothesis is suggested below:
H2c. The negative relationship between fan identification and account acceptance is
mediated by the diversity degree of social media opinion climate. The conceptual framework is visualised in Figure 1 below.
Figure 1. Conceptual framework.
Methodology Issue Example
There are two main reasons why the example of Netflix and Kevin Spacey crisis is eventually chosen. First of all, House of Cards, the collaboration between those two parties, is a very successful television show with a huge fan base. In order to better measure fan
Fan Identification Account Acceptance Social Media Use Secondary Crisis Communication Social Media Opinion Climate H1 H2a H2b H2c H2a H2b H2c
identification level, a media product that respondents are more likely to recognise and identify with is terribly important. Moreover, the way Netflix handled the situation is very typical of the practice, since it strictly applied the matching strategy suggested in SCCT. Unlike some other cases in which mixed or mis-matching strategies were used, this example illustrates the issue perfectly.
In 2017, one of the most shocking scandals in the history of Hollywood was revealed by an article in New York Times, listing the detailed sexual harassments against Harvey Weinstein throughout the past decades. This not only marked the beginning of a new stage of the #MeToo movement, which is a viral movement on social media against sexual
harassment and assault all around the globe, but also lifted the curtains of several other scandals in Hollywood, among which the most influential one involves the Oscar, Golden Globe, and Tony-winning actor Kevin Spacey (Eguchi, 2018). At that time, Kevin Spacey had been the lead actor and executive producer of the hit web television show House of Cards for five seasons since 2013—produced as the first original show from the streaming service giant Netflix—and the sixth season was in production. House of Cards has been a great success, receiving extremely positive reviews from both the critic and audiences. With several Emmy and Golden Globe trophies in its pocket throughout these years, it made history as the first online-only original television show to top the nominations in major awards in 2013 (Holpuch, 2013).
Crisis timeline can be seen later in Table 1 (BBC News, 2019).
Throughout this process, the decision made by Netflix received extremely polarised reactions from the public, and initiated wild discussion on social media. Some people thought that Netflix did the right thing and the issue should be dealt with seriously, while other people, especially fans of the show, believed that Netflix should not have fired Kevin Spacey.
2017 advance towards him when he was 14 back in 1986.
Spacey quickly responded by issuing a statement on his Twitter account, saying that he had no memory of the incident and he was sorry if the unpleasant encounter did happen.
Netflix followed up by saying it was troubled by the accusation.
October 31, 2017
2 more accusations against Spacey emerged.
Netflix reacted by announcing its suspension of the sixth season of House of Cards production, and that it needed time to review the current
situation.
November 1 to 3, 2017
5 other male victims came forward, making similar accusations.
November 3, 2017
Netflix announced that it would no longer be in association with Spacey.
November to December, 2017
More than 10 other new accusations were made.
A lot of them were filed by minors at the time when the harassment took place.
Police were involved to investigate, both in Britain and US.
December 4, 2017
Netflix confirmed that the sixth season of House of Cards would resume production in 2018 without Kevin Spacey.
Instead, the show will be focusing on the character of Robin Wright, who is the lead female in the series.
By the end of January, 2018
More than 30 individuals had made accusations against Spacey, ranging from sexual harassment to attempted rape.
For them, Spacey’s character and his critically-acclaimed acting are too important for House of Cards, and those allegations were not proved yet, lacking hard evidence.
In the end, the sixth season of House of Cards aired on November 2, 2018 on Netflix without the character of Kevin Spacey, whereas the truth behind the accusations and the scandal still remains fairly unclear. Clearly, Netflix used the strategy of rebuild to cope with this preventable crisis, distancing itself from Kevin Spacey, who was, under the circumstance, its employee, to generate new reputational assets and move on. This response is also in align with SCCT (Coombs, 2007).
Sample
Respondents were drawn through a combination of sampling methods of non-probability convenience and snowball sampling. Firstly, non-non-probability convenience sampling was employed, as the researcher had no budget for the study, and reaching out to people from the researcher’s network saved both time and money (Tashakkori & Teddlie, 2003). Participants were reached through various social media platforms, including Facebook, WhatsApp, Weibo (Chinese Twitter), and WeChat (Chinese WhatsApp). On one hand, the researcher sent out the invitation and asked them to participate in the study one by one through these channels. On the other hand, the researcher also posted the survey in different group chats.
In addition, because the researcher herself has access to some people identifying as House of Cards fans, on the foundation of convenience sampling, non-probability snowball sampling was also used. For instance, after sending out the request to a fan in the researcher’s network through social media, the researcher kindly asked that person to pass on the link of the study to more fans in his/her network or community. Since the level of fan identification is an essential variable for this study, the access to the fan community is extremely crucial. Through snowball sampling, this study reached more potential respondents who were likely
to rate higher on the fan identification level, compared to most of other participants approached through convenience sampling. What’s more, fan communities are relatively difficult to reach, as members of the group cannot be easily detected or identified, and they tend to gather together based on their shared interest (Baltar & Brunet, 2012). Thus, snowball sampling helped the researcher to get more responses through insider knowledge, without being intrusive (Baltar & Brunet, 2012).
In total, 235 completed questionnaires were received (N = 235). Participants came from diverse cultural backgrounds, covering more than 30 nationalities. 29.8% of the respondents identified themselves as male (N = 70), and 67.2% as female (N = 158). The average age of the group was 35 (M = 34.7, SD = 13.35). 63% of them knew Netflix as a company (N = 148), and 34.5% of the participants had watched House of Cards before (N = 81).
Control Variable
The control variable for this study is the perceived reputation of Netflix, consisting of both cognitive reputation, which is knowledge-based, and affective reputation, which is emotion-based (Schwaiger, 2004; Walsh & Beatty, 2007). Studies have indicated that one’s perception of a company’s reputation is likely to influence his/her standpoint during
corporate crises (Dean, 2004; Kiambi & Shafer, 2016; Zheng, Liu, & Davison, 2017). Hence, this study has this variable measured and then controlled.
Questionnaire Design
This study has applied the research method of survey, with the help of a cross-sectional self-administered online questionnaire, powered by Qualtrics. The method of a quantitative survey study can reach a relatively large sample, while at the same time, stay relatively low-cost (Creswell, 2009). As this study is interested in self-reports from the respondents regarding their opinions at a given time, there is no need for different waves or
follow-up questionnaires. Furthermore, online survey is capable of gathering large quantities of data within a relatively short amount of time, and respondents are allowed to complete the study wherever, whenever, and however they want, as long as they have access to a digital device and the Internet (Sue & Ritter, 2007). This also makes it much more convenient for respondents recruited through snowball sampling to pass on the invitation, without the restriction of geographic distances (Sue & Ritter, 2007).
It takes around five minutes in total to complete the survey, and the questionnaire is available in three languages: English, simplified Chinese, and traditional Chinese. The majority of the respondents are expected to understand English well enough for this study, but as the researcher is Chinese and has a large number of connections with Chinese people from different educational backgrounds, having the Chinese version can effectively avoid the situation where language barriers get in the way. All the items were translated from English to Chinese by the researcher herself.
The questionnaire consists of six sections. The survey begins with a short introduction of the study, including an informed consent statement that the participants need to actively click on “I agree” before proceeding. The second part of the questionnaire asks the
participants to rate on their perceived reputation of Netflix, and if they have never heard of the company, the items will not be shown. Then, the survey wants to know about the respondents’ self-identified fan level of House of Cards, and if they have not watched the show before, this part will be skipped. After that, a short description of the Kevin Spacey scandal and how Netflix handled it is presented, on the same page following the introduction of the illustration is the part measuring the account acceptance. The fifth section asks about participants’ social media use, secondary crisis communication, and social media opinion climate. Lastly, the survey comes to an end with some demographic questions and a thank-you note.
Measurements and Operationalisation
There are five main variables in the research questions and hypotheses. The
independent variable is the level of fan identification, and the dependent variable is account acceptance, with three mediator variables being social media use, secondary crisis
communication, and social media opinion climate. Except for general social media use, the answers for each item are provided on the basis of a 7-point Likert scale, ranging from 1 = strongly disagree to 7 = strongly agree.
For fan identification, there has been several existing scales developed by different scholars (Obst, Zinkiewicz, & Smith, 2002; Reysen & Branscombe, 2010; Rudski, Segal, & Kallen, 2009). However, the scale created by Groene and Hettinger (2016) has organically combined the aforementioned scales and been proved valid. Thus, their scale has been eventually applied for this study and slightly adapted to fit into the theme of House of Cards. It is comprised of 15 statements, and typical ones include “I am emotionally connected to House of Cards” and “I devote a lot of energy to House of Cards”. When it comes to account acceptance, following the precedents of Coombs and Holladay (2008) as well as Van Der Meer and Verhoeven (2014), the 6-item scale developed and validated by Blumstein and his colleagues has been used (1974), covering statements such as “Netflix’s response is
acceptable” and “Netflix’s response is sincere”.
General social media use has been evaluated by the question “on average, how much time do you spend on each social media platform on a daily basis?”. The list of social media includes Facebook, Twitter, Instagram, WhatsApp, YouTube, Tumblr, WeChat, and Weibo. They are either the most used ones worldwide, especially for fandom communities, or the most popular ones in the Greater China area, where a large proportion of the respondents come from (Bennet, 2014; Ghaznavi & Taylor, 2015; Coelho, Oliveira, & Almeida, 2016; Zhou, Zhang, Liu, Zhang, Bai, & Zhu, 2017; Liu, 2015; Gan & Wang, 2015). The response
options are 1 = none, 2 = less than one hour, 3 = about one hour, 4 = about two hours, 5 = about three hours, and 6 = four or more hours. This is a widely-used scale format in the field, since it is capable of not only showing the total amount of use, but also the differences
amongst diverse platforms (Ohannessian, 2009; Valenzuela, 2013; Whaite, Shensa, Sidani, Colditz, & Primack, 2018). As a result, although there are numerous other valid scales for social media use, given the wide variety of platforms involved in this study, this scale format is eventually chosen (Sigerson & Cheng, 2018).
A 3-item scale has been used to rate respondents’ secondary crisis communication involvement, based on the scale developed by Zheng, Liu and Davison in 2018, including items like “I have forwarded this crisis message on social media” and “I have posted negative comments about this incident on social media”. To investigate social media opinion climate, a 3-item scale has been implemented, which is adapted from Zheng, Liu and Davison’s original scale as well (2018). There are quite a number of existing scales measuring public opinion climate, but not many of them are able to test social media opinion climate in particular (Tsfati, 2003; Lindenmeier, Schleer, & Pricl, 2012). The one developed by Zheng and his colleagues has absorbed insights from different areas of research, and created a valid scale that is especially designed for the opinion climate on social media (2018). Examples such as “My friends shared the same opinion as mine on social media” and “My contacts I followed on social media shared the same opinion as mine” are included.
The control variable is measured by a 6-item scale created by Schwaiger (2004). Statements include “As far as I know, Netflix is recognised worldwide” and “I regard Netflix a likable company”. The complete questionnaire items are available in the Appendix.
Results Measurement Validation
As most scales are existing ones that have been proved to be valid and have been implemented in the Chinese language before (Ngai & Jin, 2016; Zheng, Liu, & Davison, 2018), factor analysis is not necessary for them. Only two factor analysis were conducted on the variables of fan identification and social media use, since they have not been validated in Chinese yet. The first principal axis factor analysis was performed on fan identification, and it shows that the 15 items form a unidimensional scale. Although there are two components with eigenvalue higher than 1, at 9.72 and 1.19 respectively, the latter one is only slightly above 1, and the scree plot also suggests that only one factor is extracted. This factor explains 64.8% of the variance. All the items correlate positively with that factor, with the highest factor loading at 0.92 (“being a fan of House of Cards defines me”).
The second principal axis factor analysis was carried out on social media use. The results, including the scree plot, demonstrate that there are two components with eigenvalue higher than 1 (3.25 and 1.54 respectively), explaining 59.92% of the variance together. Facebook, Twitter, Instagram, WhatsApp, YouTube, and Tumblr form one factor, while WeChat and Weibo form another. This can be read as the former factor being prevalent global social media platforms and the latter being dominant social media platforms from the Greater China region. Even though the statistic results imply that they should be treated as two factors, with theoretical backup, the study continues to regard them as measuring the overall social media use (Lien & Cao, 2014; Chan & Guo, 2013; Li & Chen, 2014). More about this will be touched upon in the Discussion section.
Reliability tests were then run to check all the scales. Overall, all but one scale have the value of Cronbach’s Alpha above 0.80. Especially for the independent variable fan identification and dependent variable account acceptance, their Cronbach’s Alpha are at 0.96 and 0.91 respectively. The control variable perceived reputation of Netflix is also reliable (α = .80), and the same goes for two of the mediation variables (secondary crisis communication,
α = .90; social media opinion climate, α = .88). Considering the fact that all the scales have been translated into a different language, the reliability level is exceptionally satisfactory.
However, the scale measuring social media use is not so neat when it comes to the reliability test, with a Cronbach’s Alpha value only at 0.47. Following the unsatisfactory factor analysis on this variable, this may be the result of the fact that the scale used in this study have covered eight social media platforms and combined items prevalent in both Western and Eastern cultures. More about this will be discussed in the Discussion section later on.
Hypotheses Testing
During the assumption checking for regression analysis, it was found that the data is not normally distributed enough. Histogram produced satisfactory result, whereas scatterplot showed that the data selection process could have been improved, and the Conclusion part will get back to this. Despite of this, a simple regression was first conducted to explore the main relation. For H1, which predicts that fan identification level is negatively associated with account acceptance under the circumstance, it can be concluded that it is not supported. The regression model is significant, explaining 10% of the variance when holding the control variable constant (R2 = .10, F(2, 232) = 12.87, p < .001). However, the correlation itself is extremely weak, and not statistically significant either (b = - .08, t = -1.34, p = .18, 95% CI [-0.20, 0.04]). In other words, the direct effect of this model is not significant, and H1 is rejected.
Then, three possible mediation relations were tested with the help of PROCESS. As far as H2a goes, it is interested in the possible mediation role social media use has on the relation between fan identification and account acceptance. The mediated relationship is not significant (b = .01, 95% CI [-0.00, 0.03]). The prediction of fan identification on social media use is not significant either (b = -.28, t = -1.13, p = .26, 95% CI [-0.76, 0.21]), and the
same goes for social media use on account acceptance (b = -.02, t = -1.42, p = .16, 95% CI [-0.05, 0.01]). Based on the above-mentioned aspects, H2a has to be rejected too.
Continuing on to H2b, which explores the possible mediation effect of secondary crisis communication on the main relation, a significant relationship can be found. Fan identification has an indirect effect on account acceptance through secondary crisis
communication (b = -.04, 95% CI [-0.10, -0.01]). Additionally, fan identification also has a weak to moderate effect on secondary crisis communication (b = .26, t = 3.63, p < .001, 95% CI [0.12, 0.40]), and secondary crisis communication in turn has a weak effect on account acceptance (b = -.17, t = -2.55, p < .05, 95% CI [-0.30, -0.04]). Thus, H2b is partially supported.
The last hypothesis H2c proposes that social media opinion climate also plays a mediating role between the main association, which is fan identification and account
acceptance. Results indicate that this hypothesis is partially supported, since the relationship exists as an indirect effect as well (b = .03, 95% CI [0.00, 0.07]). The effect of fan
identification on social media opinion climate is also statistically significant, even though the correlation itself is quite weak (b = .18, t = 2.16, p < .05, 95% CI [0.02, 0.35]), and so is the association between social media opinion climate and account acceptance (b = .16, t = 2.85, p < .05, 95% CI [0.05, 0.27]).
Control Variable
Both the regression analysis and the mediation tests through PROCESS have proved that the effect of the control variable, which is the perceived reputation of Netflix, is
statistically significant in the model (b = .16, t = 5.04, p < .001, 95% CI [0.10, 0.22]; b = .16, t = 4.39, p < .001, 95% CI [0.09, 0.24]). To put it in another way, it is correct to control this variable for this study, as it has a significant influence on the overall findings.
Figure 2. Conceptual framework with results.
(*p <.05, **p <.001)
Discussion
In the current era, where only one accusation made on the Internet can cause
disruptive damage for large corporations, corporates are getting more and more on the edge when handling crises. Largely empowered by modern technology, media fans are taking the advantage of social media to be united, express their ideas, and make an impact when needed. Their power cannot be underestimated, especially for entertainment companies relying
greatly on fan base. Even by following certain guidelines developed by scholars, such as SCCT, the outcome can still be highly unpredictable.
This study has answered two research questions, derived from the problem,
concerning the relationship between fan identification, account acceptance, and social media. Revisiting the first question—“to what extent is the level of fan identification associated with the level of account acceptance of the rebuild strategies deployed by companies during
Fan
Identification Acceptance Account
Social Media Use Secondary Crisis Communication Social Media Opinion Climate -0.07 -0.28 0.26** 0.18* -0.02 -0.17* 0.16*
preventable corporate crises?”—it has been found that there is no direct association between these two variables, rejecting the first hypothesis. For the second question, “how is the association between fan identification and account acceptance mediated by social media use, secondary crisis communication, and social media opinion climate?”, predictive relationships can be identified via secondary crisis communication and social media opinion climate, but not through social media use. Hence, for the three hypotheses derived from this research question, one is rejected, while the other two are partially supported.
Although with theoretical support beforehand, as can be seen in the final results, the significant association between fan identification and account acceptance does not exist. Since the direct effect is non-existing, there is no mediation in the model at all. No matter how much of a strong presence there seems to be on social media discussion, the disparities between fans and the general public are not statistically significant in the population. Without the direct link between fan identification and account acceptance, it practically implicates that public relations professionals at entertainment companies do not necessarily need to pay a lot of extra attention to this particular group of stakeholders. When following specific guidelines or strategies to deal with corporate crisis, what is considered acceptable by the general public is always the most important thing.
Nonetheless, fan identification is found to have indirect influence on account acceptance through secondary crisis communication and social media opinion climate. The more one identifies as a fan, the more secondary crisis communication he/she engages in, which in turn, leads to less account acceptance. To a certain extent, this is still in align with what the researcher had in mind based on theoretical framework, as higher level of fan identification does lead to lower level of account acceptance when facing the rebuild strategy applied by the company during preventable crises, only taking a detour. The more one
and share relevant information on social media (Lukach, 2012; Ho & McLeod, 2008). During this process, he/she is also more likely to further dislike the decision made by the corporate, because participating in secondary crisis communication can strengthen the already strong attitudes (Zheng, Liu, & Davison, 2018).
Nevertheless, the other potential mediator social media opinion climate generates far more intriguing insights, as the result goes the other way around compared to what the researcher thought based on literature review. The more one identifies as a fan, the more homogenous his/her social media opinion climate is, as the researcher predicted. Yet, the more homogenous social media opinion climate leads to higher level of account acceptance. In other words, fan identification indirectly but positively predicts account acceptance through social media opinion climate, which goes completely against the previous finding and what the researcher proposed.
Having said all that, it seems that, judging from the effect size, the influence of secondary crisis communication is slightly stronger compared to the one through social media opinion climate (b = -.04 compared to b = .03). Furthermore, as both indirect models are extremely weak, looking at four separate predictive routes in the model—independent variable to mediators, and then respective mediators to dependent variable—the one through secondary crisis communication is stronger as well (b = .26, b = -.17 compared to b = .18, b = .16). To a certain extent, this takes another step forward by proving that fan identification does, from a more comprehensive perspective, slightly but negatively predict account acceptance under the circumstance.
This has implicated that the public opinion in the social media environment is of vital importance, especially nowadays. Secondary crisis communication and social media opinion climate have direct influence on account acceptance, and companies should definitely keep an eye out for relevant activities on social media. On the base of regular monitoring on
keywords mentioning, it would also be a good idea to keep track of the opinion climate and the traces of secondary crisis communication in certain communities—for example, fan communities and followers of influencers or online opinion leaders, particularly during crisis time, to be better prepared for what is yet to come and have an overview of the reception of what has been done. What’s more, it is also recommended by this study that SCCT should be applied with careful thinking and flexible adaptation. The SCCT model has oversimplified certain matters, and when referring to that in practice, more aspects should be taken into consideration other than merely crises type.
Besides those aforementioned practical implications, this research has contributed to theoretical advancement as well. To begin with, the study has integrated the concepts of fan identification, account acceptance, as well as social media, showed that they are tightly connected to each other, and paved the way for upcoming research to further validate those connections. Secondly, to a certain degree, the study has challenged SCCT by providing evidence that the so-called matching strategies to combat corporate crises are not always so fruitful. It concerns much more than simply the crisis type, but also the nature of the
audiences together with communication channels.
Another interesting finding this study has picked up, which is also valuable for future research, is that there seems to be significant differences between social media platforms from Eastern and Western cultures. As the scale of calculating the time spent on respective platforms and then adding them up is widely-used in the field when measuring social media use (Ohannessian, 2009; Valenzuela, 2013; Whaite, Shensa, Sidani, Colditz, & Primack, 2018), following this commonly-applied practice, this study has treated “social media use” as one coherent variable. Nevertheless, factor analysis has suggested that there are two
components within that variable, one being the domestic platforms used in the Greater China area, and the other being those international.
A lot of studies have been done on the possible cultural differences between social media use (Pentina, Zhang, & Basmanova, 2013; Cho & Park, 2013). However, they all tend to focus on the differences within one platform across cultures, for instance, the behavioural differences between people from America and Europe when using Facebook or Instagram (Vasalou, Joinson, & Courvoisier, 2010; Sheldon, Rauschnabel, Antony, & Car, 2017), or the differences between networking sites serving similar function—such as Facebook and Renren (Qiu, Lin, & Leung, 2013; Li & Chen, 2014). Scholars have acknowledged that people from different cultures use different social media outlets, which makes perfect sense, but they have failed to pinpoint its possible consequences for academic research measuring the
comprehensive social media use (Fietkiewicz, Lins, & Budree, 2018; Choi, Kim, Sung, & Sohn, 2011; Li & Chen, 2014). The threat of mixing those less international ones in the measurement scale is yet to be identified, even in the studies focusing on the scale validation and development for social media use (Rosen, Whaling, Carrier, Cheever, & Rokkum, 2013; Olufadi, 2016; Sigerson & Cheng, 2018). Thus, the study has another scientific implication by having helped the development of social media and culture studies, as it showcases strong evidence that the culture in which the platform originates has significant impact on the possible academic outcome. The urge and necessity of further investigating the reasons behind these differences have been clearly identified, which has not been done before by the academia.
Conclusion
This quantitative study is dedicated to answering two leading research questions, namely “to what extent is the level of fan identification associated with the level of account acceptance of the rebuild strategies deployed by companies during preventable corporate crises?” and “how is the association between fan identification and account acceptance
mediated by social media use, secondary crisis communication, and social media opinion climate?”, through the method of a survey. No direct association between fan identification and account acceptance can be established, while secondary crisis communication and social media opinion climate do have significant relevance standing between fan identification and account acceptance. In the end, it is safe to say that fan identification indirectly and
negatively predicts account acceptance of the rebuild strategies deployed by companies during preventable corporate crises, through secondary crisis communication.
As any studies, this one is not without limitations, especially in terms of methodology. Firstly, during the data collection stage, not enough House of Cards fans were reached.
Ideally, the number of people who had watched the show should be more or less the same as the number of people who had not, providing a more balanced sample for this study. This may also have an effect on the fact that the sample is not normally distributed enough. Notwithstanding, taking on the perspective of the percentage of fans, in the end this sample still represents the population relatively well, since people who have watched the show is indeed the minority when put in the bigger picture.
Secondly, the use of a single illustration has certain drawbacks. This real-life crisis example of Netflix and Kevin Spacey’s scandal has its own characteristics, and it makes the results hard to be generalised. Furthermore, mainly due to the lack of theoretical development and the limitations of the scale as touched upon in the previous section, the Cronbach’s Alpha value of the mediation variable social media use is extremely low for a measurement scale used in a study. This poses as a questionable disadvantage as well, as it may have affected the final results regarding social media use.
This research has also pointed out some new directions for future studies. For instance, other variables influencing account acceptance can be looked into. Since fan identification does not have a direct and significant effect on it, what else may be of great importance then?
Or, asking the question from an inductive perspective, why is the discussion around Netflix and Kevin Spacey so wild on social media, yet the results do not seem to reflect the thoughts expressed online? One of the reasons might be that the self-reporting nature of survey studies has certain shortcomings and people may conceal their real opinion, and if that is the case, it is possible to investigate this problem further through content analysis and/or interviews.
Moreover, other types of crisis communication strategy used in SCCT can be
explored. For example, another illustration of this genre of crisis in the entertainment industry recently is the one involving the scandal of director James Gunn and Disney. Yet, Disney took on a completely different approach to handle the situation. Interesting results can be expected when comparing different strategies used under similar circumstances and their levels of account acceptance, to further validate or challenge SCCT.
Lastly, compensating the victims is one of the most typical conduct the company employs during preventable corporate crises, at least according to SCCT (Coombs, 2007). However, in the illustration used in the study, Netflix did nothing with regard to the victims caught up in Kevin Spacey’s scandal, even with allegations made by other Netflix employees who were on the set of House of Cards with the actor then (Henderson, 2017). This can be of interest for future studies as well, as one can argue that celebrity scandals are not so directly related to the corporate compared to other crises, such as product-harm, and the corporate does not need to take up the responsibility of reimbursing. Therefore, the comparisons between celebrity scandals and other types of preventable crises can be further investigated. The same goes for the possible logic behind the differences between social media platforms originated in different cultures. More importantly, with the process of globalisation and the fast development of social media platforms, when measuring social media use on a global scale, a more appropriate and comprehensive scale should be developed.
Although the sixth, which is also the last, season of House of Cards experienced steep plunge in terms of audience reception without Kevin Spacey and his character, it is, still, absolutely the right sacrifice to make for Netflix and its future business development. However, whether other companies share this vision or whether die-hard fans can overcome their grudge is a different question. Let’s leave it to time, shall we?
References
Avery, E. J., Lariscy, R. W., Kim, S., & Hocke, T. (2010). A quantitative review of crisis communication research in public relations from 1991 to 2009. Public Relations Review, 36(2), 190–192. doi:10.1016/j.pubrev.2010.01.001
Baltar, F., & Brunet, I. (2012). Social research 2.0: Virtual snowball sampling method using Facebook. Internet Research, 22(1), 57-74. doi:10.1108/10662241211199960
BBC News. (2019, January 7). Kevin Spacey timeline: How the story unfolded. Retrieved from https://www.bbc.com/news/entertainment-arts-41884878
Bennett, L. (2013). Researching online fandom. Cinema Journal, 52(4), 129-134. Retrieved from http://www.utexas.edu/utpress/journals/jcj.html
Bennet, L. (2014). Fan/Celebrity interactions and social media: Connectivity and
engagement in Lady Gaga fandom. In L. Duits, K. Zwaan, & S. Reijnders (Eds.), The Ashgate Research Companion to fan cultures (pp. 109-134). Farnham, UK: Ashgate Publishing.
Billings, A., Qiao, F., Conlin, L., & Nie, T. (2017). Permanently desiring the temporary? Snapchat, social media, and the shifting motivations of sports fans. Communication & Sport, 5(1), 10-26. doi:10.1177/2167479515588760
Bird, C. M. (2002). Phenomenological realities or “Quinntown,” life in a cyber community. Journal of American & Comparative Cultures, 25(1/2), 32-37. doi:10.1111/1542- 734X.00006
Blumstein, P. W., Carssow, K. G., Hall, J., Hawkins, B., Hoffman, R., Ishem, E., . . .
Zimmerman, D. L. (1974). The honoring of account. American Sociological Review, 39(4), 551–566. doi:10.2307/2094421
Brown, N. A. (2014). The convergence of situational crisis communication theory and social media: Empirically testing the effectiveness of sports fan-enacted crisis
communication (Doctoral dissertation). Retrieved from
https://ir.ua.edu/bitstream/handle/123456789/1955/file_1.pdf?sequence=1 Brown, N. A, Brown, K. A, & Billings, A. C. (2015). “May no act of ours bring
shame”: Fan-enacted crisis communication surrounding the Penn State sex abuse scandal. Communication & Sport, 3(3), 288-311. doi:10.1177/2167479513514387 Cha, H., Suh, J. W., & Kim, J. R. (2015). The effect of issue obtrusiveness, issue congruence,
and response strategies on the acceptance of crisis communication messages. Asian Journal of Communication, 25(3), 307-326. doi:10.1080/01292986.2014.978878 Chan, M., & Guo, J. (2013). The role of political efficacy on the relationship between
Facebook use and participatory behaviors: A comparative study of young American and Chinese adults. Cyberpsychology, Behavior, and Social Networking, 16(6), 460- 463. doi:10.1089/cyber.2012.0468
Cho, S. E., & Park, H. W. (2013). A qualitative analysis of cross-cultural new media research: SNS use in Asia and the West. Quality & Quantity, 47(4), 2319-2330. doi:10.1007/s11135-011-9658-z
Choi, S. M., Kim, Y., Sung, Y., & Sohn, D. (2011). Bridging or bonding? A cross-cultural study of social relationships in social networking sites. Information, Communication & Society, 14(1), 107-129. doi:10.1080/13691181003792624
Choi, Y., & Lin, Y. H. (2009). Consumer response to crisis: Exploring the concept of involvement in Mattel product recalls. Public Relations Review, 35(1), 18–22. doi:10.1016/j.pubrev.2008.09.009
Chung, S., & Lee, S. (2016). Crisis communication strategy on social media and the public’s cognitive and affective responses: A case of Foster Farms salmonella outbreak. Communication Research Reports, 33(4), 341-348.
Claeys, A., Cauberghe, V., & Vyncke, P. (2010). Restoring reputations in times of crisis: An experimental study of the situational crisis communication theory
and the moderating effects of locus of control. Public Relations Review, 36, 256–262. doi:10.1016/j.pubrev.2010.05.004
Claeys, A., & Cauberghe V. (2014). What makes crisis response strategies work? The impact of crisis involvement and message framing. Journal of Business Research, 67(2), 182-189. doi:10.1016/j.jbusres.2012.10.005
Click, M. A., Lee, H., & Holladay, H. W. (2013). Making monsters: Lady Gaga, fan identification, and social media. Popular Music and Society, 36(3), 360-379. doi:10.1080/03007766.2013.798546
Coelho, R. L. F., Oliveira, D. S. D., & Almeida, M. I. S. D. (2016). Does social media matter for post typology? Impact of post content on Facebook and Instagram metrics. Online Information Review, 40(4), 458-471. doi:10.1108/OIR-06-2015-0176
Consalvo, M. (2003). Zelda 64 and video game fans: A walkthrough of games, intertextuality, and narrative. Television New Media, 4(3), 321-334. doi:10.1177/1527476403253993
Coombs, W. T. (1995). Choosing the right words: The development of guidelines for the selection of the “appropriate” crisis-response strategies. Management
communication quarterly, 8(4), 447-476. doi:10.1177/0893318995008004003 Coombs, W. T. (2006). The protective powers of crisis response strategies: Managing
reputational assets during a crisis. Journal of Promotion Management, 12(3-4), 241- 260. doi:10.1300/J057v12n03_13
Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of Situational Crisis Communication Theory. Corporate Reputation Review, 10(3), 163-176. doi:10.1057/palgrave.crr.1550049
Coombs, W. T. (2011). Ongoing crisis communication: Planning, managing, and responding. Thousand Oaks, CA: Sage Publications.
Coombs, W. T., & Holladay, S. J. (2008). Comparing apology to equivalent crisis response strategies: Clarifying apology's role and value in crisis communication. Public Relations Review, 34(3), 252-257. doi:10.1016/j.pubrev.2008.04.001
Coombs, W. T., & Holladay, S. J. (2014). How publics react to crisis communication efforts: Comparing crisis response reactions across sub-arenas. Journal of Communication Management, 18(1), 40–57. doi:10.1108/JCOM-03-2013-0015
Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods approaches. Thousand Oaks, CA: Sage Publications.
Dean, D. H. (2004). Consumer reaction to negative publicity: Effects of corporate reputation, response, and responsibility for a crisis event. The Journal of Business
Communication (1973), 41(2), 192-211. doi:10.1177/0021943603261748
Deluca, K. (2018). Shared passions, shared compositions: Online fandom communities and affinity groups as sites for public writing pedagogy. Computers and Composition, 47, 75-92. doi:10.1016/j.compcom.2017.12.003
Du Plessis, C. (2018). Social media crisis communication: Enhancing a discourse of renewal through dialogic content. Public Relations Review, 44(5), 829-838.
doi:10.1016/j.pubrev.2018.10.003
Eguchi, S. (2018). Layers of homonormativity in Kevin Spacey's coming-out scandal. QED: A Journal in GLBTQ Worldmaking, 5(2), 98-104. doi:10.14321/qed.5.2.0098
Fietkiewicz, K. J., Lins, E., & Budree, A. (2018, July). Investigating the generation-and gender-dependent differences in social media use: A cross-cultural study in
Germany, Poland and South Africa. In International Conference on Social Computing and Social Media (pp. 183-200). Springer, Cham. doi:10.1007/978-3-319-91485-5_14
Fillis, I., & Mackay, C. (2014). Moving beyond fan typologies: The impact of social integration on team loyalty in football. Journal of Marketing Management, 30(3-4), 334-363. doi:10.1080/0267257X.2013.813575
Fong, C. P., & Wyer, R. S. (2012). Consumers’ reactions to a celebrity endorser scandal. Psychology & Marketing, 29(11), 885-896. doi:10.1002/mar.20571 Gan, C., & Wang, W. (2015). Uses and gratifications of social media: A comparison of
microblog and WeChat. Journal of Systems and Information Technology, 17(4), 351- 363. doi:10.1108/JSIT-06-2015-0052
Ghaznavi, J., & Taylor, L. D. (2015). Bones, body parts, and sex appeal: An analysis of #thinspiration images on popular social media. Body image, 14, 54-61.
doi:10.1016/j.bodyim.2015.03.006
Grady, M. (2014). Fandom at the crossroads: Celebration, shame and fan/producer
relationships/fan culture: Theory/practice. The Journal of American Culture, 37(1), 84. doi:10.1111/jacc.12118
Groene, S. L., & Hettinger, V. E. (2016). Are you “fan” enough? The role of identity in media fandoms. Psychology of Popular Media Culture, 5(4), 324-339.
doi:10.1037/ppm0000080
Gwinner, K., & Swanson, S. R. (2003). A model of fan identification: Antecedents and sponsorship outcomes. Journal of services marketing, 17(3), 275-294.
doi:10.1108/08876040310474828
Harriss, C. (2017). The producer as fan: Forensic fandom and The Good Wife. Journal of Communication Inquiry, 41(4), 368-381. doi:10.1177/0196859917712233
employee and made atmosphere on set toxic’. Retrieved from
https://www.telegraph.co.uk/films/2017/11/03/kevin-spacey-sexually-assaulted-house-cards-employee-made-atmosphere/
Ho, S. S., & McLeod, D. M. (2008). Social-psychological influences on opinion expression in face-to-face and computer-mediated communication. Communication Research, 35(2), 190–207. doi:10.1177/0093650207313159
Holpuch, A. (2013, July 18). Netflix makes history as House of Cards takes top Emmy nominations. Retrieved from
https://www.theguardian.com/tv-and-radio/2013/jul/18/emmys-nominations-netflix-house-of-cards
Hong, S., & Kim, B. (2018). Exploring social media use in university crisis communication: An experiment to measure impact on perceived crisis severity and attitudes of key publics. Journal of Contingencies and Crisis Management, 27(1), 61-71.
doi:10.1111/1468-5973.12242
Jahng, M., & Hong, S. (2017). How should you tweet?: The effect of crisis response voices, strategy, and prior brand attitude in social media crisis communication. Corporate Reputation Review, 20(2), 147-157. doi:10.1057/s41299-017-0022-7
Johnson, A. R. (2005). When a celebrity is tied to immoral behavior: Consumer reactions to Michael Jackson and Kobe Bryant. In G. Menon & A. R. Rao (Eds.), Advances in consumer research (pp. 100-101). Duluth, MN: Association for Consumer Research. Kellner, D., & Deraa, H. (2008). Fandom: Identities and communities in a mediated world.
Contemporary Sociology: A Journal of Reviews, 37(6), 559-560. doi:10.1177/009430610803700621
Kiambi, D. M., & Shafer, A. (2016). Corporate crisis communication: Examining the interplay of reputation and crisis response strategies. Mass Communication and Society, 19(2), 127-148. doi:10.1080/15205436.2015.1066013
Kim, S., Zhang, X. A., & Zhang, B. W. (2016). Self-mocking crisis strategy on social media: Focusing on Alibaba chairman Jack Ma in China. Public Relations Review, 42(5), 903-912. doi:10.1016/j.pubrev.2016.10.004
Kozman, C. (2013). The Tiger Woods scandal in the media: Measuring attribute effects on the public. International Journal of Sport Communication, 6(2), 214-233.
doi:10.1123/ijsc.6.2.214
Laufer, D., & Coombs, W. T. (2006). How should a company respond to a product harm crisis? The role of corporate reputation and consumer-based cues. Business Horizons, 49(5), 379–385. doi:10.1016/j.bushor.2006.01.002
Laufer, D., Gillespie, K., McBride, B., & Gonzalez, S. (2005). The role of severity in consumer attributions of blame: Defensive attributions in product-harm crises in Mexico. Journal of International Consumer Marketing, 17(2–3), 33–50.
doi:10.1300/J046v17n02_03
Li, X., & Chen, W. (2014). Facebook or Renren? A comparative study of social networking site use and social capital among Chinese international students in the United States. Computers in Human Behavior, 35, 116-123. doi:10.1016/j.chb.2014.02.012
Liang, Y., & Shen, W. (2016). Fan economy in the Chinese media and entertainment industry: How feedback from super fans can propel creative industries’ revenue. Global Media and China, 1(4), 331-349. doi:10.1177/2059436417695279
Lien, C. H., & Cao, Y. (2014). Examining WeChat users’ motivations, trust, attitudes, and positive word-of-mouth: Evidence from China. Computers in Human Behavior, 41, 104-111. doi:10.1016/j.chb.2014.08.013
Lindenmeier, J., Schleer, C., & Pricl, D. (2012). Consumer outrage: Emotional reactions to unethical corporate behavior. Journal of Business Research, 65(9), 1364-1373. doi:10.1016/j.jbusres.2011.09.022
