The emotional crisis communication of Germanwings and Malaysia Airlines
Name: Valerie Kirsch Student ID: 10841792
Master Thesis
Graduate School of Communication Corporate Communication Supervisor: Suzanne de Bakker Date of completion: January 26, 2016
Abstract
The master thesis deals with the emotional crisis communication of Germanwings and Malaysia Airlines. Both companies had to face a plane crash and thereby made use of
different kinds of emotions to communicate with their public. I studied their communication on Facebook and analysed posts and comments.
I assumed that the more emotions a company shows, the less negative the
communication of the public would be. The results confirm this assumption, meaning that more emotions shown by a company resulted in less negative communication by the public towards the company. This indicates that using emotions during a crisis generally helps a company to show a more human side and therefore protect their reputation.
Moreover, results showed that most used emotions were sadness and sympathy, which is connected to the crisis types, namely plane crashes that led to human victims. Therefore, people were likely to express their sadness about the incident and their sympathy towards the families of the victims.
The overall conclusion of the thesis is that it depends on the crisis type to state which emotions should best be used by a company. That means, there is no general suggestion of what emotions shown by a company experiencing a crisis lead to a better communication of the public but one always has to analyse the type of the crisis.
1 Organizational crises are a threat to its reputation and the organization will therefore have an interest in applying a strategy to repair their damaged image (Benoit, 1997). At the same time, crises also include other facets, such as different kinds of emotions. Those will be on the side of the organization as well as on the side of the public (Van der Meer &
Verhoeven, 2014). The organization facing a crisis will have to choose whether to show or suppress such emotions. Since corporations are institutions and therefore non-human, they are usually run on a rational level. Hence, it might seem odd to express emotions that refer to a human side. Moreover, some people might consider the expression of emotions, such as shame and regret as weak and therefore inappropriate, which is why a company could choose to not do so publicly. However, van der Meer and Verhoeven (2014) argue that showing emotions, such as shame and regret during a crisis might have a positive impact on the organization’s reputation.
Until now, there is some research on how organizations should behave during a crisis (Benoit, 1997; Claeys, Cauberghe, & Vyncke, 2010; Coombs, 1998, 2006, 2007). However, not much is known about the effects that emotions shown from the organizational side will have on the public’s side and the public’s perception of the organization’s reputation. Most of the conducted studies within that field, such as that from Van der Meer and Verhoeven (2014) are experiments. They analysed the behaviour of people in reaction to communicated
emotions by companies during a crisis. However, they only focused on the emotions of shame and regret and left other types of emotions out. Therefore, it is relevant to analyse a broader range of different emotions to gain a better understanding of the connection between emotions and crisis communication.
In addition, Kim and Cameron (2011) found evidence that people react more positive towards a corporation during a crisis when the company focuses on the well-being of crisis victims rather than on law or punishment. Both studies within the field of emotional crisis
2 communication used experiments and therefore focused on the behaviour of people on a rather offline and private level. It is left open to analyse how people behave online regarding emotional crisis communication. One could argue that social media users feel more
anonymous online and therefore have more courage to express inappropriate behaviour. However, I argue that a crisis, such as a plane crash, causes such strong emotions and sympathy that people will express the same feelings online as they do offline. People might express their feelings more specifically online, due to the safe feeling of anonymity, but I do not see a reason why they would totally change their expression of emotions. The aim of this study is therefore to find out to what extend online emotional crisis communication will influence the publics’ communication.
Because the web becomes more interactive regarding a rather two-way communication with one’s stakeholders instead of a former one-way communication (Constantinides & Fountain, 2008), it is relevant for organizations to understand what effects their emotions might have on their customers. Therefore, my research questions are:
RQ1: To what extend do emotions, communicated on social media platforms during a crisis, have influence on the social media communication of the public towards the
organization?
RQ2: To what extend does emotional crisis communication lead to a less negative communication by the public instead of a more positive communication?
3 Theoretical background
Organizational image and reputation
Corporate image is the current vision that the audience has of an organization and it can change quickly due to recent events. In other words, the image is what a person thinks immediately when for instance seeing a corporate logo. In contrast to that, corporate reputation is the steady judgement the audience has of an organization due to consistent behaviour in the past. It does not change as quickly as a corporate image but it develops over time. Both image and reputation are important because stakeholders will prefer a company with a positive image and reputation (Gray & Balmer, 1998).
Reputation develops through different kinds of news a person receives about the organization. Those can be via “interactions with an organization” (Coombs, 2007, p. 164), meaning that one’s reputation develops through two-way communication. That can be through social media, where one can interact with an organization. Moreover, reputation forms through “mediated reports about an organization” (Coombs, 2007, p. 164), meaning what one reads or sees in the news about an organization or what the organization
communicates through advertisement. Reputation also forms through “second-hand
information from other people” (Coombs, 2007, p. 164). Those can be what others tell us or what one reads online on weblogs. In connection with my research question, one can therefore argue that corporate communication serves as a factor when the public develops its reputation about the organization. Therefore, the frames, which the organization uses during a crisis are an important factor in reputation formation.
In addition, Gensler, Völckner, Liu-Thompkins and Wiertz (2013, p. 253) state that a corporate brand is not exclusively defined by the company itself anymore but that there has been a shift towards “consumers as pivotal authors of brand stories in the branding process.” That means that organizations should see the consumer’s perception of the image and
4 reputation of the company as important. During a crisis, the audience might receive negative information about a company and therefore communicate negatively about it. Hence, it is important for an organization to repair its image during a crisis because otherwise it might damage the reputation (Benoit, 1997).
Benoit (1997) states that an organization’s image is threatened when it is believed to be responsible for an act and when the act itself is considered heinous or offensive. In other words, an organization’s image is threatened when a bad incident happened and the
organization is considered to be responsible for it. Moreover, Benoit (1997, p. 178) points out that “perceptions are more important than reality.” This indicates that it is less relevant if an organization is actually responsible for the crisis than if the audience perceives it to be since image and reputation are rather subjective concepts. Coombs (1998, p. 186) found evidence that “image damage increases with perceptions of organizational responsibility for the crisis.” With regard to this, the perception that a person has depends on the framing of the situation (Entman, 1993). Taking both aspects together, one can assume that a person’s perception regarding the image, reputation and responsibility of a company during a crisis depends on how the situation is framed. Therefore, a company should be interested in framing a crisis in a way that it will lead to the protection of its reputation. I argue, that people will experience different emotions during a bad crisis, such as a plane crash. Therefore, using emotions as frames by the organization during a crisis can serve as a reputational protection because it makes the organization seem more human.
Organizational crisis types
To understand the type of an organizational crisis better, Coombs (2006) identifies three crisis clusters. Those are the victim cluster, the accidental cluster and the preventable cluster. The victim cluster indicates that an organization is the victim of a crisis, e.g. due to a natural disaster, false rumors, workplace violence in the form of an employee attacking the
5 organization or and external person threatening the image of the organization (Coombs,
2006). When a crisis is ascribed to the accidental cluster, it means that the actions of an organization unintentionally led to the crisis. Characterizations of that can be a technical breakdown error or a technical accident (Coombs, 2006). Within the preventable cluster, an organization took a risk consciously and therefore acted inappropriately. Characterizations are human breakdown accidents and recalls, organizational misdeed with or without injuries or violated laws (Coombs, 2006).
There are three factors that build the reputational threat: crisis responsibility, crisis history and relationship history. In general, how greater the organization’s responsibility of the crisis is, the greater the reputational threat is. This is because people experience anger and less sympathy towards the organization when they identify an organization as responsible. That means that more organizational responsibility will create more negative feelings towards the organization and therefore the reputation is threatened. That means that the crises within preventable cluster are the biggest threat to the reputation, crises within the accidental cluster are a medium threat and crises within the victim cluster are a minimum threat (Coombs, 2007). However, there are two crisis intensifiers that can increase the reputational threat. The first is crisis history, meaning that the reputational threat is higher when the organization has experienced similar crises in the past. The second is relationship history, regarding how well the organization treated their stakeholders in the past. The reputational threat thereby is higher when the organization treated their stakeholders badly in the past (Coombs, 2006, 2007). Understanding the type of crisis is important with regard to crisis communication, since different crisis types might result in different communication of the public. That means, the more responsibility a company has for the crisis, the worse the public’s communication might be.
6 Theory of framing
Entman (1993, p. 52) defines framing as selecting “some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described.” In other words, everyone can have a different perception of reality and therefore, the same event can be framed differently by different people. As Entman (1993) lists in his definition, frames define problems, diagnose causes, make moral judgements and suggest remedies.
Van Gorp (2007) also defines a frame used in the news media as an invitation to the reader to understand the text in the specific way the author intended to. Therefore, he states that framing starts with a journalist who shapes the news in a particular way and then the sense of that frame is possibly adopted by the readers or rather the audience. Moreover, he points out that frames are connected to the culture that they are used in. Van der Meer, Verhoeven, Beentjes and Vliegenthart (2014) point out that frames help to make sense of a situation and therefore, especially help during crisis situations to understand what is
happening.
In connection to crisis communication, one can say that using a strategy means to frame the crisis in a specific way. With regard to this, Schultz, Kleinnijenhuis, Oegema, Utz and Van Atteveldt (2012) analysed the case of the BP oil spill crisis in 2010 and showed that the framing of the media and the PR differed in the beginning. The PR frame focused on the problem of the oil spill and on solutions and apparently tried to avoid taking responsibility.
The media’s frame was more complex in the beginning and focused on the company’s responsibility, consequences of the oil spill and different actors that were involved during the
crisis. However, after a while the media partially adopted the frame of the PR, meaning that
7 authors assume that the media partially took over the company’s framing (Schultz et al., 2012).
With regard to frames alignment, Van der Meer et al. (2014) studied the dynamics of framing in crises situations and found three periods of frame alignment. The first period indicates that PR, the media and the public will have different frames at the beginning of a crisis. After a while, those frames will align, which is the second period. However, after the crisis-frame alignment, the PR, media, and the public will use different frames again (Van der Meer et al., 2014). The authors argue that frame alignment during a crisis is a temporary status because the three domains of PR, the media and the public need to “reduce ambiguity and uncertainty and to provide coherence to the issue by aligning in their frame
communication” (Van der Meer et al., 2014). In other words, the public has the aim to make sense of the situation that might seem chaotic in the beginning of the crisis, when not all information is provided and a general overview is not given yet. Therefore, there will be a natural response of adopting the frame that the organization offers.
Moreover, the organization has an interest to show their own view of the crisis so that the public will not adopt a more negative frame of a third party that might lead to a damaged reputation. During a crisis, it might be different for the public to understand what is happening and therefore they need a way to make sense of the situation. In connection with emotions, I assume that the public will show the same or similar kinds of emotions as the spokesperson of the organization does, since that is a way to make sense of the crisis and therefore to
understand what is happening. Hence, my first hypothesis is:
H1: The public will adopt the emotions framed by an organization and therefore express the same kinds of emotions online.
8 Organizational crisis communication
To work against a threat of reputation, an organization will make use of a crisis response strategy. Coombs (2006) introduced three options that an organization has. Those are the deny response option, the diminish response option and the deal response option.
The deny response option includes attacking the accuser, who claims that there is an issue. Moreover, it includes denying that there is a crisis at all or finding a scapegoat for the crisis and therefore blaming someone else (Coombs, 2006). An example would be that when there is a negative report in the press about a company, in which the organization is blamed for an issue. The organization would use the deny response option when blaming the journalists and possibly state that they were informed badly or did not research properly.
The diminish response option includes finding an excuse and thereby trying to
minimize the organization’s responsibility for the crisis. Moreover, it includes justification of the situation, meaning that the organization tries to minimize the crisis damage by stating that it is not as bad (Coombs, 2006).
The dealing response option implies ingratiation, meaning that the organization tries to remind stakeholders of past positive works. It also includes showing concern for the victims, offering compassion for the victims, in the form of money or gifts, showing regret by saying that the organization feels bad about the situation and apologizing for the incident and thereby taking full responsibility (Coombs, 2006). Coombs (2007) also calls the dealing option a rebuild strategy and states that this is the strategy that addresses victims best. In addition, Claeys, Cauberghe and Vyncke (2010) and Van der Meer and Verhoeven (2014) found evidence that the rebuild strategy generally leads to a more positive reputation than the diminish strategy.
However, not only the strategy of how to deal with the crisis is important, but also the medium that is used for communication. Utz, Schultz and Glocka (2013, p. 40) conducted an
9 experiment where they found out that “crisis communication via social media resulted in a higher reputation and less secondary crisis reactions such as boycotting the company than crisis communication in the newspaper.” The authors explain this finding with the statement that the use of social media by an organization signals willingness to inform quickly and to start a dialogue with its stakeholders (Utz et al., 2013).
In addition, Gensler et al. (2013) state that there has been a shift towards interactivity due to the rise of social media, which will lead to branding processes and therefore the process of building a reputation. I therefore conclude that companies should put emphasis on the use of social media while applying a crisis communication strategy. People valuing a company’s willingness for a dialogue on social media might affect their perception of image and reputation of the company in a positive way. Therefore, I analysed crisis communication on social media to conduct my research.
Emotions
In terms of my research questions, as I already pointed out in the introduction, a company can choose to express emotions or not. Showing emotions would form an emotional appeal that aims to touch the consumers’ feelings. The opposite would be a rational appeal, meaning that the organization would try to persuade the consumers through rational reasons. Researchers have argued that the effectiveness of the appeal depends on the match with the product type. That means, in advertisement an emotional appeal is best used if the product suits to appeal to the consumer’s emotions and a rational appeal is best used if the product is rather utilitarian (Albers-Miller & Stafford, 1999). In connection with the study on hand, the crisis itself can be regarded as the product since it is what the message is about. A crisis, such as a plane crash, is generally emotional for everyone and therefore the literature suggests that an emotional appeal suits better than a rational appeal.
10 Moreover, Albers-Miller and Stafford (1999) summarize previous results that state that emotional appeals generally lead to more positive reactions within the customers. Also, emotional appeals seek to make the customer feel good about the product. Taking both aspects together, people might prefer emotional appeals during crisis communication because it makes them feel better about the tragic incident. In addition, literature suggests that
emotional appeals will lead to more positive responses towards the organization. In this study, I aim to find out if choosing the emotional appeal has advantages regarding the protection of the organizational reputation during a crisis. However, Van der Meer and Verhoeven (2014) state that expressing emotions during a crisis will rather reduce negative opinions on the public’s side than increasing positive feelings towards the company. Hence, my second hypothesis is:
H2: Showing emotions during a crisis by the organization will lead to less negative communication of the public towards the organization.
Emotional crisis communication
Emotions are part of everyday life and usually “occur in response to an event, usually a social event, real, remembered, anticipated or imagined” (Ekman, 1993, p. 385). The list of different emotions is long and Shaver, Schwartz, Kirson and O’Connor (1987) state that there is no basic definition of an emotion. However, people are able to categorize emotions and assign different kinds of feelings to one bigger feeling or emotion. Therefore, the authors conducted an experiment, where they let 112 participants categorize 135 emotion words into six basic sets. Their results show that six basic concepts are revealed, which are love, joy, anger, sadness, fear and surprise, even though surprise is a rather small category since it only includes three emotion words. Each concept consists of several more differentiated emotions, e.g. among others, sadness contains shame and regret (Shaver et al., 1987).
11 Moreover, the authors divided the emotion words into positive and negative emotions. The outcome shows that all emotion words included in love, joy and surprise are positive. The emotion words that are part of anger and fear are all negative. Of the 37 emotion words that belong to sadness, only the emotion sympathy is categorized as positive and melancholy is categorized as neutral, whereas the other 35 words are negative (Shaver et al., 1987).
Emotions can be expressed verbally, through the use of language and/or non-verbally, through the use of facial expressions, postures and/or gestures (Ekman, 1993). In other words, a person can simply say how he or she is feeling and thereby express an emotion verbally. Emotions can also be shown non-verbally. This can either happen spontaneously, e.g. when someone makes a joke and we laugh or smile, or with intention, e.g. when we have an audience. In the latter case, emotions can thereby be expressed in an exaggerated manner or hold back because we do not want the others to know about our feelings (Parkinson, 2005). In terms of my study, an organization would express emotions intentionally as they know about their audience. Because I will focus on social media communication, an organization will be likely to express emotions in written words, by saying how they feel. Another possibility would be to post a video, where a spokesperson can also express emotions non-verbally.
Taking emotions and crisis communication together, expressing emotions during a crisis can be considered as a communication strategy. As explained earlier, the rebuild
strategy regarding crisis communication includes showing concern for the victims, expressing regret and apologizing for situation (Coombs, 2006). Coombs (2006) does not explicitly mention emotions. However, since regret is one of the emotions that Van der Meer and Verhoeven (2014) analysed in their study and regret is also mentioned with regard to the rebuild strategy, I argue that showing emotions is generally part of this strategy. That means,
12 when a company shows emotions, such as expressing regret, shame and sorrow for what happened they make use of the rebuild strategy to protect their reputation.
Kim and Cameron (2011) found general evidence that emotions framed by an organization will influence the public’s response. They used the emotions of sadness and anger and studied relationships with the emotions a person had due to the news framing of the organization. Their results indicate that during a corporate crisis that involves human victims, the public rather wants to know about the victims and their possible reliefs than about law, justice or punishment regarding the company’s wrongdoing (Kim & Cameron, 2011). Therefore, people will be receptive for emotions regarding the tragedy and its victims. Kim and Cameron (2011, p. 845) suggest that “when a corporate crisis involves human victims, the responsible company might need to develop its response messages by maximizing the human interest frames and emphasizing how the victims’ situations have been relieved.” Due to their statement, I assume that the public will rather comment with concern about the victims than about the role the organization played or plays within the crisis. Therefore, my third
hypothesis is:
H3: During a crisis, the public will express more emotions online regarding the crisis victims than the organization itself.
In line with this finding, Van der Meer and Verhoeven (2014) conducted an experiment where they analysed the relationship between communicated emotions by an organization during a crisis and the public response. They also found evidence that the reputation of an organization during a crisis is better after communicating emotions than no emotions. That means, “the display of these emotions by spokespersons affects the way the public makes sense of the role of the organization in the crisis” (Van der Meer & Verhoeven, 2014, p. 533). In contrast to Kim and Cameron (2011), Van der Meer and Verhoeven (2014) used the emotions of shame and regret. They did not find a different public reaction between
13 the emotions of shame and regret and left it open to further research to analyse other kind of emotions. Therefore, according to my second research question, I aim to find out what kinds of emotions will lead to the protection of the reputation.
Moreover, Van der Meer and Verhoeven (2014) point out that using a strategy that involves expressing emotions during a crisis will protect the company’s image in terms of “reducing the public’s negative affects process (i.e., anger), rather than eliciting feelings of sympathy” (Van der Meer & Verhoeven, 2014, p. 533). This indicates that especially negative emotions play an important role during crisis communication since people will not be likely to have positive feelings due to the crisis. This also indicates that the positive outcomes of an effective crisis communication are not necessarily positive emotions because the crisis itself might be so terrible that people are not able to have positive emotions. However, the positive outcomes of a good crisis communication will therefore be that there are less negative emotions regarding the company (Van der Meer & Verhoeven, 2014). Moreover, the authors argue that emotional communication by the organization has a cognitive effect on the public because they observe the emotions that the organization shows and thereby judge the
organization and make sense of the situation. This part, regarding sensemaking and framing, is also described earlier.
Previous literature has only studied few kinds of emotions, for instance shame and regret but never took all possible different kinds of emotions into account. Therefore, I cannot base a hypothesis regarding the most used negative emotions on that. However, literature suggests that people will rather be concerned about the victims of the crisis, than about the organization. Assuming that a happened crisis was a one-time incident, there is no reason for the public to feel fear, since they are not put in a dangerous situation because the dangerous situation already has happened the moment the public responses to it. Therefore, I assume that the public will rather show feelings of anger and sadness. Anger will rather be about the fact
14 that the incident happened at all and could not have been prevented. Sadness will rather be about the fact that people suffered or even died. The only positive feelings the public could express in connection to a crisis could be sympathy towards the victims. However, this is one emotion of plenty and the most logically expressed emotions, such as anger and sadness, are negative. Therefore, my fourth hypothesis is:
H4: The emotions used during a crisis by the organization as well as by the public will mostly be negative.
15 Methods
In order to answer my research questions and to test my hypotheses, I conducted a quantitative content analysis. Thereby, I used the two cases of Germanwings and Malaysia Airlines that both had to face a plane crash. Both incidents led to an organizational crisis since the companies had to communicate to the public what happened and why it happened. Both cases are comparable because both crashes were not caused due to a technical error but due to human acts.
In the following, I will describe the cases, the research units, the variables and the research design I made use of.
Case of Germanwings
On Tuesday, 24th March 2015, a plane of the airline Germanwings crashed in the French Alps. It was travelling from Barcelona to Düsseldorf and had 144 passengers and 6 crewmembers on board, who all died. Most of the victims were German or Spanish (Spiegel Online, 2015a).
On Friday, 27th March 2015, it came out that the reason for the crash was the co-pilot of the airplane, who made the plane crash on purpose. Several documents were found in his apartment that revealed that his doctor had declared him not to be able to work due to his depressions (Spiegel Online, 2015b). Germanwings declared in a press release that they did not receive the sick certificate from the co-pilot (Germanwings, 2015).
The plane crashed because of the illness of the co-pilot, which can be categorized as an employee attacking the organization. According to Coombs (2006) this is part of the victim cluster, meaning that the responsibility of Germanwings is rather low.
Case of Malaysia Airlines
On Thursday, 17th July 2014, an airplane of Malaysia Airlines crashed in the Ukraine
16 place. There were 298 people on board of the plane, who all died. The plane was travelling from Amsterdam to Kuala Lumpur and most of the passengers were Dutch (Spiegel Online, 2014).
Even though experts assumed early that the plane was fired off from the ground, evidence for that was found more than one year later, in October 2015. Experts made an assumption of the region where the rocket was fired off from, which was at that time in hands of pro-Russian separatists (Spiegel Online, 2015c). In addition, it was announced that the product of the rocket was Russian (Schmid, 2015). However, until now there is no final evidence regarding the responsible party of the rocket.
Since it was not the fault of the company, this crisis is also part of the victim cluster. Therefore, the company has little responsibility for what happened.
Materials
The analysis consist of two actor parts. One are the messages of the companies and the other one are the messages of the public in response to those of the comapnies. The two organizations are Germanwings and Malaysia Airlines. The research material regarding the two companies are their Facebook posts. The starting date is that of each plane crash and ended with the last Facebook post that related to the plane crash. Regarding Germanwings, 29 Facebook posts were used and regarding Malaysia Airlines 25 Facebook posts were used. Each Facebook post was coded in terms of the relevant subjects, which are described afterwards.
The public is another research unit. I analysed the responses they gave in connection to the Facebook posts of the companies, which were in the form of comments. Therefore, each comment was coded in terms of the relevant variables. Germanwings received 17,792 comments in total and Malaysia Airlines received 996 comments in total. However, there was a high amount of comments that did not include any content towards the subject, e.g. when
17 someone linked another person to the post or when the comment included only a few letters or random signs that did not make sense. Moreover, there was a high amount of comments where people discussed the incidents without including emotions. In addition to that, there was a high amount of comments that went off topic, e.g. when some commenters started to discuss religion. Therefore, all of those comments that did not include emotions by the sender or that were not connected to the incident were removed from the sample. That left 6,194 comments of Germanwings and 310 of Malaysia Airlines.
Variables
The variables that are relevant for the analysis are “emotion cluster”, “emotion set”, “tone” and “addressor”. After the description in the following, there is a figure that presents the relationship between the variables. A more detailed description is included in the
codebook that is attached in the appendix.
Emotion set. Each emotion cluster is part of one of the six emotion sets. By coding each emotion cluster, I will generate a variable that shows the overall emotion set. The sets are joy, love, fear, anger, sadness and surprise and each emotion cluster belongs to one of those concepts. That means, if a sender expresses e.g. shame in a message, the emotion cluster will be shame and the emotion set will be sadness, since shame is part of the overall set of sadness. The emotion set variable is nominal.
Tone. Besides coding the expressed emotions, I will also code the tone of each message, which can be positive, negative or neutral. Shaver et al. (1987) divided the emotion sets into positive and negative. Love, joy and surprise are generally positive emotions and fear, anger and sadness are generally negative, with the exception of sympathy, which is part of sadness but still positive and melancholy, which is also part of sadness but neutral. If a sender
18 sadness and the tone negative. An example of the emotion sadness and its clusters and tones is shown beneath.
However, it is possible that a message contains irony or sarcasm and therefore requires the interpretation of the coder. This variable regarding the tone is ordinal.
Emotion cluster. The basic variables of this research are the emotions used by the
organizations and the public. Therefore, I coded what kinds of emotions were expressed in the Facebook posts and comments. To do so, I made use of the emotion sets and emotion clusters, described by Shaver, Schwartz, Kirson and O’Connor (1987). The list contains 25 emotion clusters that all belong to one of the six basic emotion sets. The clusters are affection, lust, longing, cheerfulness, zest, contentment, pride, optimism, enthrallment, relief, surprise, irritation, exasperation, rage, disgust, envy, torment, suffering, sadness, disappointment, shame, neglect, sympathy, horror and nervousness. The full list, including all emotion words, clusters, sets and the tone are shown in the appendix.
Each cluster includes several emotion words that help to assign different expressions to one of the cluster. That means each post will be coded in terms of the emotions expressed by the sender. Therefore, each emotion cluster or a conjugation of the word will be coded. If there are emotions expressed that cannot clearly be assigned to one of the emotion words, they will be coded as “other”. If a sender expresses e.g. shame in a message, the category “shame” will be coded as emotion cluster. This variable is nominal.
19 Figure 1: Overview of emotion set sadness
Addressor. With regard to the social media messages by the public, I also coded whom their comment was addressed to. These were either the victims, the company or other, e.g when someone expressed that they were generally shocked by the incident. Thereby, I was able to differentiate if a specific emotion was expressed towards the victims (e.g. sadness about the tragic incident) or towards the company (e.g. anger about their behaviour). This variable is nominal.
Results
The emotions used by Germanwings were sadness and fear. 90% of all used emotions were sadness and 30% of the used emotions were fear. It is important to add that fear always occurred in combination with sadness but sadness was also expressed alone. Moreover, the emotions used within the sadness set were mostly sympathy and sadness. The following table shows the used emotions of Germawings.
Emotion set: Sadness
Tone: Negative Emotion cluster: Suffering Negative Sadness Negative Disappointment Negative Shame Negative Neglect Positive Sympathy
20 Used emotions Germanwings N Percentage Love 0 0% Anger 0 0% Sadness 27 75% Fear 9 25% Joy 0 0% Surprise 0 0% Total 36 100%
Table 1: Used emotions by Germanwings
88.3% of all expressed emotions by the public are sadness. Second biggest emotion set is fear (6.6%), followed by love (4.5%) and anger (1.5%). Joy and surprise were never expressed.
Public’s emotions to Germanwings N Percentage Love 278 4.4% Anger 89 1.4% Sadness 5,395 86.1% Fear 506 8.1% Joy 0 0 Surprise 0 0 Total 6.268 100%
Table 2: Used emotions by Germanwings' public
Because the emotion set of sadnes is very big, it is also mentionable to have a look at the emotion clusters used within that set. There, we can see that 82.2% of all emotions within the sadness set were sympathy. Only 6.8% of the emotions within the sadness set were actual sadness and a very few comments expressed shame (0.1%).
Public’s sadness cluster N Percentage Sympathy 5,022 92.3% Sadness 413 7.6% Shame 5 0.1% Total 5.440 100%
Table 3: Emotion set sadness of Germanwings' public
Malaysia Airlines expressed the emotion sets sadness (25%), joy (4.2%) and surprise (4.2%). It is worth mentionable that the sadness set only included the emotion cluster of sympathy. That means, the sadness cluster was not expressed, only the sympathy cluster, which is part of the overall sadness set.
21 Used emotions
Malaysia Airlines N Percentage
Love 0 0% Anger 0 0% Sadness 6 75% Fear 0 0% Joy 1 12.5% Surprise 1 12.5% Total 8 100%
Table 4: Used emotions by Malaysia Airlines
The public expressed 71.6% sadness towards the case of Malaysia Airlines, followed by love (19%), anger (7.1%) and fear (2.6%).
Public’s emotions N Percentage
Love 59 19.5% Anger 22 7.3% Sadness 222 73.3% Fear 8 2.6% Joy 0 0% Surprise 0 0% Total 303 100%
Table 5: Used emotions by Malaysia Airlines' public
When looking at the sadness set expressed by the public, we see that 65.5% of the expressed emotions within that set were sympathy, 7.4% sadness and 0.3% shame.
Public’s sadness cluster N Percentage Sympathy 203 89.4% Sadness 23 10.1% Shame 1 0.5% Total 227 100%
Table 6: Emotion set sadness of Malaysia Airlines' public
To test my first hypothesis, I ran a correlation that included the emotion sets used by the companies and the public. The output for Germanwings is shown beneath. We see that the correlation between the sadness sets is significant (p < .0001) but the correlation is very weak and negative (r = -.049). That means the more sadness the company expressed, the less sadness the public expressed.
22 The correlation between the sets of fear is significant (p < .0001) but weak and
positive (r = .066). That means the more fear the company expressed, the more fear the public expressed.
The only emotion sets that do not correlate are the company’s set of fear and the public’s set of anger (p = .609).
Germanwings
Sadness Fear
Public Love Sig. .00 .00
Pearson Correlation .06 .07 Anger Sig. .00 .61 Pearson Correlation -.05 -.01 Sadness Sig. .00 .00 Pearson Correlation -.05 -.09 Fear Sig. .00 .00 Pearson Correlation .04 .07
Table 7: Output correlation emotion sets Germanwings and public
Because the sadness sets have a negative correlation, H1 is declined for the case of Germanwings.
For the case of Malaysia Airlines, we see that the sets of sadness correlate and have a weak, positive effect (p < .0001; r = .20). That means the more sadness the company
expressed, the more sadness the public expressed.
We also see that the public’s set of anger and fear do both not correlate with the company’s sets of joy and surprise.
Even though the sets of sadness correlate and have a positive effect, the sets of joy and surprise were not adopted by the public. Therefore, H1 is declined for the case of Malaysia Airlines.
23 Malaysia Airlines
Sadness Joy Surprise
Public Love Sig. .03 .01 .00
Pearson Correlation -.12 .16 .24 Anger Sig. .10 .27 .72 Pearson Correlation -.09 -.06 .02 Sadness Sig. .00 .11 .00 Pearson Correlation .20 -.09 -.18 Fear Sig. .04 .52 .31 Pearson Correlation -.12 -.04 -.06
Table 8: Output correlation emotion sets Malaysia Airlines and public
To test my second hypothesis, I ran a correlation between all emotions of the company and the negative and positive tone of the public. In the case of Germanwings, the negative tone of the public correlates with whether or not Germanwings used emotions and the effect is negative (p = .004; r = -.15). The positive tone of the public dose not correlate with whether the company used emotions or not (p = .098). Therefore, H2 is accepted for the case of Germanwings. All Emotions Germanwings Public Negative tone Sig. .00 Pearson Correlation -.15 Positive tone Sig. .10 Pearson Correlation .08 Table 9: Output correlations all emotions by Germanwings and tone by public
In the case of Malaysia Airlines, the negative tone of the public correlates with
whether or not the company used emotions and the effect is negative (p = .006; r = -.266). The positive tone of the public does not correlate with the emotions of the company (p = .011). Therefore, H2 is accepted for the case of Malaysia Airlines.
All Emotions Malaysia Airlines Public Negative tone Sig. .01 Pearson Correlation -.27 Positive tone Sig. .01 Pearson Correlation .25
24 In order to test my third hypothesis, I first ran a frequency table that shows that the public showed emotions towards Germanwings 365 (6.7%) times and 5,057 (93.3%) times towards the victims. To test if the difference is significant, I ran a one-sample t-test. The mean regarding the case of Germanwings of the public’s addressor (M = .93, SD = .26) was higher than the test value of 0.5. The mean difference of 0.43 is significant (p < .0001). Therefore, H3 is accepted for the case of Germanwings.
Towards Malaysia Airlines, the public showed emotions 104 times (37.7%) and towards the victims of Malaysia Airlines 172 times (62.3%). The one-sample t-test revealed that the mean of the public’s addressor (M = .62, SD = .49) was higher than the test value of 0.5. The mean difference of 0.12 is significant (p < .0001). Therefore, H3 is accepted for the case of Malaysia Airlines.
The following figure shows a graphic of the distribution of emotions towards the addressors.
Figure 2: Emotions used by the public towards addressor
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Germanwings Malaysia Airlines
Emotions used by the public
25 To test my fourth hypothesis, I first ran a frequency table. With regard of the tone of all Germanwings posts and comments, 5.9 % was negative, 55.4% were both negative and positive and 38.7% were positive. To test which tone was used most, I conducted a one-sample t-test. The mean of the tone (M = 1.33, SD = 0.58) was higher than the test value of 1 and therefore, the mean was more directed to positive emotions. The mean difference of 0.33 was significant (p < 0.0001). Therefore, H4 is declined for the case of Germanwings.
With regard of the tone of al Malaysia Airlines posts and comments 11.9% were negative, 5.5% were both negative and positive and 82.6% were positive. To test which tone was used most, I conducted a one-sample t-test. The mean of the tone (M = 1.71, SD = 0.67) was higher than the test value of 1 and therefore, the mean was directed more toward positive emotions. The mean difference of 0.71 was significant (p < 0.0001). Therefore, H4 is declined for the case of Malaysia Airlines.
In the following, two figures show graphic of the used tones and the used emotion sets.
Figure 3: Tone of all used emotions
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Germanwings Malaysia Airlines
Tone of emotions
26 Figure 4: Occuring emotion sets
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Germanwings Malaysia Airlines
Ocurring emotion sets
27 Discussion and conclusion
The most important conclusion of this study is that emotions expressed by the
organization during a crisis correlate with less negative communication of the public towards the organization. Moreover, the results showed that showing more emotions on the side of the organization does not correlate with a more positive communication by the public. This finding is completely in line with the statement made by Van der Meer and Verhoeven (2014) who point out that using emotions will protect the company’s image in terms of “reducing the public’s negative affects process (i.e., anger), rather than eliciting feelings of sympathy.”
With regard to the type of crisis, a plane crash, one can say that people generally face negative emotions about the fact. That means, a good communication strategy implemented by the company rather helps to diminish negative feelings towards the company instead of increasing positive ones, such as making the public happy, since happiness would be something that people cannot face regarding a plane crash.
When looking at the results of the two cases analysed in the study, one sees that the effect of less negative communication is higher with regard to the case of Malaysia Airlines. This can be explained due to the difference in crisis of the two cases. Since the plane by Malaysia Airlines was shot down by a third party, Malaysia Airlines is victim of the crisis and therefore does not face responsibility for the plane crash. Therefore, the reputational threat is not very big and people do not necessarily feel negative about the company itself. The plane of Germanwings crashed because of the illness of the co-pilot, who made the plane crash on purpose. This can be categorized as an employee attacking the organization (Coombs, 2006) and is therefore also part of the victim cluster. However, even though both cases are part of the victim cluster, Malaysia Airlines faced a cause from the outside of the company, whereas Germanwings faced a depressed co-pilot who intentionally killed all passengers. One can argue that Germanwings had some sort of responsibility for that, since they should care about the well-being of their employees. As Benoit (1997) states, everyone has a different
28 perception of the reality and therefore people can perceive the responsibility of a company differently. In general, the public rather showed sympathy towards Malaysia Airlines and it seems that the public sees them as less responsible for their crisis than Germanwings.
To conclude, even though the effect of less negative communication towards the company is higher for Malaysia Airlines, this does not necessarily mean that Germanwings implemented a worse communication than Malaysia Airlines. The reason for that is rather the difference in crisis type and responsibility. Moreover, one could argue that since
Germanwings had a slightly higher responsibility than Malaysia Airlines, they applied a good communication strategy since the public’s communication was still less negative when the company made use of emotions.
The difference between those cases leads to the general conclusion of this study. Effective emotional crisis communication depends on the type of crisis and the specific characterizations of the crisis. That means, it is not possible to define a general rule about which emotions have to be used when. Every crisis differs and therefore, the responsibility of the company differs. This is shown in the examples of Germanwings and Malaysia Airlines. Future research should consider this connection between crisis type and emotional crisis communication further.
Kim and Cameron (2011) state that when a crisis involves human victims the public will mostly be interested in their situation. In line with this statement, the findings of my study showed that in both cases, the public showed more emotions towards the victims than the company itself. Therefore, I conclude that when a crisis is connected to human victims, the well-being or rather the situation of those victims and their families is more important to the public than the role the organization plays or played.
Comparing the two cases, one sees that the public showed more emotions towards Malaysia Airlines than towards Germanwings, which is also an indicator for the difference in
29 crisis. That means, the public had sympathy towards Malaysia Airlines as a company since they did not find them responsible for the crisis and rather felt sorry for them. Moreover, comparing the results of both cases, one could conclude that the lower the company’s responsibility of the crisis, the higher the communication towards the company. However, since both companies are part of the victim cluster, and therefore both have a general low responsibility, this idea would have to be investigated further in future research.
Another aim of this study was to find out the different kinds of emotions and the tones that were generally used during a crisis. Surprisingly, for both cases the overall tone of used emotions was positive instead of negative, as expected. The reason for that is that one of the most used emotions clusters was sympathy, which is part of the overall sadness set, but is positive. Since both crisis faced several human victims, it is logical that the people felt
sympathy for their families. That result indicates that the emotions differ in terms of the crisis types. In connection to plane crashes that include human victims, people will feel sorry and show sympathy and therefore express a high amount of positive feelings.
Another surprising outcome of the study was that the public did not always adopt the feelings framed by the organizations. With regard to the case of Germanwings, the correlation between the sadness sets was even negative, meaning the more sadness the company
expressed, the less sadness the public expressed. However, interesting for the case of
Germanwings is that the sadness set of the company and the anger set of the public correlated negatively. That means, the more sadness the company expressed, the less anger the public showed. This generally shows that using emotions as a communication strategy leads to less negative feelings on the public’s side, which indicates that this strategy helps to protect the organizational reputation.
The only set that was used on both sides of Malaysia Airlines and the public was the sadness set, which correlated positively. However, the company also used joy and surprise as
30 emotions which were not mirrored on the public’s side. Interesting about that is that joy was used in connection to hope regarding the future. Since the public did not express the same hope, it is questionable whether showing that emotion is good or not. However, since the public still showed less negative communication towards the company, apparently it did not harm the company to have made use of joy.
The study on hand adds to the existing literature about crisis communication. Until now, all conducted studies within that field were experiments that analysed the behaviour of people on a rather private level. The study on hand analysed how the public reacts online towards a crisis. Moreover, several emotions were analysed. Before, only shame and regret had been studied in an experiment and sadness and anger had been analysed in another experiment. The study on hand reveals that in addition to the previous studied emotions also fear and love are expressed during a crisis. Especially love is an important emotion, in order to analyse whether the public shows affection towards the company or not. Fear is an emotion that is especially occurring in connection to the studied crises regarding plane crashes, since people expressed nervousness about flying, terrorism and murderer.
Moreover, by analysing both studies of Germanwings and Malaysia Airlines, the study on hand reveals that it depends on the crisis types to decide which emotions are used best. Both companies used different kinds of emotions but still received less negative
communication towards their companies.
The study on hand also has some limitations. First, only Facebook communication was analysed. Other studies could include more online communication, such as Twitter. Since a Tweet has a limited amount of signs, whereas a Facebook comment can be unlimitedly long, it is possible that there is a difference in communication between those mediums.
Moreover, the amount of Facebook comments of the two cases vary. Germanwings received a few thousand comments, whereas Malaysia Airlines only received a few hundreds.
31 One possible explanation for that is that Germanwings is more popular on Facebook.
However, it could also be that a reason for less comments is the degree of responsibility. Maybe people feel the urge to comment publicly when they judge the company to have a higher responsibility. However, that is a factor to be analysed by future research.
In addition, I coded the emotion clusters but then analysed the overall emotion sets. Since there was a difference in the clusters of sadness and sympathy, it would have been interesting to rather analyse the emotion clusters instead of the sets. Thereby, one would receive a more detailed information about the used emotions.
References
Albers-Miller, N. D., & Stafford, M. R. (1999). An international analysis of emotional and rational appeals in services vs goods advertising. Journal of Consumer Marketing, 16(1), 42–57. http://doi.org/10.1108/07363769910250769
Benoit, W. L. (1997). Image repair discourse and crisis communication. Public Relations
Review, 23(2), 177–186. http://doi.org/10.1016/S0363-8111(97)90023-0
Claeys, A.-S., 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(3), 256–262. http://doi.org/10.1016/j.pubrev.2010.05.004
Constantinides, E., & Fountain, S. J. (2008). Web 2.0: Conceptual foundations and marketing issues. Journal of Direct, Data and Digital Marketing Practice, 9(3), 231–244.
http://doi.org/10.1057/palgrave.dddmp.4350098
Coombs, W. T. (1998). An Analytical Framework of Crisis Situations: Better Responses From a Better Understanding of the Situation. Journal of Public Relations Research, 177–191. http://doi.org/10.1207/s1532754xjprr1003
Coombs, W. T. (2006). The Protective Powers of Crisis Response Strategies. Journal of
Promotion Management, 12(June 2015), 241–260. http://doi.org/10.1300/J057v12n03
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. http://doi.org/10.1057/palgrave.crr.1550049
Ekman, P. (1993). Facial Expression and Emotion. American Psychologist, 48, 376–379. Entman, R. M. (1993). entman-framing.pdf. Journal of Communication, 43(4), 51–58. Gensler, S., Völckner, F., Liu-Thompkins, Y., & Wiertz, C. (2013). Managing Brands in the
Social Media Environment. Journal of Interactive Marketing, 27(4), 242–256. http://doi.org/10.1016/j.intmar.2013.09.004
Germanwings. (2015). Copilot des Germanwings Flugs 4U 9525: Germanwings lag keine Krankmeldung vor. Retrieved November 17, 2015, from
https://www.germanwings.com/de/4u/unternehmen/presse/archiv/copilot-des-germanwings-flugs-4u-9525--germanwings-lag-keine-kra.html
Gray, E. R., & Balmer, J. M. T. (1998). Managing Corporate Image and Corporate Reputation. Long Range Planning, 31(5), 695–702. http://doi.org/10.1016/S0024-6301(98)00074-0
Kim, H. J., & Cameron, G. T. (2011). Emotions Matter in Crisis: The Role of Anger and Sadness in the Publics’ Response to Crisis News Framing and Corporate Crisis Response. Communication Research, 38(6), 826–855.
http://doi.org/10.1177/0093650210385813
Parkinson, B. (2005). Do facial movements express emotions or communicate motives?
Personality and Social Psychology Review : An Official Journal of the Society for Personality and Social Psychology, Inc, 9(4), 278–311.
http://doi.org/10.1207/s15327957pspr0904_1
Schmid, F. (2015). Abschuss von MH17: Unbequeme Fakten - nicht nur für Moskau. Retrieved November 17, 2015, from http://www.spiegel.de/panorama/abschuss-mh17-wer-die-verantwortung-zu-tragen-hat-a-1057623.html
Schultz, F., Kleinnijenhuis, J., Oegema, D., Utz, S., & van Atteveldt, W. (2012). Strategic framing in the BP crisis: A semantic network analysis of associative frames. Public
Relations Review, 38(1), 97–107. http://doi.org/10.1016/j.pubrev.2011.08.003
Shaver, P., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion knowledge: Further exploration of a prototype approach. Journal of Personality and Social Psychology,
52(6), 1061–86. http://doi.org/10.1037/0022-3514.52.6.1061
Spiegel Online. (2014). Möglicher Abschuss: Malaysische Passagiermaschine über Ukraine abgestürzt. Retrieved November 17, 2015, from
http://www.spiegel.de/panorama/malaysische-passagiermaschine-ueber-der-ukraine-abgestuerzt-a-981631.html
Spiegel Online. (2015a). Germanwings-Absturz in den Alpen: An der Felswand zerschellt. Retrieved November 17, 2015, from www.spiegel.de/panorama/germanwings-absturz-alle-bekannten-fakten-a-1025384.html
Spiegel Online. (2015b). Germanwings-Absturz: Co-Pilot war für Unglückstag krankgeschrieben. Retrieved November 17, 2015, from
http://www.spiegel.de/panorama/justiz/germanwings-absturz-co-pilot-war-fuer-unglueckstag-krankgeschrieben-a-1025956.html
Spiegel Online. (2015c). MH17-Abschlussbericht: Niederländische Ermittler bestätigen Abschuss durch Buk-Rakete. Retrieved November 17, 2015, from
http://www.spiegel.de/panorama/justiz/flug-mh17-wurde-abgeschossen-a-1057565.html Utz, S., Schultz, F., & Glocka, S. (2013). Crisis communication online: How medium, crisis
type and emotions affected public reactions in the Fukushima Daiichi nuclear disaster.
Public Relations Review, 39(1), 40–46. http://doi.org/10.1016/j.pubrev.2012.09.010
Van der Meer, T. G. L. A., & Verhoeven, J. W. M. (2014). Emotional crisis communication.
Public Relations Review, 40(3), 526–536. http://doi.org/10.1016/j.pubrev.2014.03.004
Van der Meer, T. G. L. A., Verhoeven, P., Beentjes, H., & Vliegenthart, R. (2014). When frames align: The interplay between PR, news media, and the public in times of crisis.
Public Relations Review, 40(5), 751–761. http://doi.org/10.1016/j.pubrev.2014.07.008
Journal of Communication, 57(1), 60–78.
Appendix 1: Tables and figures Figure 1: Overview of emotion set sadness
Figure 2: Emotions used by the public towards addressor Figure 3: Tone of all used emotions
Figure 4: Occuring emotion sets
Table 1: Used emotions by Germanwings
Table 2: Used emotions by Germanwings' public Table 3: Emotion set sadness of Germanwings' public Table 4: Used emotions by Malaysia Airlines
Table 5: Used emotions by Malaysia Airlines' public Table 6: Emotion set sadness of Malaysia Airlines' public
Table 7: Output correlation emotion sets Germanwings and public Table 8: Output correlation emotion sets Malaysia Airlines and public
Table 9: Output correlations all emotions by Germanwings and tone by public Table 10: Output correlations all emotions by Malaysia Airlines and tone by public
Appendix 2: Codebook Project description
The aim of this study is to find out to what extend online emotional crisis
communication will lead to the protection of an organization´s reputation. Therefore, the study consists of two parts, which are the analysis of two companies’ communication on the one hand and the analysis of the public’s communication on the other hand. Throughout this study their relationship and in addition, the effect of the companies on the public are meant to be analysed.
Two research questions will be answered by the following study. Those are:
RQ1: To what extend do emotions communicated on social media platforms during a crisis result in a less negative social media communication of the public towards the
organization?
RQ2: What emotions communicated by an organization during a crisis lead to a less negative social media communication of the public towards the organization?
Units of analysis
The units that will be analysed consist of two parts that represent the companies and the public. The first part are Facebook Posts of Germanwings and Malaysia Airlines that are in relation with the plane crashes. The other part will be the comments of the public on those Facebook Posts.
Units of context
The units of context of the companies will be the company name, and the used emotions. The units of context of the public will be the used emotions and the addressor of their message.
The emotions consist of an overall emotion set, the tone, and an emotion cluster that includes several emotion words that describe the cluster. The emotion word that is bold in the
text is the word that indicates the name of the cluster. That means, in the first emotion cluster the word “affection” is bold and therefore the whole cluster will be called affection cluster later on in the analysis. I will code the emotion cluster they find in the messages with help of the description and the emotion words.
Procedure
In order to answer the research questions and to test the hypotheses, Facebook posts will be coded. The coder will have a list of the posts of the companies or a list of comments from the public. For each unit, the companies or the public’s comments, there is one specific codebook that will be used. “Codebook 1 – Company part” is the codebook that includes the coding scheme for the Facebook posts and Tweets that were published by the companies with regard to the plane crashes. “Codebook 2 – Public part” is the codebook that includes the coding scheme for the comments that were published by the audience in answer to the previous post from the company.
Q3 of both codebooks deals with a detailed description of the emotion clusters. Since the emotion clusters from the companies’ and the public’s side were coded the same way, Q3 is summarized for the company and the public part. Therefore, the description of emotion clusters Q3 that is shown in the first part is also applicable for “Codebook 2 – Public part” and therefore contains information about companies and individuals.
Each post will be numbered and the comments regarding the same post will be labelled with the same number, so that the two parts from the company and the public can be connected afterwards.
Codebook 1 – Company part General information
Q1: Company name
(0) Germanwings, (1) Malaysia Airlines
Q2: Number of the Tweet/Post (in order to be able to connect the Tweet/Post with the public’s comment later on)
Q3: Emotions (applicable for company and public part) Variable
Description Coding
Emotion set
Tone Emotion cluster
Love Positive Adoration Affection Love Fondness Liking Attraction Caring Tenderness Compassion Sentimentality
Emotions that are part of affection are “warm, positive feelings directed to individuals.” It means to worship something or someone very much or to love and respect someone deeply. It includes to feel attached to someone, e.g. “the child won everyone’s heart.” Friendship and relationship contain affection that developed over time (The Free Dictionary, 2015a). It also means to prefer something, e.g. “He preferred whiskey over wine.” Affection includes to be concerned about the wellbeing of others and to be able to experience a deep awareness of the suffering of another accompanied by the wish to relieve it (The Free Dictionary, 2015b).
(0) No (1) Yes
Part of this cluster is also when a customer likes a company and expresses his or her positive view about it, e.g. “I like flying with Germanwings” or “I trust Germanwings.”
Love Positive Arousal Desire Lust Passion Infatuation
Emotions connected to lust are usually part of sexuality and can mean to be sexually aroused or to have the desire for another person (The Free Dictionary, 2015p). Infatuation rather connects to “a foolish and usually extravagant passion or love or admiration” and is often called puppy love. It is often temporary during
adolescence and a short-lived passion (The Free Dictionary, 2015l).
(0) No (1) Yes
Love Positive Longing Longing is a strong continuous desire or need and is often not fulfilled (The Free Dictionary, 2015o).
(0) No (1) Yes
Joy Positive Amusement Bliss Cheerfulness Gaiety Glee Jolliness Joviality Joy Delight Enjoyment Gladness Happiness Jubilation Elation Satisfaction Ecstasy Euphoria
Emotions that are part of the cheerfulness cluster are feelings of delight and to feel entertained. Because it is often mentioned in the context of entertainment having a feeling of pleasure or satisfaction is connected to it. It refers to being very happy and in a good mood and it can be associated with carefreeness. Feelings that are part of this cluster show signs of “extreme gratification aroused by something good or desired” and “a condition of supreme well-being and good spirits”(The Free Dictionary, 2015n).
(0) No (1) Yes
Joy Positive Enthusiasm Zeal
Zest Excitement Thrill Exhilaration
The emotions within the zest cluster are feeling a “vigorous and enthusiastic enjoyment” (The Free Dictionary, 2015ac).
Moreover, the emotions mean to have a strong positive feeling towards a person and can lead to pushing ahead with something or to accomplish something in a positive way, e.g. “he could hardly conceal his excitement she agreed” (The Free Dictionary, 2015i).
(0) No (1) Yes
Joy Positive Contentment Pleasure
Those emotions mean to be happy with one’s situation in life and feeling satisfied. This can also be felt in a situation, e.g. when a
(0) No (1) Yes
person experiences “the pleasure ofhis company” and therefore feels satisfaction (The Free Dictionary, 2015c).
Joy Positive Pride Triumph
Those emotions are “A sense of one's own proper dignity or value” and lead to self-respect.” It results in a high opinion about oneself (The Free Dictionary, 2015t).
(0) No (1) Yes
Joy Positive Eagerness Hope Optimism
Those emotions mean that someone has the feeling that everything will turn out to be alright or that a specific wish will be fulfilled, e.g. “in spite of his troubles he never gave up hope” (The Free Dictionary, 2015s).
(0) No (1) Yes
Joy Positive Enthrallment Rapture
“A feeling of great liking for something wonderful and unusual” and to be spell bounded, e.g. “the magic show enthralled us”(The Free Dictionary, 2015g).
(0) No (1) Yes
Joy Positive Relief Relief is the easing of a burden, pain or distress, e.g. “the news of their safe arrival came as a great relief” (The Free Dictionary, 2015v).
(0) No (1) Yes
Surprise Positive Amazement Surprise Astonishment
This emotion is caused when something unanticipated happens and one feels wonder about it. If one is surprises, they did not expect the situation, e.g. “for her surprise he said no” (The Free
(0) No (1) Yes
Dictionary, 2015z). One expresses surprise when saying “we could never have imagined…”
Anger Negative Aggravation Irritation Agitation Annoyance Grouchiness Grumpiness
Those emotions represent a mental condition and mean that someone is annoyed or irritated by someone or something, e.g. “having to stand in line so long was a real bummer” (The Free Dictionary, 2015m). This cluster is coded as yes when someone expresses that someone else’s behaviour is exorbitant, e.g. “
(0) No (1) Yes
Anger Negative Exasperation Frustration
Feelings of dissatisfaction that mostly result from unfulfilled needs or unresolved problems and therefore mean to experience being thwarted in attaining one’s goals, e.g. “her constant complaints were the main source of his frustration” (The Free Dictionary, 2015j). This cluster is coded as yes when a person express dissatisfaction about a company’s or person’s behaviour, e.g. “It’s bad that they did that” or “Unfortunately the media do not show respect anymore.”
(0) No (1) Yes
Anger Negative Anger Rage Outrage Fury Wrath Hostility Ferocity Bitterness Hate Loathing Scorn Spite Vengefulness Dislike Resentment
Those emotions are connected to strong negative feelings and are expressed in intense anger, or violent behaviour. It is also
connected to a high temper or even passionate madness, e.g. “she fell into a rage and refused to answer” (The Free Dictionary, 2015u). This cluster is coded as yes when a commenter or a representative of a company expresses their anger towards a company or a person, e.g. “I hope he will burn in hell.”
(0) No (1) Yes
Anger Negative Disgust Revulsion Contempt
“A great loathing or distaste
aroused by someone or something” and a strong feeling of disliking something, e.g. “this spoilt food disgusts me” (The Free Dictionary, 2015e).
(0) No (1) Yes
Anger Negative Envy Jealousy
“A feeling of discontent and resentment aroused by and in conjunction with desire for the possessions or qualities of another” (The Free Dictionary, 2015h).
(0) No (1) Yes
Anger Negative Torment Torment can be unbearable physical pain or extreme mental distress. Both are expressed in intense suffering, e.g. “it pains me to see my children not being taught well in school” (The Free
Dictionary, 2015ab).
(0) No (1) Yes
