Co-construction of a crisis
Online public’s crisis collaboration and thereof resulting management of strategic communication during crisis situations
Master’s Thesis
Graduate School of Communication, University of Amsterdam Master’s program Communication Science
Vera Spring 11181508 Supervisor: Dr. James Slevin Date of Completion: 03.02.2017
Abstract
During corporate crises, the communicative interaction between the organization in crisis, news media and the public becomes highly complex. Their messages have the potential to shape public opinion and influence the evolution of the crisis. Especially the public
experienced empowerment in this regard through the rapidly growing social media platforms. However, only little is known about online public’s contribution and influence to the
communicative discourse. Hence, it is crucial for communication managers to understand the communicative dynamics between these actors among different phases of the crisis in order to efficiently manage corporate crisis communication. This study applies a computer assisted semantic-network analysis on the recent Samsung Galaxy Note7 crisis, studying frame dynamics between Samsung, news media, and the online public on Facebook and Twitter. The findings document that over the course of the crisis, frames used by the analysed actors increasingly diverge form each other. Especially the online public communicates extensively about the analysed crisis. Faced with the fast moving social media activities of the online public, it is crucial for organizations in crisis to engage by the means of immediate crisis communication.
Keywords: crisis communication, framing theory, online public, automated content analysis
Co-construction of a crisis. Online public’s crisis collaboration and thereof resulting management of strategic communication during crisis situations
During corporate crises, the organization in crisis, news media and the online public play a key role for the construction and evolution of the crisis. Especially in our mediated society, different actors’ communication plays a fundamental role for crisis magnitude, evolution, and consequences (Sellnow & Seeger, 2013). While crises are often caused by external circumstances, actors socially co-construct the meaning of a series of events in their communicative interaction (Kleinnijenhuis, Schultz, Oegema, 2015; Schultz & Raupp, 2010). Confronted with a continuous stream about corporate scandals, actors try to make sense of these rapidly unfolding events (Weick, 1988). The organization in crisis needs to
communicate with its stakeholders in order to advise them how they can protect themselves from the crisis and to help them cope psychologically with the crisis (Coombs & Holladay, 2008). Moreover, the organization needs to communicate to protect its reputation, which is threatened by the crisis (Coombs, 2007). At the same time, news media provide crucial information (Sorribes & Rovira, 2011), and offer interpretations of the events (Van Gorp, 2007). In doing so, they can significantly influence other actor’s understanding of the crisis and guide their actions. Thus, news media can cause or prevent crisis escalation (Van der Meer, Verhoeven, Beentjes, & Vliegenthart, 2014) and consequently play a decisive role for the crisis development. In addition, the public, visible through social media, takes part in the negotiation about the crisis (Meyers, 2012). Recent studies found that the public perceives social media sites as being more credible than traditional news media (Johnson & Kaye, 2004; Sweester & Metzgar, 2007), and that the public turns primarily to social media sites for
immediate crisis information (Sung & Hwang, 2014). Even journalists increasingly access social media sources in order to gather information in times of crisis (Van der Meer, Verheoven, Beentjes, & Vliegenthart, 2016). Hence the online public can be considered of vital importance in crisis communication. Despite its relevance, the communicative dynamic
between these actors has been widely overlooked (Van der Meer et al., 2014). Thus, further research is needed to understand how these actors interact and co-construct the crisis.
To advance understanding of crisis situations and actors’ interactions, it is essential to unfold the underlying communicative process. Framing theory offers a powerful body of academic literature to study crisis communication patterns and their evolution over time (Schultz, Keinnijenhuis, Oegema, Utz, & Van Atteveldt, 2012). Particularly, analysing how frames vary in communication across actors promises critical insight into the communicative interaction and its consequences.
The framing of crises by relevant actors can have a major influence on the
development of crises (Schultz et al., 2012). Traditionally, the audience had no active role in the frame building process. Since the advent of social media, the public gained power to collaborate in crisis frame building (Castells, 2007; Van der Meer & Verheoven, 2013). So far, framing research has predominately concentrated on frames in media content and has mainly disregarded public’s role in the framing process (Borah, 2011). In particular, only little research has been conducted regarding the public framing of a crisis in social media and its timing and development (Schultz et al., 2012). Public social media expressions may challenge media frames by introducing different attributes (Liu, 2010). The relevance of differing frames lays in the influence established media outlets can have on stakeholders’ perception of organizations (Hallahan, 1999). Frames are able to direct public’s attention to certain aspects and to withhold others (McCombs, Shaw, & Weaver, 1997).
This knowledge gap may pose a threat to organizations as the public opinion
influences organizations’ reputation and image severely (Coombs, 2007). Without a prompt corporate reaction, i.e. framing of the crisis from the organization’s point of view (Holladay, 2010), the audience might build the initial frames on the basis of rumours. On the other hand, consumers may defend the organization on social media and therefore represent valuable sources of support. They might support the frame building process in favour of the
organization (Gummerus, Liljander, Weman, & Pihlström, 2012). In addition, the higher level of credibility assigned to public crisis coverage compared to traditional mass media (Sweetser & Metzgar, 2007), enhances public’s collaboration in framing events especially in crisis situations (Wigley & Fontenot, 2011).
This knowledge will help organizations to plan and manage their strategic
communication during a crisis and consequently helps them to protect their reputation in times of crisis. It also enlarges knowledge about the relative power of the three different actors regarding the communicative construction of the crisis event (Cornelissen, Carroll & Elving, 2009). Understanding the active role of the public in the process of meaning
construction helps organizations to react appropriately upon stakeholders’ requests and to utilise social media in crisis (Schultz, Utz & Göritz, 2011).
In sum, this study sets out to investigate crisis frames and their development across the actors, namely organization in crisis, news media, and the online public, by addressing the research question: To what extent do frames used by the organization in crisis, the online public and the news media differ during corporate crisis and how does this relate to different phases of the crisis?
Theoretical Framework
Volkswagen cheating in emissions tests, widespread corruption within FIFA, Nestlé selling unsafe Maggi noodles in India. No organization is immune to crises. Moreover, globalization leads to events, such as crises, “happening on a world scale” (Giddens, 2002, p. 4). For instance, products are launched globally and consequently crises are globally relevant. Furthermore, the omnipresence of social media and smartphones augmented the opportunities for information dissimilation about threats and missteps (Holladay & Coombs, 2013), and thus makes crises more visible by creating a global arena. Indeed, some social media tools (i.e. Facebook) seem to increase the possibility of detecting threats and organizational misconduct. In turn, same social media tools may also facilitate spreading rumours
concerning threats (Holladay & Coombs, 2013). Sources for such crises can origin from both, increasing legal liabilities as well as increasing moral liabilities corporations are facing
nowadays. In addition, offended stakeholders are more likely to generate crises (The changing landscape of liability, 2004) by actively monitoring organizations’ activities and raising their voices when suspecting misconduct (Coombs, 2002). From a sociological perspective, the high speed of change and expanding scope and growth of new modern institution resulting in risk culture and increased uncertainty in late modernity (Giddens, 2002), enhance the
likelihood of crises. All together, this leads to a continuous stream of information about corporate crisis.
Communication during corporate crisis
A corporate crisis is the “perception of an unpredictable event that threatens important expectancies of stakeholders related to health, safety, environmental, and economic issues, and can seriously impact an organization’s performance and generate negative outcomes” (Coombs, 2015, p. 3). Such events are characterized by low probability and severe
consequences for the organization in crisis. Its most fundamental objectives (Weick, 1988) and its reputation are at risk (Coombs, 2015). On one hand, crises have the potential to create comprehensive disruption resulting in great harm. On the other hand, they can also represent opportunities for “constructive change, growth and renewal” (Sellnow & Seeger, 2013, p. 1).
In the context of a crisis, the aspect of communication is generally acknowledged to be highly important (Coombs, 2007). Although “crisis have real origins, they are constituted in the communicative interplay between several actors, whose perception of a series of events as a “crisis” produces real consequences” (Kleinnijenhuis et al., 2015, p. 4) for corporate and public actors as well as news media. Research has repeatedly demonstrated that
communication can have an impact on magnitude, duration, and consequences of corporate crises (e.g. Cornelissen et al., 2009; Schultz et al., 2012), which is why it is imperative to understand the role of communication among different actors during crises.
The advent of social media also influenced crisis communication. Members of the public are no longer passive recipients of information from traditional news media but have become active communicators seeking out and generating information (Coombs, 2015). Ultimately, they both instigate pressure on organizations and present allies during crisis situations.
Framing
Framing theory offers a powerful approach to investigate the dynamic process of meaning construction and negotiation (Gamson & Modigliani, 1989). Underlying assumption is that a single issue can be regarded from multiple angles (Chong & Druckman, 2007). Frames in communicating text (e.g. press releases, news articles) highlight some information of an issue and thereby make it more noticeable and meaningful to the audience in order to influence the receiver’s opinion about the issue (Entman, 1993; Van Gorp, 2007). By using and repeating certain key words and information sources, constructs of facts and judgements can be thematically reinforced (Entman, 1993). Thereby frames have a direct impact on how people perceive a certain issue (Coombs, 2015). Thus, by providing a certain perspective through which people interpret a given issue, frames are a critical element in the construction of social reality (Hallahan, 1999).
Communicative material can include explicit as well as implicit1 frames. Since only the analysis of implicit frames allows to study the development of frames over time and to compare frame usage across actors (Hellsten, Dawson, & Leydesdorff, 2010), the focus of this paper lies on implicit frames adopted by aforementioned three actors.
Durham (2001) notes that even though frames help to organize social interactions, they bear the risk of neglecting cultural and ideological diversity. By analysing U.S. newspapers and social media sites, this paper tries to circumvent cultural influences. Likewise, by assessing frames contributed to the crisis communication by different actors, ideological diversity can be preserved to a certain extent. An additional point of critique
regarding the analysis of frames is, that for traditional content analysis, frames need to be pre-defined in order to be detected. This may lead to biased results as frames may be framed by the researcher. However, in this research applied automated semantic-network analysis caters for a revealing of frames in an inductive, exploratory manner, whereby the pre-definition of frames can be avoided and subjectivity excluded (Leydesdorrf & Hellsten, 2005).
Crisis framing
Frames also play a critical role in crisis situations. Crises are characterised by a high level of uncertainty and ambiguity for all involved actors as an organization often lacks information in the beginning of a crisis (Coombs, 2015). During organizational crisis situations, sensemaking2 activities rapidly develop resulting in frames that guide people’s meaning construction of complex events (Benford & Snow, 2000; Van Gorp, 2007), and allowing actors to come to an understanding of the crisis (Weick 1988). Thus, frames help actors to organize a crisis event (Hellsten et al., 2010) and reduce uncertainty (Leydesdorff & Ivanova, 2014) as well as complexity (Van Gorp, 2007). This makes frames fundamental for the prevention of crisis escalation (Seeger, 2002) and public confusion or panic (Van der Meer & Verheoven, 2013).
Organizations’ crisis framing. Organizations naturally make an effort to frame their response to the crisis favourably (Sellnow & Seeger, 2013). The frame-building process is fundamental to the formation of organizational reputation (Coombs, 2007). Therefore, as well as due to their profit-oriented interests, organizations will frame the crisis event in such
manner that is likely to be accepted by stakeholders and represents the organization in the best possible way (Cornelissen et al., 2009). Since the way information about the event is framed affects public perception of the organization, it is crucial for organizations to participate in the framing process (Holladay, 2010). The goal is to influence the way in which news media frame the crisis and ultimately, to trigger positive evaluation of the organization by stakeholders and the public (Coombs, 2007). Thus, organizational framing is regarded as
successful when frames used in organizational press releases resonate as intended in the communication of other actors (Van der Meer, 2014).
News media’s crisis framing. During crisis, news media traditionally serve as primary platform for communication and information distribution (Sellnow & Seeger, 2013). News media have the potential to shape public’s perception of crisis events by highlighting certain aspects of the issue (Gitlin, 1980; McCombs, 2005) and offering interpretations of the event (Van Gorp, 2007). They either accept or reframe information provided by the
organization in crisis (Sellnow & Seeger, 2013). Oftentimes, the public’s meaning construction is based on information presented in news media (Sorribes & Rovira, 2011). Especially in crisis situations, characterized by high need for information, journalists are important gatekeepers influencing the crisis through their selection of issues and sources (Van der Meer et al., 2016). Moreover, they influence public’s construction of reality by making sense of events organising them into meaningful succession (Hallahan, 1999; Tuchman, 1978). Thus, news media can encourage or prevent crisis escalation (Van der Meer et al., 2014) and ultimately impact the evolution of the crisis (Kleinnijenhuis, Schultz, Oegema, van Atteveldt, 2013).
Online public’s crisis framing. By providing a more equal distribution of resources, the Internet empowered the public with a platform to engage in mass-to-mass communication (Castells, 2007). Social media enables the public to distribute and access information,
whereby it actively participates in activities of public relations (Smith, 2010) as the public content reflects on organizational reputation (Kelleher, 2009). Thus, social media empowered the public to actively contribute to the communicative construction of a crisis event. The online public makes use of the new empowerment and increasingly voices concerns regarding organizations (Coombs, 2002). Facing uncertainty and lack of information, the public might base its initial crisis frames on personal speculations (Van der Meer & Verheoven, 2013). How online publics in social media interpret a crisis in distinct ways is highly relevant in
order to evaluate the reputation threat the organization in crisis is facing and to craft crisis responses strategies accordingly (Coombs, 2015). The public can challenge news media frames by introducing alternative attributes (Liu, 2010) and influence public perception about an organizational crisis by contributing responses (Sweetser & Metzgar, 2007). Since people seeking information about a crisis mainly turn to social media sources (Liu, 2010; Sung & Hwang, 2014), online publics can play a crucial role in dissimilating crisis information and ultimately in shaping public meaning construction and influencing crisis development (Chong & Druckman, 2007).
Crisis framing across actors. Since the meaning of a certain event or issue can be constructed or framed from different perspectives (Van Gorp, 2007), and the sensemaking process is influenced by the respective professional identity (Cornelissen et al., 2009), it is likely that different actors establish competing interpretations and meanings of the crisis (Chong & Druckman, 2007). Therefore, they vary in their level of similarity and existence in communication over time. The problem of competing frames derives from the roles of different actors regarding opinion formation of the public. Especially during crises, frames that are communicated by the news media traditionally have a serious impact on public’s perception of the organization in crisis (Hallahan, 1991). In contrast, research also
demonstrated that the way in which the public discussed an issue online was significantly related to the way in which the media covered it (Zhou & Moy, 2007). Given the relevance of crisis frames, assessing whether and how meaning construction, displayed through frames, differs between actors can offer insights into the interactive negotiation of reality among actors. This is of vital importance to crisis communication. Particularly for corporate public relations it is essential to understand how the media interacts and endorses frames of other actors. Moreover, since news media are the primary influencer of public opinion (McCombs, 2005), it is crucial to understand if the online public on social media has an impact on news media framing of a crisis.
Although the assumption of differences in the framing of a crisis between actors, it is necessary to achieve collective sensemaking and coherent understanding of the complex situation in order to overcome the crisis (Hellsten et al., 2010; Kleinnijenhuis et al., 2015). Since actors will try to reduce uncertainty and ambiguity, and competing frames will evoke a need to provide coherence to the issue (Hellsten et al., 2010), actors might start to mutually influence, borrow, and align their frames over the course of the crisis (Leydesdorff & Ivanova, 2014; Van der Meer et al., 2014). As organization, online public, and news media start to react upon each other and negotiate meanings of crisis events (Benford & Snow, 2000; Vliegenthart & Van Zoonen, 2011), a coherent understanding of the elements that build a frame may arise. Even though this does not result in a complete overlap of frames, it may lead to temporary frame alignment between actors regarding the meaning of the crisis (Van der Meer et al., 2014).
Literature suggests that frames are not static but they develop over time with the coverage of a specific event (Hellsten et al. 2010). When actors have come to a shared understanding of the crisis, they will return to pursue their respective interests and professional goals. This means that their frames would diverge and focus on new communicative aspects related to their professional identity (Van der Meer et al. 2014). Ultimately, this process would again lead to frame diversity. To further address the dynamic character of frames, this paper analyses frame development across actors over the course of different crisis stages.
Frames and crisis stages
Frames develop over time with the coverage of a specific event (Hellsten et al. 2010). Likewise, frames constructed by actors involved in organizational crises can change over the course of different crisis stages as the amount of available information changes
(Kleinnijenhuis et al., 2015). Based on Fink’s (1986) crisis phase model, this paper
time. As this study aims to analyse communication across organization, news media, and online public during an existing crisis, the first stage dealing with crisis prevention is going to be disregarded. Only stages two to four are relevant and will be included.
Acute crisis stage. This early phase of the crisis is especially important for crisis managers since it is characterised by confusion (Coombs, 2015) and high media attention (Lerbinger, 2011) resulting in high framing impact on public perception (Regester & Larkin, 2008). However, what makes the initial phase critical is the high level of uncertainty3 that generically accompanies a crisis. The lack of information can lead to rumours and
speculations, which can have lasting negative impact on organizations in crisis (Horsley & Barker, 2002). Rumours arise through social sensemaking directed at reducing uncertainty (Bordia & DiFonzo, 2014). Taking into account actors’ rapidly emerging sensemaking activities based on their respective professional and personal identities (Cornelissen et al. 2009), it is likely that in the beginning of a crisis, organization, news media, and online public apply different crisis frames in their communication.
Empirical research has shown that frames during initial stage of the BP oil spill crisis actually differed between the organization and news media (Schultz et al., 2012).
Furthermore, analysis of the financial crisis in 2008 disclosed competing implicit frames constructed by private and public organizations differed during initial crisis communication (Schultz & Raupp, 2009). Moreover, research on various Dutch companies in crisis that occurred between 2000 and 2011 found differing implicit frames from organizations, news media, and the public during the initial phase of the crises (Van der Meer et al., 2014). Etter and Vestergaard (2015) support this notion by demonstrating that all actors, i.e. the
organization, news media, the public on Facebook, as well as NGOs differed in their initial framing of Nestlé’s KitKat crisis. Therefore, the first research question and hypothesis read as follows:
Research Question 1 (RQ1): Which frames are constructed by organization in crisis, online public, and news media during the acute stage of the crisis?
Hypothesis 1 (H1): Implicit frames applied by organization, news media, and online public differ from each other during the acute stage of a crisis.
Chronic crisis stage. As a necessity to overcome the crisis, all involved actors will direct their communication at a consistent understanding of the crisis (Seeger, 2002). Considering the interaction and mutual influencing among actors through the “collective process of negotiation over meaning” (Vliegenhart & Van Zoonen, 2011, p. 111), it can be assumed that frames used by the three actors become more similar during this phase as the crisis progresses and available information increases and uncertainty decreases (Leydesdorff & Ivanova, 2014; van der Meer et al. 2014).
Empirical findings support the notion that frames align across different actors.
Research on the financial crisis 2008 (Schultz & Raupp, 2009), the BP oil spill crisis (Schultz et al., 2012), the explosion of a Dutch chemical plant (Van der Meer & Verheoven, 2013), and on several Dutch companies in crisis (Van der Meer et al., 2014) demonstrated the alignment of frames over time across all involved actors. Based on these findings, this paper expects that same dynamic applies to frames used by organization in crisis, online public, and news media in the chronic stage of the crisis. Therefore the second research question (part a) and hypothesis read as follows:
Research Question 2a (RQ2a): Which frames are constructed by organization in crisis, online public, and news media during the chronic stage of the crisis?
Hypothesis 2 (H2): Implicit frames applied by organization, news media, and online public differ less from each other during the chronic stage of a crisis compared to the acute stage.
Since less frame diversity is assumed during the chronic stage, ultimately not all frames will survive. Previous research delivered mixed implications regarding the question
whose frames will survive the acute crisis stage. Van der Meer and Verhoeven (2013) found that after initially forming their frames based on personal speculations, the public adapted frames provided by news media. Same findings were drawn from the analysis of Nestlé’s KitKat crisis (Etter & Vestergaard, 2015). On the other hand, news media were found to adopt online public’s frames in their communication about socio-political crises in China (Zhou & Moy, 2007). Supporting the second notion, a big data analysis showed that frames constructed by the online public can influence news media coverage (Neumann, Guggenheim, Jang, & Bae, 2014). Moreover, research demonstrated that journalists increasingly turn to social media sources in crisis situations (Van der Meer et al., 2016), leading to the assumption that in doing so news media might adopt frames provided by the online public. A third way of possible frame adoption was found in communication about the fair-trade policy of Max Havelaar, where the public adopted the explicit frame provided by the organization even though frames initially differed (Van der Meer, 2014). In order to shed light into this dynamic of frame adoption, part b of the second research question reads as follows:
Research question 2b (RQ2b): Which actor’s frames are adopted by the other actors during the chronic stage of a crisis?
Crisis resolution stage. During the last stage of the crisis, organizations rebuild their reputation and return to their daily routine (Fink, 1986). Ideally, all uncertainty factors have been resolved. However, since initial confusion and focus on the crisis has dissolved in mutual understanding (van der Meer et al., 2014), this last phase gives the news media and online pubic time to assess and evaluate how the organization dealt with the crisis. Again all actors will follow their professional identity when communicating. Therefore, actors might endorse new perspectives and issues in their communication that are no more comparable across actors (van der Meer et al., 2014).
Research showed that after an alignment of frames, where the public adopted frames provided by news media, implicit frames again differed between these two actors during the
last phase of the crisis. Thus, when sufficient information is available, the public will again personalize its crisis frames (Van der Meer & Verheoven, 2013). Adding to that, Van der Meer et al. (2014) showed with their analysis of several Dutch crises, that implicit frames are no longer aligned between organization, public, and news media during the last stage. Based on these considerations, the third research question and hypothesis read as follows:
Research question 3 (RQ3): Which frames are constructed by organization in crisis, online public, and news media during the resolution stage of the crisis?
Hypothesis 3 (H3): Implicit frames applied by organization, news media, and online public differ more from each other during the resolution stage of a crisis compared to the chronic stage.
Together, RQ1-3 and H1-3 will deliver insights into how newspapers and online public frame a corporate crisis during different stages in contrast to the organization uttering the crisis. This will help organizations to understand where and how to address their crisis communication. Depending on the standpoint of the online public as supporter or detractor of the organization, organizations can construct their crisis initiatives. Same applies for
newspapers. If there is one dominant actor whose frames are adopted during the chronic crisis stage, organizations may concentrate their frame-building efforts in its direction. Figure 1 visualizes the three hypotheses.
Method Data collection
To answer the research question, an automated content analysis of U.S. newspaper coverage, corporate press releases and social media posts, and social media posts and comments published by the online public was conducted.
Samsung’s recent crisis evolving around exploding Galaxy Note7 smartphones served as a case study. Samsung is South Korea’s biggest business conglomerate comprising
numerous subsidiaries and affiliated businesses (see Samsung Newsroom, 2016). Samsung Electronics is market leader in the smartphone sector (IDC, 2016). It was selected for several reasons, (1) the crisis including exchange programs, software updates, up to total recall of devices provides an exclusive opportunity to study frames over the course of different crisis stages, (2) all three actors were communicatively participating and (3) several people were physically harmed by the crisis and it poses a threat to a vast amount of smartphone users. The crisis began on August 24, 2016, when the first case of an exploding Samsung Galaxy Note7 device surfaced in South Korea. Since then, Samsung announced a global recall of 2.5 million Note7 phones, citing faulty batteries, airlines banned Note7 devices from their planes, public transportation companies prohibited to carry such device on board, telecommunication companies halted issuing new Note7 phones due to safety issues, and Samsung permanently halted sales and production of these smartphones and asked customer to stop using their phones (Lee & Pak, 2016). Communication material was therefore collected between August 24, 2016, until the end of data collection on November 21, 2016.
The computer assisted semantic-network analysis applied in this research allows to examine large numbers of communication material. Therefore, it was possible to analyse a census sample, which in turn facilitates the study of frame development over time. The census consists of the entire communication material about Samsung’s Galaxy Note7 crisis that has been published by these three actors in the selected media (N = 28,664). A detailed table
presenting the number of articles/posts collected per actor and medium is provided in Appendix A.
For news media, eight U.S. newspapers with highest circulation rates (MediaMiser, 2016), available with complete articles on LexisNexis, were selected. Relevant articles were selected using a search string including the words ”Galaxy Note7” and “Note7” (see
Appendix B) (n = 112). Selected newspapers are listed in Appendix A.
Samsung’s press releases were manually retrieved from its newsroom on the official corporate website. In addition Samsung’s communication on the social media platforms Twitter and Facebook was included in the sample. These two social media platforms were chosen due to their general relevance. Both platforms rank under the fifth largest social media platforms by number of their active daily users (excluding Chinese platforms, Statista, 2016). Corporate tweets from account Samsung Mobile US (n.d.) were retrieved via the website allmytweets.net4. Relevant tweets were selected with help of a keyword search. Corporate messages on Samsung’s U.S. Facebook page (Samsung Mobile USA, n.d.) were extracted manually. The Samsung Mobile sites were chosen as the main discourse about the Galaxy Note7 evolved on these specific sites (e.g. not on Samsung). In total n = 101 corporate messages from Samsung were included in the analysis.
Public social media messages posted on Samsung’s Facebook were extracted with help of a Python script. Relevant messages were selected with help of a keyword search. Again, Facebook page Samsung Mobile USA (n.d.) served as basis. Public messages on Twitter were identified with help of Twitter Search API (see Appendix C) and downloaded with a Python script. This approach comes with a major limitation. Twitter selects and displays tweets that are most relevant to the respective user based on location, language, friends etc. (Developers, n.d.). Due to the scope and resources of this research, this limitation was accepted. A random sample of social media posts was manually checked in order to
validate the material. In total, n = 28,451 messages of the online public were included in the analysis.
Research Units
In order to analyse the role of crisis phases and the development of frames over time, the data was split into different groups according to their date of publication and was analysed separately. The acute first phase is defined as the first eight days (24.08.2016-01.10.2016), starting with the day the first report about an exploding Note7 smartphone surfaced and ending on the day Samsung recalls 2.5 million Note7 smartphones marking the first official statement by Samsung. In this specific case, eight days represent an accurate timeframe to get an image of the initial crisis communication, as little official information was available and several reports about exploding devices surfaced. The second phase, the chronic stage, lasts for 40 days (02.10.2016-11.10.2016). Characterised by temporary procedures until normal operations can resume, Samsung launched several exchange initiatives and system updates, ending with the permanent halt of production and sales of the Galaxy Note7 model during this phase. The last phase, the resolution stage, covers processes of image restoration. In the analysed case Samsung published apologies and mission statements in order to rebuild its reputation. Limited by the end of data collection this phase comprises 41 days (12.10.2016-21.11.2016). An overview of the research units with respective number of analysed materials is provided in Appendix D.
Automated semantic-network analysis
This study examines implicit frames as latent patterns of words that (co-)occur in communication material about Samsung’s Galaxy Note7 crisis. Since implicit frames cannot be analysed by traditional, manually coded content analysis (Hellsten et al., 2010), this paper applies a type of automated content analysis to understand meaning construction of crisis situations through words and their relative position to each other. This so-called automated semantic-network analysis determines implicit crisis frames as word networks by following
an algorithmic approach. This method allows to identify semantic changes in communication materials of different actors by measuring the meaning of words in their context. It is based on the idea that a text can be seen as a network of words that transports their meaning, with each network serving as an indicator for the frame they represent (Hellsten et al., 2010). By clustering groups of correlating words, a higher-order structure in and between texts can be identified and frames detected (Hellsten et al., 2010). This method allows to identify implicit frames in an inductive way, avoiding predefining frames as required for traditional content analysis. Therefore, as means of this study, frames will be operationalised as (co-)occurrences in communication material, generating latent semantic-networks that transmit their meaning (based on Hellsten et al., 2010).
Following a scientifically published manual by Vlieger and Leydesdorff (2011), several steps involving different software programs were taken in order to obtain the data for the automated approach. A detailed protocol of the process is provided in Appendix E.
After preparing the data ending with a principal component factor analysis, an overall image of the frame usage from Samsung, the online public, and newspapers was identified. Due to the large number of different implicit frames, the qualitative analysis focused on the three most dominant frames. These were selected based on the highest eigenvalue (EV) and the highest explained variances in each research unit (Field, 2009). This procedure is based on interpretative labelling of word clusters, therefore is a subjective comparison of frames. In this regard, this research compares implicit frames based on statistical indicators. Following an analytical process developed by Van der Meer et al. (2014), factor loadings of words that are used by two actors at a certain stage of the crisis were compared to each other (see Appendices F and G). The factor loadings indicate the degree of importance of the word in the implicit frame. The correlation coefficient Spearman’s rho implies the extent to which the words had the same importance in meaning provision and framing for two actors during a
specific stage of the crisis. Since this paper’s analysis is based on a census sample, correlations can be assumed relevant rather than the significance level.
Results
In the following section, identified frames used in communication material of Samsung, eight U.S. newspapers, and the online public were compared to each other. The analysis was performed separately for each crisis stage in order to expose frame development across crisis stages.
Table 1
Correlations among actors during different crisis stages.
Samsung/ Public Samsung/ Newspaper Public/ Newspaper Spearman's rho correlation coefficient ρ
Acute crisis stage .21 -- --
Chronic crisis stage -.02 .13 -.19
Crisis resolution stage -.04 -.36 .20
RQ1 asked for an investigation of frames used by the three actors during the acute crisis stage. However, while conducting principal component factor analysis, it became evident, that it was not possible to retrieve valid factor components for the communication material published by the newspapers (based on Cronbach’s α test). This might be due to the small research unit (n = 3). Therefore, only implicit frames used by Samsung and the online public were compared. For this description, frames were interpreted qualitatively by
comparing the three most dominant frames (with highest eigenvalue) from both actors. Looking at the frames produced by Samsung and the online public, it is observable that the thematic emphasis differs between actors. Samsung made use of frames that can be named “investigation frame” (EV = 9.53, Cronbach’s α = .99), “smartphone attributes frame“ (EV = 3.71, Cronbach’s α = .91), and “advertising frame” (EV = 1.82, Cronbach’s α = .79). Online
public’s frames can be interpreted as response to advertisement and features of the
smartphone but also emphasise reports about defect devices. The three most dominant frames are named “fan frame” (EV = 3.40, Cronbach’s α = .87), “demand frame” (EV = 3.13,
Cronbach’s α = .88), and “defect frame” (EV = 3.02, Cronbach’s α = .79). Interestingly, the crisis-related frames from Samsung ranks first regarding the eigenvalue, whereas the online public’s crisis-related frame ranks third. During the first phase of the crisis, words forming the dominant frames were complementary between these two actors. Samsung framed the crisis with words such as ‘ensure’ (1.00), and ‘thorough’ (1.00), ‘quality’ (1.00), ‘scanner’ (.85), indicating that Samsung initially focused on promoting their new smartphone but also reacted to reports about exploding phones reassuring their diligent investigation of the issue. The online public formed their frames using words like ‘chance’ (.84), ‘win’ (.78), ‘supply’ (.86), and ‘blame’ (.80), demonstrating that there was a high demand of the smartphone but also beginning critique directed at Samsung.
H1 stated that implicit frames used by the three actors differ from each other during the acute stage of a crisis. In addition to the qualitative interpretation above, H1 was addressed quantitatively. Again, newspapers were not included into the analysis, thus the statistical comparison was only conducted between Samsung and the online public. Correlations of factor loadings between the two actors were compared with correlation coefficient Spearman’s rho. Table 2 shows that the correlation between Samsung and the online public during the acute crisis stage is ρ = .21. The coefficient indicates a weak correlation between the two actors. Low factor correlations between actors suggest that mutually used words differ in their importance when it comes to framing the crisis situation (Van der Meer et al., 2014). The weak correlations of word importance between implicit framing of Samsung and the online public can also be illustrated with help of specific words. For instance, the words ‘battery’, ‘reported’, and ‘time’ produced considerably higher factor loadings in frames of Samsung (.74, 1.00, .79) than in online public’s frames (.38, .20, -.10).
Consideration of the qualitative and quantitative interpretation of the data results in the conclusion that frames used by Samsung and the online public differed during the first phase of the crisis. However, since the newspapers were not very active at that time and no implicit frames could be shown in their communication material, it was not possible to compare implicit frames across all three actors. Therefore, H1 can only be partially supported.
In order to answer RQ2a, most dominant frames used by the three actors during the chronic crisis stage were again interpreted qualitatively. Newspapers increased their
participation during this phase (n = 45). Thus it was possible to include all three actors into the comparison. In general, the most dominant frames evolve around the exchange/recall topic. Samsung again applied frames that can be called “investigation frame” (EV = 30.73, Cronbach’s α = .96), and “smartphone attributes frame” (EV = 30.60, Cronbach’s α = .98), as well as a new “exchange program frame” (EV = 20.18, Cronbach’s α = .95). Frames from the online public partially overlap with those from Samsung, in particular when it comes to the exchange program. However, the public also directs attention to safety issues. Moreover, only two significant (based on Cronbach’s α) frames were identified in online public’s
communication material. These were “exchange program frame” (EV = 3.12, Cronbach’s α = .66) “safety frame” (EV = 2.79, Cronbach’s α = .78). The newspapers introduce new topics by emphasising consequences provoked by the exploding phones for consumers as well as for Samsung. Their frames can be named “market frame” (EV = 13.50, Cronbach’s α = .89), “consequence frame” (EV = 10.79, Cronbach’s α = .87), and “recall frame” (EV = 10.48, Cronbach’s α = .88). Dominant frames contain words such as ‘collaborating’ (.97), ‘pen’ (1.00), ‘feature’ (1.00), ‘refund’ (.99), indicating that Samsung is both, still promoting its phone and at the same time demonstrating generosity regarding the solution of the problem. In contrast, the online public is exchanging information regarding safety issues and the exchange program. Words forming their dominant frames are for instance ‘participate’ (.82), ‘info’ (.76), ‘commission’ (.84), and ‘safety’ (.73). Again, words composing the dominant
frames in newspapers illustrate the new perspective they introduce. These are ‘screen’ (.73), ‘industry’ (.69), ‘brand’ (.68), ‘administration’ (.80), and ‘aviation’ (.79).
To assess H2, which assumes implicit frames by the three actors differ less from each other during the chronic stage of a crisis than during the acute stage, the similarity of frames used by the actors were compared between the two phases. As newspapers could not be included in the quantitative analysis of the first stage, it is important to look at correlations of factor loadings of mutually used words by Samsung and the online public first. This allows for comparison of similarity of these frames with the ones found for the first stage of the crisis. During the second stage, the correlation between Samsung and the online public is ρ = -.02, indicating a fundamental deviation in word importance between the two actors regarding the formation of frames. Compared to the correlation coefficient from the acute crisis stage (ρ = .21), the frames were substantially less similar than during the acute crisis stage.
As illustrated in table 1, factor loading correlation during the chronic stage between Samsung and newspapers is ρ = .13, and between online public and newspapers ρ =
-.19. These correlations suggest that the same words used by Samsung and newspapers did have a slightly similar importance in meaning provision, whereas mutually used words by online public and newspaper did not have the same importance in establishing meaning and frames. Some exemplary mutually used words illustrate these different correlations. The words ‘advise’, ‘continue’, and ‘replacement’ load highly different on frames from Samsung (.88, .62, .64) and frames from the online public (-.12, -.12, -.22). In contrast, the words ‘affected’, ‘battery’, and ‘continue’ show more similar factor loadings on frames from Samsung (.68, .56, .62) and newspapers (.54, .47, .51). Finally, words such as ‘airplane’, ‘burn’, and ‘flight’ highly differ in their importance of determining implicit frames from online public (-.27, .70, -.37) and from newspapers (.63, .35, .63). Overall, the correlation across all actors is ρ = (-.02 + .13 + -.19)/3 = -.03, representing a negative correlation of words importance among all actors.
Since only the factor loading correlations of frames from Samsung and the online public can be compared between the chronic and the acute stage of the crisis, H2 is only partially rejected. However, inclusion of the newspapers, and therefore the correlation pairs Samsung-newspaper and public-newspapers, as well as the qualitative interpretation support the rejection of H2 by showing that implicit frames across all actors are not aligned during the chronic crisis stage.
Since the assessment of H2 showed that overall, the three actors did not use similar frames, answering RQ2b became redundant. If frames are not aligned between actors during the second phase of the crisis, none of the actors’ frames established during the acute stage were adopted by the others during the chronic crisis stage. In contrast, newspapers starting to participate in discursive crisis communication lead to the introduction of new frames.
A qualitative interpretation of dominant frames used by the three actors during the crisis resolution stage was undertaken to answer RQ3. This analysis presents a mixed picture. On one hand some actors use common frames, i.e. the “recall frame” was prominent in communication material from Samsung (EV = 23.51, Cronbach’s α = .98) as well as from the online public (EV = 3.27, Cronbach’s α = .71), and the “airline ban frame” was dominantly applied by both, online public (EV = 2.66, Cronbach’s α = .70) and newspapers (EV = 13.40, Cronbach’s α = .92). On the other hand, each actor introduced its own frames. Samsung focused on the future outlining quality and research standards they plan to adhere, “mission frame” (EV = 12.29, Cronbach’s α = .88), but they also summarized financial losses due to the crisis with the “economic consequence frame” (EV = 17.23, Cronbach’s α = .97). In contrast, the online public included strong personal opinion about the recall in their discussions, which is expressed by the “recall judgement frame” (EV = 3.64, Cronbach’s α = .72). The
newspapers extended the coverage about the crisis by adding the frames “federal ban frame” (EV = 14.33, Cronbach’s α = .92) and “legal frame” (EV = 11.84, Cronbach’s α = .93). The most important words forming newspapers’ frames are ‘banned’ (.888), ‘department’ (.89),
‘air’ (.85), and ‘court’ (.97). This shows that newspapers put emphasis on the smartphone ban issued by single airlines evolving to a federal ban. But newspapers also started to discuss legal consequences for Samsung evoked by the crisis. Dominant words attributed to the frames present in the communication material of Samsung are ‘exchanged’ (1.00), ‘inconvenience’ (1.00), ‘profit’ (1.00), ‘cost’ (1.00), ‘service’ (.99), demonstrating the company’s focus on financials and improvement. Words building the online public’s most dominant frame
illustrate the personal aspect in the communication material, ‘debated’ (.95), ‘ignoring’ (.95), and ‘rushed’ (.88).
H3 assuming that implicit frames used by organization, news media, and online public differ more from each other during the resolution stage of a crisis compared to the chronic stage, was assessed by comparing the correlation of factor loadings of frames from the pairs Samsung-online public, Samsung-newspapers, and public-newspapers between this two stages. The average correlation of factor loadings among the three actors is ρ = (-.04 + -.36 + .20)/3 = -.07. Comparison with the chronic crisis stage (ρ = -.03), leads to the conclusion that differences in word importance increased during the crisis resolution stage, and hence implicit frames. Additionally the pattern of decreasing correlation can be shown on the actor-pair level: Samsung-online public (ρ2 = -.02; ρ3 = -.04) and Samsung-newspapers (ρ2 = .13; ρ3 = -.36). Interestingly, for the pair online public-newspapers similarity of implicit frames
increased during the two stages (ρ2 = -.19; ρ3 = .20). The overall as well as the detailed tendency can be underlined when looking at specific words and their importance in forming frames. For example, factor loadings of all actors remarkably differ for the words ‘credit’ and ‘replacement’, whereas only Samsung ascribes ‘exchange’ a different word importance: Samsung (1.00, .95, .98), online public (.14, .36, .45), newspapers (-.29, .66, .49). In sum, H3 is supported.
Conclusion
Frames play a critical role in the development of a crisis. A collective understanding of the situation, expressed through mutually used frames is a necessity in order to overcome a crisis (Seeger, 2002). With the advent of social media, the online public became an important actor collaborating in crisis frame building (Castells, 2007). Although public opinion built upon dominant frames influences corporate reputation and image severely (Coombs, 2007), little is known about the interaction of this newly empowered actor with more established actors such as organizations and news media. Previous research suggests that after initially diverse crisis frames established by the three actors in their individual sensemaking, actors’ frames become more similar in the course of collective sensemaking in the proceeding of the crisis. During the last phase of the crisis frames again diverge (Van der Meer et al., 2014). In order to advance knowledge about the frames produced by these actors and their
development, the research presented in this paper pursued to answer the question whether and to what extent frames used by news media, the online public, and the corporation going through a crisis differ from each other and how this relates to different phases of a crisis.
This study shows that during the acute crisis stage, implicit frames used by Samsung and the online public differ. This result indicates that in the beginning of the crisis, both actors individually try to make sense of the underlying situation through their first communication. The immediate emerging sensemaking and initial production of frames might therefore follow the respective professional or personal identity, which results in competing frames. This finding is in line with the theoretical concept of sensemaking (Cornelissen et al. 2009; Weick, 1988) and confirms previous research on frame-building between these two actors (Van der Meer et al., 2014). However, qualitative interpretation of the frames established by these two actors indicates, that frames used by the online public can be understood as answers to
Samsung’s frames and do not entail completely new aspects or interpretations of the situation. In addition, dominant frames of both actors not only evolved around the crisis but also around
the excitement aroused by the new smartphone. Interestingly, newspapers were not very active in their communicative role during the first phase of the crisis. Thus, no news media frames could be compared with frames of the two other actors during the initial crisis phase. This lag in participation may be attributed to characteristics of newspaper production, such as the news gathering process (Blood, 2003) and traditional journalism values, i.e. accuracy (Society of Professional Journalists, 2014). Consequently, public opinion formation started out on basis of corporate communication material and public discussion in social media.
Furthermore, this research found that during the chronic stage of the crisis, frames of Samsung and the online public differ even more compared with the acute stage. This finding contradicts existing research on framing by these two actors, which found frame alignment between company and public during the second phase of the crisis (Van der Meer et al., 2014). This result may indicate that the two actors did not start to engage in mutual
influencing and borrowing of frames (Vliegenhart & Van Zoonen, 2011). Instead, the result suggests that organization and online public focus on different aspects of the crisis situation. The absence of a collective effort to understand the crisis could indicate that the public has another meaning of the crisis than the organization has. In addition, newspapers began to actively participate in the communication about the crisis during this period. They introduce new frames and perspectives to the discussion and in doing so enlarge communication’s scope. Again, this finding objects previous research focusing on frame development across actors over time, which found frame alignment during the second phase of crises between these (Van der Meer et al., 2014). However, newspapers’ role to extend the overall discourse about a crisis is in line with previous research on framing of Nestlé’s KitKat crisis, where newspapers were found to contribute a more diverse and balanced coverage (Etter & Vestergaard, 2015). It could be argued that because of the delayed participation of
newspapers, the mutual influencing, resulting in similar frames, is lagged and will occur at a later stage of the crisis. The dissimilarity in frames used by the three actors is also indicated
by the qualitative analysis. Whereas the organization’s frames imply a focus on actions directed at resolving the crisis, the online public’s frames suggest emphasis on the safety issue. Frames used by news media indicate a broader analysis of the crisis, for instance by outlining possible market consequence. Consequently, the actors did not adopt each other’s frames in the course of meaning negotiation during the chronic stage of the crisis. In terms of communication contributing to public opinion, the entrance of newspaper coverage adds new perspectives and aspects to this process unfolding over time. After a first period of
undisturbed negotiation between the organization in crisis and the online public, news media provide new stimuli and ideas to the ongoing discourse of reality construction.
Finally, during the crisis resolution stage, frames used by the three actors differ more from each other than during the chronic stage. This finding is in line with previous research, which found a de-alignment of frames between these three actors during the third and final phase of a crisis (Van der Meer et al., 2014). However, the de-alignment in this case preceded an alignment in the second phase of the crises, which is not the case for this study. Moreover, this finding contradicts above mentioned suggestion that an alignment of frames is deferred. In addition, qualitative interpretation of the frames adds to this notion by showing that during the crisis resolution stage, the company focuses on rebuilding reputation, as suggested by theory (Fink, 1986). On the other hand, the dominant frame used by the online public suggests that it starts to evaluate the company’s crisis management at this point of the crisis.
Furthermore, news media again enriched communication by contributing more technical frames. These findings indicate that the actors applied frames according to their respective goals and interests.
Interestingly, analysis on the pair level shows, that the similarity of implicit frames between the online public and newspapers increases in the third phase of the crisis compared to the second. This finding suggests support for previous research arguing for the proximity of these two actors (Etter & Verstergaard, 2015; Neuman et al., 2014). However, since the
mutually used frame was only introduced in the resolution stage, as no actor used it during a previous stage, it is not possible to conclude whether the online public adapted the news media’s frame or vice versa.
In respond to the presented research problem, the current research delivers several theoretical and practical implications. Over the course of the crisis, frames used by the three actors increasingly diverge form each other. Especially the expected similarities in frames used by the organization, the online public, and news media did not occur over the course of the crisis. On one hand, this raises the question whether collective sensemaking did not occur at all or whether frame alignment is an inadequate measurement to grasp the expression of collective sensemaking and consistent understanding of the crisis among actors. This understanding challenges suggestions by other scholars (Seeger, 2002; Van der Meer et al., 2014). On the other hand, one could ask whether or not a crisis can ever be drawn to a close in regard to the new possibilities social media offers for framing. Permanent access to social network sites allows for constant dissimilation of information, ideas, and opinions. Through social media the public becomes an actual part of the crisis communication response (Veil, Buehner, Palenchar, 2011). As for the constant stream of new information about corporate crises, it is questionable if a crisis can ever be resolved. This research shows that especially the online public communicates intensively about the analysed crisis. The qualitative
interpretation of frames reveals that the online public contributes its own distinct frames to the discourse. However, it is not possible to conclude if the public’s empowerment enabled by social media has positive or negative consequences for the organization in crisis. Derived from presented results, several practical implications can be drawn. First, faced with the fast moving social media activities of the online public, it is crucial for organizations in crises to engage in immediate crisis communication. In doing so, they contribute to the public’s sensemaking activities and benefit from the fact that other actors have not yet started to participate. Second, newspapers have repeatedly shown to be an important actor in extending
the discourse and providing alternative information. Thus, it is important for organizations to provide broad background information tailored to newspapers’ needs over all stages of the crisis. Third, during the last phase of the crisis, the public starts to evaluate the organization. Since public opinion is crucial for corporate reputation (Coombs, 2007), it is essential for organizations to actively communicate during this phase in order to positively contribute to the evaluation. Lastly, faced with the proximity between online public and newspapers, it is critical for organization to be active in social media and monitor public social media activities in order to foresee possible communication trends and to act preventively.
Discussion
By studying frame development among different actors and crisis stages, this research shows that the online public’s new power in terms of the collaboration of crisis frames has neither an explicit negative nor positive influence on organizations. Furthermore, the findings challenge existing literature on crisis framing by different actors, which predominantly found a phase of similar frames among different actors after initially diverse frames in the beginning of a crisis situation. However, with the presented findings, this research relates to a recently published study, which analysed crisis frames of AirAsia and the online public in their communication material about AirAsia’s plane crash in 2014. The authors did not find frame alignment between the actors at any point of the crisis (Gerken, Van der Land, & Van der Meer, 2016). Together, these studies underline the new dynamic in crisis framing across different actors following online public’s empowerment.
Theoretically, this study extends the field by empirically analysing communication dynamics of the organization in crisis, news media and the online public. By means of a semantic-network approach, the study provides further insight into the dynamic and complex process of meaning construction and framing over time, and thereby answers the call for a more complex approach regarding the analysis of the interactive and inter-organizational negotiation of reality between different actors in times of crisis (Schultz & Raupp, 2010).
Furthermore, this study overcomes the predominant organization-centrism in public relations and crisis research (Schultz et al., 2011) by including the online public as vital actor
influencing the crisis’ development. Applying the concept of frames shows how the public represented in social media frames a crisis differently than other established actors, whereby they contribute to the discursive negotiation of a corporate crisis with distinct frames used in their communication. The findings illustrate the public’s prominence in crisis communication and particularly framing, supporting a more complex and non-hierarchical approach (Castells, 2007) to the frame-building processes among the actors organizational public relations, news media, and the public. Methodically, applying semantic-network analysis to analyse frames based on word/document co-occurrence, this study contributes to a less prominent field of empirical methods. Moreover, validating a newly developed analytical technique (Van der Meer et al., 2014), allowed to quantitatively assess the development of implicit frames across different actors and crisis stages.
This research faces certain limitations, which must be considered when interpreting the results. First, as crises are by definition unpredictable and every crisis can be seen as an anomalous and unique event, generalizability is restricted (Easton, 2010) and drawing implications for future crisis situations is problematic. However, since crises occur with increasing frequency and scale, it is possible to observe similarities and patterns (Sellnow & Seeger, 2013). Never the less, every case study enriches the body of literature by providing new perspectives on this field. Moreover, casual explanations drawn from one case may serve as foundation to develop theory beyond this single case (Easton, 2010). Second, the
automated analysis is based on the assumption that word networks are sufficient to capture the general meaning of the text. This comes with the risk to obscure relevant information.
However, disregarding the syntax allows to analyse large amounts of communication material (Grimmer & Stewart, 2013). Third, as mentioned previously, applied social media data collection method comes with some drawbacks. Identifying relevant data using search API
delivers results most relevant to the user’s profile (Developers, n.d.) and thus, tends to deliver a biased sample. However, in regard of this paper’s resources and a lack of alternatives this limitation has been accepted.
Future research should further investigate each actor’s power in influencing public opinion, in particular extend knowledge about the public’s role in meaning negotiation. Especially the disclosed pattern of public opinion unfolding as a process over time with involved actors starting to communicatively contribute at different points in time deserves closer investigation. In addition, research should focus on collective sensemaking as a need to overcome the crisis and its imperative. Is the absence of collective understanding, i.e. frames problematic with respect to the crisis development or is it possible that different discourses co-exist and perhaps even complement each other? Characteristically, this kind of research approaches the public as well as the newspapers as one actor. In contrast to describe the organization as one corporate actor, treating the public and newspapers as single entities leads to a loss of information. Thus, future research should assess different groups of the online public regarding their similarities and differences in framing a certain issue in order to gain valuable insights into the structure and characteristics of this newly empowered group of individual actors.
Notes
1. As described before, frames are constructed through specific use of language (Entman, 1993) based on the choice of particular words (Leydesdorff & Hellsten, 2006). The network in which these words are presented provides words with their specific meaning (Hellsten et al., 2010). This understanding is based on the analytical approach that communication materials can contain both implicit as well as explicit frames. Explicit frames arise form specific vocabulary used by the author to
emphasise a certain perspective when describing a topic (Hellsten et al., 2010). These explicit frames are best identified with help of traditional content analysis, where texts are coded according to predefined categories (Krippendorf, 1980). However, the majority of frames used in communication materials are not explicit but implicit (Hellsten et al., 2010). Implicit frames can be recognized in “latent dimensions of communication, and they are generated because of spurious correlations between word (co-)occurrences in communications” (emphasis in original) (Hellsten et al. 2010, p. 593). Thus, regarding implicit frames, the meaning of words is created in the semantic context in which they are used (Hellsten et al. 2010). Consequently, unlike explicit frames, implicit frames are not directly observable and ask for a computer-assisted method of automated content analysis.
2. Sensemaking can be described as the interpretation, categorization, labelling, and explanation of certain issues by individuals within an organization (Cornelissen et al. 2009; Weick, Sutcliffe & Obstfeld, 2005). The sensemaking process is influences by the individual’s professional identity, including such as beliefs, values, motives and experiences. Thus actors’ professional identity tends to guide actions and
communication since each actor refers to different beliefs and standards when assessing the significance of an event or issue (Cornelissen et al., 2009).
3. Uncertainty during crises has different sources, such as unknown nature and cause of the crisis, unknown involvement of stakeholders as well as unknown scope and duration of crisis impacts (Lu, 2013).
4. The website allmytweets.net allows to collect all recently published tweets of one specific account.
References
Benford, R. D., & Snow, D. A. (2000). Framing processes and social movements: An overview and assessment. Annual review of sociology, 26, 611-639. doi: 10.1146/annurev.soc.26.1.611
Blood, R. (2003). Weblogs and journalism: Do they connect. Nieman reports, 57(3), 61-63. Retrieved from http://www.rebeccablood.net/clips/03fall_nieman.pdf
Borah, P. (2011). Conceptual issues in framing theory: A systematic examination of a decade's literature. Journal of communication, 61(2), 246-263. doi: 10.1111/j.1460-2466.2011.01539.x
Bordia, P., & DiFonzo, N. (2004). Problem solving in social interactions on the Internet: Rumor as social cognition. Social Psychology Quarterly, 67(1), 33-49. doi: 10.1177/019027250406700105
Castells, M. (2007). Communication, power and counter-power in the network society.
International journal of communication, 1, 238-266. doi: 1932-8036/20070238
Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science,
10(1), 103–126. doi: 10.1146/annurev.polisci.10.072805.103054
Coombs, W. T. (2002). Assessing online issue threats: Issue contagions and their effect on issue prioritisation. Journal of Public Affairs, 2(4), 215-229. doi: 10.1002/pa.115 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. doi: 10.1057/palgrave.crr.1550049
Coombs, W.T. (2015). Ongoing crisis communication: Planning, managing, and responding. Los Angeles: Sage.
Coombs, W. T., & Holladay, S. J. (2008). Comparing apology to equivalent crisis-response strategies: Clarifying apology’s role and value in crisis communication. Public
Cornelissen, J. P., Carroll, C. E., & Elving, W. J. L. (2009). Making Sense of a Crucial Interface: Corporate Communication and the News Media. In L. Chouliaraki & M. Morsing (Eds.), Media, Organizations and Identity (pp. 129-145). New York: Palgrave Macmillan. doi: 10.1057/9780230248397_7
Developers (n.d.). Twitter developer documentation. The search API. Retrieved from https://dev.twitter.com/rest/public/search
Durham, F. D. (2001). Breaching powerful boundaries: A postmodern critique of framing. In: S. D. Reese, O. H. Grandy Jr., & A. E. Grant (Eds.), Framing public life: Perspectives, on media and our understanding of the social world (pp.123-137). Mahwah, NJ: Lawrence Erlbaum Associates.
Easton, G. (2010). Critical realism in case study research. Industrial Marketing Management, 39(1), 118-128. http://dx.doi.org/10.1016/j.indmarman.2008.06.004
Entman, R. M. (1993). Framing: Toward a Clarification of a Fractured Paradigm. Journal of Communication, 43(4), 51-58. doi: 10.1111/j.1460-2466.1993.tb01304.x
Etter, M. A., & Vestergaard, A. (2015). Facebook and the public framing of a corporate crisis. Corporate Communications: An International Journal, 20(2), 163-177. doi:
10.1108/CCIJ-10-2013-0082
Field, A. (2009). Discovering statistics using SPSS. London: Sage Publications Limited. Field, A. (2013). Exploratory Factor Analysis. In M. Verboord (Ed.), Quantitative methods in
media and communication (pp. 207–261). London: Sage Publications.
Fink, S. (1996). Crisis management: Planning for the inevitable. New York: American Management Association.
Gamson, W. A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology, 95(1), 1–37. doi: 10.1086/229213
