Preparing for a Crisis?
The Significance of Crisis Response Strategy and Source of Information in the Context of Crisis Communication via Social Media
Master’s Thesis
Graduate School of Communication, University of Amsterdam Master’s programme Communication Science
Student: Margarita R. Boncheva Student number: 12343676 Supervisor: Dr. James Slevin Date of completion: 29-01-2020
Abstract
When a crisis happens, organizations should choose the right communicative strategies to respond to the crisis in order to limit reputational damage. In the era of social media, effective crisis communication still largely depends on the choice of a crisis response message. Yet, research has not agreed on how this message should look like. Source of information is another key factor for social media crisis communication. With the large amount of content on social media, organizations are often challenged to get their opinion heard. It is believed that only organizational sources that are perceived as credible will be selected among the plethora of information online and would not become buried under the prevalence of other views. Which organizational source is perceived as the most credible, though, seems
insufficiently empirically validated. Therefore, in a 2 x 3 factorial design (N = 329), this study explored how crisis response messages and different organizational sources disclosing crisis information affect stakeholders’ post-crisis evaluation of organizational reputation. And also, whether source of information moderates the effectiveness of crisis response strategies. In this way, the study answered the need for additional research regarding the effectiveness of crisis response message and source strategies and was able to collect evidence to provide
organizations with insights on how to strategically match a crisis response strategy and source of information to optimize their crisis response. According to the findings, rebuild as a
strategy results in a higher post-crisis reputation of the organization compared to a denial strategy. Also, stakeholders were found to react better to crisis response strategies coming both from the regular employee and the organization than to strategies coming from the CEO. When the regular employee was the source of information, the reputation of the organization was the most positive. Practically speaking, considering the findings, organizations should no longer rely on the “conventional wisdom” that the CEO should deliver the crisis message on social media and should aim for using accommodative strategies in their response.
Preparing for a Crisis? The Significance of Crisis Response Strategy and Source of Information in the Context of Crisis Communication via Social Media
Consider the recent Boeing 737 Max scandal regarding an automated system on their planes contributing to deadly crashes; Marriott’s massive data breach that affected the records of up to 500 million customers; The Facebook saga of failures to ensure user data privacy; Ikea’s furniture that killed multiple children and lead to millions of recalls, or the horsemeat found in their meatballs. Such examples of high-profile crises involving various organizations are in the media around the world constantly. They serve as reminders that no organization is resistant to crises and that all organizations should be prepared if a crisis befalls them.
Organizational crises present a time of ambiguity and risk for both organizations and their stakeholders. To reduce the negative impact of a crisis, organizations should respond appropriately, otherwise their reputation most likely will suffer (Ma & Zhan, 2016). Over the past few decades, scholars have paid considerable attention to find out what strategies would be most effective for crisis communication (Claeys & Opgenhaffen, 2016). Despite all the knowledge made available, organizations still often fail to prevent damage to their reputations (Coombs, 2015). This implies the need for more crisis research and the importance of
considering additional communication aspects (van Zoonen & van der Meer, 2015).
With the emergence of social media, crisis management has become a more complex field, and organizations face new challenges in performing effective crisis communication (Stephens & Malone, 2010; Freberg, 2012). Due to the more dialogic, interactive and immediate nature of social media, a vast amount of crisis information and misinformation spreads rapidly and more broadly than it used to before with traditional media (Coombs, 2015). With the large amount of content on social media, organizational opinion often becomes blurred or confused by the many-to-many communication model made available by social media (Freberg, 2012). Consequently, if organizational messages are hindered from
reaching out to the public as intended, crisis communication becomes less effective. In this regard, a major question for organizations that aim to communicate more effectively online in times of crisis appears to be: what makes an organizational source and its message standing out as more credible, and thus, as preferred among the plethora of information online?
Among others, effective social media crisis communication is said to be about using the potential for dialogue, monitoring and being prepared for a crisis, and, most importantly, about choosing the right timing, message and source to deliver it (e.g., Eriksson, 2018). In regard to timing, researchers have agreed that organizations should provide a quick response to a crisis and up-to-date information on a regular basis to remain in control over the situation (Arpan & Roskos-Ewoldsen, 2005). However, even when previous research has paid
substantial attention to assessing the effects of different crisis response message strategies on post-crisis outcomes (Claeys & Opgenhaffen, 2016), there appears to be no universal
agreement of how an effective message should look like (Eriksson, 2018; Tkalac Verčič, Verčič, & Coombs, 2019). Moreover, despite the increasing importance of sources and their credibility perception for effective crisis communication, especially in the context of online crisis communication (Liu, Austin, & Jin, 2011; Eriksson, 2018), too little is known by scholars and organizations about the possible effects of different organizational sources disclosing crisis information online on the post-crisis outcomes and the effectiveness of crisis response strategies (Arpan, 2002; van Zoonen & van der Meer, 2015; Tkalac Verčič et al., 2019; Kim & Park, 2017). Therefore, the research questions of this study will be:
RQ: What is the effect of type of crisis response strategy (denial vs. rebuild) and source of information (organization, executive employee, or a regular employee) on post-crisis evaluation of organizational reputation? And also, is the effect of type of crisis response strategy on organizational reputation different for the different sources of crisis information?
The few recent studies that have examined how the public evaluates different crisis response strategies from different sources on social media provide a somewhat incomplete and rather controversial depiction of the situation and imply the need for more research (van Zoonen & van der Meer, 2015; Tkalac Verčič et al., 2019; Kim & Park, 2017). By answering the above-stated questions, a gap in knowledge will be addressed. On the one hand, filling this gap will add to the current literature in the field of crisis communication on social media. On the other hand, evidence will be collected to provide helpful insights to organizations about how to strategically match a crisis response message and source of information, thereby optimizing their crisis response in the online environment of information wealth. Given the centrality of a reputation for organizations, finding the best combination of crisis response strategy and source to respond in a crisis situation would be of great importance for
organizations and practitioners. They would be able to prepare by making informed choices, and in this way, address crises in the context of social media more effectively than before.
Therefore, building upon existing research, aiming to further investigate the impact of crisis response messages and source strategies on organizational reputation, an online factorial experiment with 329 participants is conducted. To address and re-test the effectiveness of crisis response strategies in the context of social media, this study draws on the
well-established situational crisis communication theory (SCCT) (Coombs, 2007). Also, since how people perceive different crisis-response strategies from different organizational sources on social media is less well established (e.g., van Zoonen & van der Meer, 2015), the current literature on source credibility is taken as a framework to explain why specific sources of information might be perceived as more credible than others.
Theoretical Framework
The theoretical foundation of the current study begins by defining a crisis and addressing the importance of organizational reputation in crises. Next, the focus is on crisis communication and the new challenges organization face in the era of social media. The presumed impact of crisis response strategy and source of information on reputation is discussed afterwards. Each section ends with a hypothesis derived from the literature. The framework concludes with a proposed moderation hypothesis and a conceptual map.
Organizational Crisis and Reputation
Organizational crises, which are widespread nowadays, can affect both organizations and numerous individuals in society. Fearn-Banks (2011, 2) defines a crisis as: “a major occurrence with potentially negative outcomes affecting an organization, company, or industry, as well as publics, products, services or good name. It interrupts normal business transactions and can sometimes threaten the existence of the organization.”
As underlined in this definition, one of the potential adverse outcomes of a crisis is significant damage to the reputation of an organization. A positive reputation, though, is essential for organizational survival and success (Coombs, 2007). According to Fombrun and van Riel (2004), a reputation is considered as a valuable, intangible asset of organizations that has to do with how stakeholders perceive an organization. Among others, an organization that has a favorable reputation can attract new customers and employees, increase financial
performance, and create investment interest (Fombrun & van Riel, 2004).
Crisis Communication and Social Media
Since high levels of uncertainty characterize crises, effective organizational
communication represents the dialogue an organization has with its stakeholders before, during, and after a crisis. The key objective thereby is to provide information to stakeholders and to preserve organizational assets such as reputation and sales (Yang, Kang & Johnson, 2010; Coombs, 2015).
With the advances in communication technologies, crisis management has become a more complex field, and organizations face new challenges in performing effective crisis communication (Stephens & Malone, 2010; Freberg, 2012). Social media has made
communication transmission easier, faster, and more visible, and enabled crises to spread and develop in new ways (Coombs, 2015). Social media platforms such as Facebook and Twitter have changed the traditional communication model of ‘one-to-many’ such that nowadays, the audience can actively participate in the crisis conversation. This means that lots of
information and misinformation spread rapidly and more broadly than before. On the one hand, social media platforms afford organizations new possibilities to communicate. On the other, in this context of information abundance, organizations’ opinions often become blurred or confused by the noise of the many-to-many communication model made available by social media (Freberg, 2012). Besides, unconfirmed messages, such as rumors and mischief can become dominant on social media, exposing organizations to new reputational risks. These changes put organizations under great pressure, prompting an increasing need for guidance in how social media can be used more efficiently in crisis to reach stakeholders and to protect organizational assets.
SCCT and Crisis Response Strategies
The publics’ increasing use of social media during crises means that organizations cannot afford to neglect integrating social media into their crisis management. As Xu and Wu (2015) argue, the only choice is how to do this. In the last decades, crisis communication
scholars have examined how organizations should respond to crises, coming up with useful tactics to guide organizational response (Ma & Zhan, 2016). Besides using social media to create a dialogue with the public, providing timely and updated crisis information, being prepared, and monitoring the Internet for emerging issues, effective social media crisis communication is said to be about choosing the right message (Eriksson, 2018).
As one of the most prominent theories in crisis communication, situational crisis communication theory (SCCT) (Coombs, 2007) provides a set of guidelines for how crisis managers can use crisis response strategies to protect a reputation from the effects of a crisis. The theory suggests that crisis managers should match strategic crisis responses to the level of crisis responsibility and the reputational threat posed by a crisis. According to SCCT,
organizations in crises can choose from three groups of primary response strategies to protect or repair their reputation. The strategy should be appropriate for the situation and the potential reputational harm (Tkalac Verčič et al., 2019). An organization that uses denial strategies indicates that it is not responsible for the event, framing the message in a way that removes any connections between the organization and the crisis. An organization that uses diminish strategies tries to downplay the crisis itself and the organization’s part in it. The third, more accommodative strategy, is the so-called rebuild strategy in which an organization attempts to improve its reputation by apologizing and offering material and/or symbolic forms of aid (Coombs, 2007; Ma & Zhan, 2016). While defensive strategies are recommended when organizational crisis responsibility is weak, accommodative strategies are advocated when crisis responsibility is stronger (Coombs, 2007; Claeys & Cauberghe, 2014).
It needs to be emphasized that traditional crisis communication research, including SCCT, has often been criticized for taking a firm-oriented perspective, considering the public to be passive receivers of a firm’s crisis communication (Zheng, Liu, & Davison, 2018). However, as noted, the advent of social media has changed the public’s role from passive
receivers to active content creators and opinion generators. While some of the best practices used in traditional media are expected to remain effective in the context of social media, other may require adaptation (Freberg, 2012). Still, it seems that a vast majority of scholars
recommend using traditional crisis communication strategies, such as providing an apology as a tool for effective crisis communication online (Eriksson, 2018). Numerous authors have suggested that defensive crisis response strategies (i.e., denial and diminish strategies) are often less effective than a more accommodative strategy, such as offering an apology (i.e., rebuild strategies) (Kim, Avery, & Lariscy, 2009; Coombs & Holladay, 2008; van Zoonen & van der Meer, 2015, Dardis & Haigh, 2009; van der Meer & Verhoeven, 2014), and that blaming others is not effective response message strategy online (Eriksson, 2018). Some researchers have even argued that accommodative strategies should be chosen by
organizations in crisis regardless of crisis responsibility (Lee, 2005; Kiambi & Shafer, 2016). Nonetheless, other studies argue that defensive messages are most often used by organizations (Kim et al., 2009) since they present a useful way to reduce anger, criticism, and the spread of online rumors (Roh, 2017). To re-test and confirm the effect of crisis response strategies in the context of crisis communication via social media, the following hypothesis was formed:
H1: The reputation of an organization after using a rebuild crisis response strategy will be more positive than the reputation of an organization using a denial crisis response strategy.
Source of Information and Credibility Perception: Traditional vs. Social Media
According to Hovland and Weiss (1951), the source (i.e., a person, organization, or the media) is a multidimensional predictor factor of communication effectiveness besides the content and the medium of a message. Numerous studies have revealed that the more credible people perceive information, the stronger the effects are on attitudes and behavior (Miller &
Krosnick, 2000; Nan, 2009). The degree of credibility people attribute to a piece of
information is said to largely depend on, among others, the source of information (Hovland, Janis, & Kelley, 1953). Source credibility is defined as “judgments made by a perceiver (…) concerning the believability of a communicator (O’Keefe, 2002, p. 181).”
In traditional news formats, such as newspapers or television news, there is usually a restricted number of sources and high barriers for access to the public broadcasting of information (Metzger & Flanagin, 2013). The information that is chosen for publication is done so through a process known as gate-keeping (Shoemaker & Vos, 2009). Gatekeeping is a filtering process by which content creators, such as journalists, decide what information should be presented to the public (Westerman, Spence, & van der Heide, 2012). Beforehand, gatekeepers are assumed to be checking information to ensure its credibility. In this
environment of information scarcity, credibility is granted to sources believed in providing reliable information such as the government or an organization, business leaders and officials on formal positions (Westerman et al., 2012, Miller & Kurpius, 2010).
Social media presents a very different environment from that of traditional media, namely one of information abundance. The amplified possibilities for the general public to participate in the news creation has changed the traditional process of gatekeeping and has led to various questions regarding the credibility of news content and online information (Bakker, Trilling, de Vreese, Helfer, & Schönbach, 2013). Since social media has become a heavily used source of information through which the public receives crisis-related information (van Zoonen & van der Meer, 2015), how people make judgments about the credibility of
organizational sources and the messages released by these sources online becomes a significant question.
In times of crisis, crisis communication, including the one coming from organizations and employees online, is not subjected to media gatekeeping processes. Consequently,
stakeholders have to determine the relevance, newsworthiness, and credibility of the information they receive themselves (Utz, Schultz, & Glocka, 2013). In this regard, researchers have argued that an organizational source that is perceived as credible can
influence stakeholders’ crisis perceptions and their probability of accepting the organizational messages as primary sources over the number of others (Coombs, 2015). Also, it has been suggested that when publics receive a crisis response from a communicator with greater credibility in a crisis, they are more likely to think it is acceptable and truthful and, in turn, have a high-level intent of supportive behavior (Kim & Park, 2017; Arpan & Roskos-Ewoldsen, 2005).
Organization vs. Executive Employee
In a crisis, organizations need to address all their main stakeholders (Coombs & Holladay, 2008; Verhoeven, van Hoof, ter Keurs., & van Vuuren, 2012). Online, this can be done through various channels and sources. For instance, a crisis message can be sent out from the organizational account on social media formulated on behalf of the organization. In this case, the source of the message is not a person but rather a complex institutional structure with a history of practice and information (Metzger et al., 2003).
In this regard, Snoeijers, et al. (2014) argue that having a human face as a source will have a greater effect on the effectiveness of social media crisis communication than using an anonymous organizational social media account. People can relate easier to spokespersons than to the organization, as people often feel more empathy for a spokesperson rather than a cold, impersonal entity (Carroll, 2016). Hence, a written or a video message can be sent out from the private or the organizational account on behalf of an executive employee, for instance (i.e., here understood as an employee belonging to the upper management, such as a CEO).
Seeger and Ulmer (2001) argue that if an organization is facing a severe crisis, CEOs are often chosen because of their authority and credibility. It is assumed, namely, that the CEO’s leadership position can provide virtue to the message they communicate (Seeger & Ulmer, 2001) and offer the audience better impressions of crisis recovery (Lee, Kim, & Wertz, 2014). Also, Jaques (2012) claims that the CEO’s involvement demonstrates the organizational commitment to the crisis, and the title should be a cue of expertise. Jaques (2012) further notes that it is a “conventional wisdom” within much of the crisis literature that the CEO should be the communicator during a crisis. To test and provide empirical evidence-based support to this notion, Tkalac Verčič et al. (2019) compared the effect of a CEO and a public relations practitioner disclosing crisis information on the post-crisis reputation of an organization and found no significant difference between the sources, concluding that the selection of source has little effect on how people perceive crisis strategies. To empirically test the added value of a CEO as a source of information compared to the organization itself disclosing crisis information, in this study, the following hypothesis was formed:
H2a: The reputation of an organization where an executive employee was the source of crisis information will be more positive than the reputation of an organization where the
organization itself was the source of crisis information.
Regular Employee as the most credible Source
Van Zoonen and van der Meer (2015) compared a sales employee to an organization as a source of crisis information, arguing that when communicating through his private account, the sales employee was perceived as more credible and trustworthy than the organization. This might be true, they claim, because employees can utilize their private social media channels to inform others, communicating not as institutional entities, but as
individuals. As such, employees may be seen as independent sources of information from official organizational sources.
Moreover, as noted, it is quite common practice for organizations to choose a highly-ranked person, such as a CEO, a president or a managing director to disclose information in the media during a crisis. However, according to the Edelman Trust Barometer (2013), regular employees, especially those with technical expertise, are seen as more credible than chief executives. This would suggest that although a highly-ranked employee may add virtue to the message, demonstrating that the situation has the highest priority for the organization, such a spokesperson might not always be the most appropriate choice (Carroll, 2016).
In the present study, it is argued that while regular employees may indeed be seen as unrestricted in their online communication by the organization, and as not being forced by the organization to say specific things on their private accounts; Executive employees, because of their higher rank position, may still be seen as restricted when utilizing social media. Even when using their private account, executive employees, traditionally recognized as credible sources, may still be perceived as acting in favor of and representing the organization, and thus, as less credible than regular employees. In contrast, despite their relative lack of authority, regular employees (i.e., here understood as all employees bellow the upper management, including middle and lower management) may be seen as having relevant expertise due to their firsthand knowledge or know-how with topic or situation, and therefore might be perceived by others as possessing a great deal of experiential credibility (Flanagin & Metzger, 2008, 2013). Therefore, the following hypothesis was formed:
H2b: The reputation of an organization where a regular employee was the source of crisis information will be more positive than the reputation of an organization where an executive employee or the organization was the source of crisis information.
To test if the effectiveness of crisis-response strategies will be different for the different sources of crisis information, or in other words, to asses the potential combined effects of crisis communication strategy and source of information, the following hypotheses were formed:
H3: The positive effect of a rebuild strategy on organizational reputation, compared to a denial strategy, will be more prominent if: a) an executive employee is the crisis information source, rather than if it is the organization itself; b) a regular employee is the crisis
information source, rather than if it is an executive employee or the organization itself.
Figure 1. Conceptual Map
Methods Participants
The participants for the experiment were recruited conveniently both from a pool of acquaintances of the researcher, as well as from the online platform Survey circle (i.e., a research platform based on the principle of mutual support). The only requirement to take part in the experiment was for participants to be at least 18 years old. The number of participants
H2 a, b
H1 H3 a, b
Source of crisis information
who fully completed the questionnaire was N = 331. One participant, who indicated to be below the age of 18, and another one, who appeared to be an outlier when checking the assumption for the statistical analyses, were excluded from the final data. Thus, the analyses were conducted with a total of 329participants. Further, four participants did not indicate their country of origin, while one gave an impossible value for age. Consequently, this resulted in five missing values. The eventual sample consisted of about 58% females. All respondents were between 18 and 73 years old (M = 29.23, SD = 9.23). In total, participants from 49 countries took part in the experiment, the majority being from continental Europe. Regarding their level of education, above 75% of participants indicated holding some form of an academic degree. All in all, 90% of all participants indicated having a Facebook profile.
Research Design
The aim of the experiment was to measure how the public evaluates different crisis response strategies from different sources on social media. To approach the question, crisis response strategy was held as an independent variable, source of information as a moderator variable, while organizational reputation as a dependent variable. So, a 2 (Crisis response strategy: denial vs. rebuild) × 3 (Source of information: organization vs. executive employee vs. regular employee) between-subject factorial experimental design was conducted. This resulted in six experimental conditions all in all. Each participant was randomly allocated to one of the conditions and exposed to either a denial or a rebuild message combined with a source of information, either being the organization, an executive employee, or a regular employee.
Procedure
Upon the start of the experiment, participants either got an invitation containing a link leading to an online questionnaire via social media or they could follow the link on Survey circle. The introductory page included some general information about the experiment and some brief instructions. After declaring consent, participants were allowed to commence with the study. In the first block, a news release by a well-known news agency was shown to participants in which the crisis and the organization were depicted. The article started with “London (Reuters).” Thus, participants could identify the source of the text as a neutral press agency. To eliminate participants’ prior knowledge and perception of an organization, a fictitious baby formula producer called Babylinno was created. Participants were asked to read the news release, which was 239 words long, and reported that the consumer product safety commission declared that testing results indicate that a certain batch of Babylinno milk-based powdered infant formula was found to be contaminated with Salmonella. The article further reported that no cases of illness associated with the consumption of this product had been reported so far. Afterward, depending on which one of the six experimental
conditions they were assigned to, participants read a message from the Facebook account of one of the three sources (the organization, an executive employee, or a regular employee) talking about the crisis. The six mock-up messages thus differed in the crisis response strategy applied by the source (denial or rebuild strategy) and in which one of the mentioned above three sources talks about the crisis. After exposure to the manipulation material, participants were requested to fill out a questionnaire to measure perceived source credibility and
organizational reputation. A manipulation check followed just after. Next, participants were asked some questions concerning demographics (e.g., gender, age) and were debriefed at the end of the questionnaire. It was explained that the situation was fictional and merely created for the sake of the study.
Experimental Materials
To manipulate crisis response strategy, two response strategies were selected, namely denial and rebuild as examples of a defensive and a more accommodative crisis response strategy, respectively. In the case of a denial strategy used, the message released by the different sources sharing information about the accident contained information attempting to remove any connection between the organization and the crisis. Scapegoating, as a type of a denial strategy, was used by the organization or each communicator to blame a third-party outside of the organization for the crisis (Coombs, 2007). An example of the denial strategy message is: “We are truly saddened that such an accident has occurred, but unfortunately, our company could not foresee and prevent it. To our current knowledge, the bacteria was
introduced during the manufacturing process at a factory of our contractor company Nutribo, which produces some of our formulas. In this case, it is clear that the provided by Babylinno safety regulations were not followed, and we cannot be held responsible!”
Further, rebuild strategies include apology (i.e., accepting full responsibility and asking for forgiveness) and compensation (i.e., offering compensation to victims) (Coombs, 2007). Thus, in the case of a rebuild strategy used, the message about the accident released by the different sources contained the following information: “We are truly saddened that such an accident has occurred and want to offer our sincere apologies to all parties affected. To our current knowledge, the bacteria was introduced during the manufacturing process at one of our factories. Urgent measures are being taken to prevent such accidents of happening in future and anyone who has been affected will be compensated.” The content of the messages was kept similar between the sources to ensure that the general message would be the same.
To manipulate source of information, participants were either exposed to the verified Babylinno Facebook timeline, the timeline of an executive employee of Babylinno (i.e., a CEO), or a regular employee (i.e., a Production Supervisor). Personal information about the
employees was added to their accounts, to make a clear distinction between the three
Facebook accounts. The biography of the fictitious executive employee Oliver Jones specified that he is a Chief Executive Officer of Babylinno and a father of three from London.
Additionally, the profile picture of the executive employee’s account showed a man in his fifties, and on the left side of the account, several personal pictures of the employee were shown. The background of the timeline was black, provided by default by Facebook. The Facebook account of the lower-level regular employee was manipulated similarly. The biography of the fictional employee Jack Davies specified that he is a Production Supervisor at Babylinno and a father of three from Watford. The additional manipulations were exactly the same as by the executive employee. In contrast, the organizational Facebook account had a more professional look, compared to the employees’ accounts. The biography of the
Babylinno account stated that this is the official account of the company for the UK and contained some contact information. A logo showed that this is the verified account of the organization, and the background showed a picture of an infant with its mother. An example of the three Facebook accounts is shown in Appendix A.
Measures
Much work has attempted to distinguish the distinct dimensions of perceived source credibility (Hellmueller & Trilling, 2012). Van Zoonen and van der Meer (2015) use a five-item scale by Metzger, Flanagin, and Zwarum (2003) that identifies the five-items believability, accuracy, trustworthiness, bias, and completeness of information as consistent dimensions of source credibility. Since Hellmueler and Trilling (2012) identify competence of a source as another important dimension of source credibility, this study uses the five items discussed above and adds a sixth one, namely competence of a source to measure source credibility. The items were measured on a 7-point Likert scale ranging from 1 (strongly disagree) to 7
component factor analysis indicated that the scale was unidimensional (one component with Eigenvalue above 1.00), explaining 58% of the variance in the six original items. The reliability of the scale was satisfactory but increased after one item was deleted (i.e., bias) (Cronbach’s α = 88). All items were combined into a mean variable (M = 4.06, SD = 1.20).
Post-crisis organizational reputation was measured using an adopted and validated organizational reputation scale by Coombs and Holladay (2002, 2008) that consists of five items. The items were measured on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). An example item is: “The organization is concerned with the well-being of its publics.” The reliability of the 5-item scale was satisfactory after two negatively framed items were reverse coded beforehand (Cronbach’s α = 89). All items were combined into a single mean variable (M = 4.33, SD = 1.28).
Results Randomization Check
A Chi-square test of independence was conducted to check if participants’ gender was comparable over all six experimental conditions. The Chi-square test had the independent variable experimental conditions from the data set placed in columns, and the dependent variable gender in rows. No significant relationship was found between the two variables, χ2
(10, 329) = 8.54, p = .576.
To check if participants’ age was comparable across the six experimental conditions, a Two-way ANOVA was conducted. This ANOVA had experimental conditions as an
independent variable and age as a dependent variable. The ANOVA showed that participants’ mean age was not significantly different across the conditions, F (5, 322) = 1.27, p = .278. Following, one could assume that the randomization of participants across conditions was successful in terms of participants’ gender and age.
Manipulation Check
A manipulation check verifying whether participants experienced their condition of the experimental factor as manipulated was conducted toward the end of the experiment to avoid any undesired effect of the order of items to the acquired results. Operationalization of the manipulation check for the variable crisis response strategy included the item: “To what extent do you believe that the organization/communicator took responsibility for the crisis?” A scale ranging from 1 (not at all) to 7 (very much) was provided to participants to answer the question. To check if participants’ responses differed in the two groups, an Independent-samples T-test was conducted with crisis response strategy as a factor and manipulation1 as a dependent variable, t (327) = -15.41, p < .001, 95% CI [-1.97, -1.52], d = 1.70. One can assume that the participants in both groups have perceived the factor as manipulated since participants in the rebuild condition scored significantly higher on the scale (M = 3.57, SD = 1.05) compared to participants in the denial condition (M = 1.82, SD = 1.00).
The manipulation/exposure check of the variable source of information consisted of the item: “Who was the source of crisis information on Facebook? Three different options were provided below: The organization, an executive employee, or a regular employee/lower-level management. A Chi-square test that had source of information in columns and
manipulation2 in rows confirmed that the manipulation of the source of information was successful and that the majority of participants correctly identified the right source, χ 2(4, 329)
= 178.47, p < .001, tau = .27.
Test of Hypotheses
Hypotheses 1 to 3 were tested using a two-way ANCOVA, which had crisis response strategy (denial or rebuild) and source of information (the organization, an executive
a dependent variable while controlling for source credibility perception (See Table B1 in Appendix B).
Hypothesis 1 suggested that the reputation of an organization after using a rebuild crisis response strategy will be more positive than the reputation of an organization using a denial crisis response strategy. The analysis of covariance revealed a significant main effect of crisis response strategy on organizational reputation, while controlling for source
credibility perception, F (1, 322) = 39.11, p < .001. This effect size was weak, η2= .01 Thus, 1.0 % of the variance in the outcome variable reputation could be explained by the crisis response strategy. Confirming the assumptions of hypothesis 1, when a rebuild crisis response strategy was used by the source disclosing information, the organizational reputation was significantly higher-rated (M = 4.91, SD = 1.05) than in the case of using a denial strategy (M = 3.74, SD = 1.22).
To test the main effect of source of information on the post-crisis evaluation of organizational reputation, hypothesis 2 was formulated. Hypothesis 2a predicted that the reputation of an organization where an executive employee was the source of crisis information will be more positive than the reputation of an organization where the
organization itself was the source of crisis information. Further, hypothesis 2b suggested that the reputation of an organization where a regular employee was the source of crisis
information will be more positive than the reputation of an organization where an executive employee or the organization was the source of crisis information. The analysis of covariance revealed a significant main effect of source of information on post-crisis evaluation of
organizational reputation while controlling for source credibility perception, F (2, 322) = 4.33, p = .014. The effect size was very weak, η2 = .001. Thus, less than 1.0 % of the variance
in the outcome variable reputation could be explained by source of information. Contrary to the assumptions of hypothesis 2a, a Post-hoc Bonferroni test showed that when the executive
employee was the source of crisis information, the reputation of an organization was significantly less positive (M = 4.11, SD = 1.32) compared to when the organization was disclosing crisis information (M = 4.28, SD = 1.30, p = .042) (See Table B2 in Appendix B for all group means and SDs). Further, partially supporting hypothesis 2b, the test showed that when the regular employee was the source of crisis information (M = 4.59, SD = 1.17), the reputation of the organization was more positive than when the CEO was the source of crisis information (p = .028). In contrast to what was also argued in 2b, no significant difference was found between the regular employee condition and the organization (p = 1.00). Thus, Hypothesis 2a was not supported, while hypothesis 2b was partially supported.
Hypothesis 3a proposed that the positive effect of a rebuild strategy on corporate reputation, compared to a denial strategy, will be more prominent if an executive employee is the crisis information source, rather than if it is the organization itself. Also, hypothesis 3b proposed that the positive effect of a rebuild strategy on corporate reputation, compared to a denial strategy, will be more prominent if a regular employee is the crisis information source, rather than if it is an executive employee or the organization itself. Contrary to the
assumptions of both hypotheses 3a and b, the analysis of covariance did not reveal a significant two-way interaction effect between crisis response strategy and source of information on organizational reputation in any condition, F (2, 322) = .20, p = .822 (See Figure B1 in Appendix B). Hypotheses 3 a and b were rejected.
When it comes to the assumptions of the conducted analysis of covariance, even when most of the other assumptions were more or less met (i.e., equal group size, no outliers and normally distributed data, linear relationship), it should be noted that the Levene’s F-test appeared to be significant, F (5, 323) = 4.10, p = .001. This means that the assumption of equal variances in the population has been violated. Also, the assumption of the homogeneity of regression slopes was half-met. Therefore, the results should be interpreted with caution.
Conclusion and Discussion
In the event of a crisis, organizations need to choose the right communication strategies to respond in order to limit reputational damage (Tkalac Verčič et al., 2019). The choice of a crisis response message, recognized by traditional crisis research as important (e.g., Coombs, 2007), remains one crucial aspect of effective crisis communication in the context of social media (e.g., Eriksson, 2018). Another key factor for effective social media crisis communication is source of information (van Zoonen & van der Meer, 2015; Kim & Park, 2017). With the large amount of content on social media, organizations are often challenged to get their opinion heard (Freberg, 2012). It is believed that only organizational sources that are perceived as credible will be selected among the plethora of information online and would not become buried under the prevalence of other views.
Previous findings on the use and effectiveness of crisis response message strategies appear to be contradictory, creating the need for additional research (e.g., Tkalac Verčič et al., 2019). Moreover, the impact of different organizational sources on crisis-related outcomes in the context of social media seem to be insufficiently empirically validated (e.g., Kim & Park, 2017). Thus, this study asked the following questions: “What is the effect of type of crisis response strategy (denial vs. rebuild) and source of information (organization, executive employee, or a regular employee) on post-crisis evaluation of organizational reputation?” Moreover, “is the effect of type of crisis response strategy on organizational reputation different for the different sources of crisis information?” Using an experimental design, the present study intended to add to current knowledge by addressing the limitations in the existing literature and thus provide organizations with guidance on whether and how to strategically match crisis response strategy and source of information to optimize their social media crisis communication.
Partially answering the study’s first question, the findings of this study confirm that the choice of crisis response strategy largely influences organizational reputation. The results found a significant difference between a rebuild strategy and a denial strategy. In line with what hypothesis one predicted, the present investigation revealed that corporate reputation is positively affected by a crisis response in which an organization shows compassion, concern or regret about the situation, and offers compensation to the parties concerned. These findings are in line with other researchers that have previously confirmed that accommodative
strategies such as providing an apology are often more efficient than the more defensive denial or diminish response strategies (e.g., Coombs & Holladay, 2008; Eriksson, 2017; Kim et al., 2009; van Zoonen & van der Meer, 2015; van der Meer & Verhoeven, 2014; Lee, 2005; Kiambi & Shafer, 2016). The findings are in contrast with Kim and Park (2017), who found no difference between defensive and accommodative strategies regarding reputation. Also, Kim et al. (2009) have argued that denial messages are most often used by organizations, which in light of the current findings does not appear to be the most effective strategy.
Adding to the answer to the first research question, the current findings show that how people ascribe blame during a crisis does not solely depend on the crisis response strategy, but it also depends on the source and perceptions of source credibility. These outcomes contradict Tkalac Verčič et al. (2019), who concluded that the selection of source has little effect on how people perceive crisis strategies. Somewhat surprisingly, though, refuting hypothesis 2a and partially supporting 2b, both when the organization and the regular employee were the sources of crisis information, the organizational reputation was
significantly more positive compared to when the CEO was the source of crisis information. Also, results showed that the reputation of the organization was more positive when the regular employee was the source of crisis communication compared to the organization.
Unfortunately, even when the results were in the hypothesised direction, the difference between the sources was not meaningful.
When it comes to social media, the empirical results of this study counter the “conventional wisdom” that the CEO should be the source disclosing crisis information during a crisis (Jaques, 2012). Moreover, the results are in contrast with what has been argued by Kim and Park (2017) and Seeger and Ulmer (2001), namely that the CEO is the most trustworthy and credible organizational source in delivering a crisis message since its leadership position can provide virtue to the message communicated.
Instead, in line with the Edelman Trust Barometer (2013), in the study, the regular employee was found to be perceived as more credible and to lead to a more positive
reputation compared to the CEO. One possible explanation of the finding might be such that the regular employee was seen as an independent source from official organizational sources (e.g., van Zoonen & van der Meer, 2015). In contrast, because of his higher rank position, the executive employee was possibly perceived as restricted when using his private social media account to disclose crisis-related information, or as acting in favor of the organization. Another plausible explanation could be that the regular employee was perceived as having relevant expertise due to his first-hand knowledge (Flanagin & Metzger, 2008, 2013). Stakeholders might have found a Production Supervisor as better qualified to address the topic of food contamination than a CEO.
Furthermore, early research on credibility has suggested that people often trust others similar to themselves more than organizations and consider sources affiliated with the organization as hiding the truth and as not being credible (Callison, 2001; Callison &
Zillmann, 2002). On the one hand, this suggestion might explain why the regular employee as a source (i.e., if seen as more similar to self) appeared to lead to a more positive reputation compared to the CEO (i.e., as source affiliated to the organization). However, this suggestion
provides no explanation of why results show no significant difference between the
organization and the regular employee in terms of reputation. In this regard, people either did not see the organization as dissimilar to themselves online, or the non-significant difference between the organization as a source and the regular employee could be attributed to other factors, such as the design of the study or possibly to the sample size.
In contrast to the first two significant findings, the third one, which was about the interaction effect between crisis response strategy and source of information on reputation, was not significant. Thus, answering the second research question, the type of source was not found to moderate the effectiveness of crisis response strategies. In contrast with what was predicted in the third hypothesis, the results suggest that the importance of applying a different crisis response strategy does not depend on whether the organization, an executive employee, or a regular employee will disclose crisis information, and vice versa.
Implications
Though not all hypotheses were supported, the results of this study have both scientific and practical implications. On the one hand, the study satisfied the need for more academic research concerning the effects of crisis response message and source strategies on post-crisis outcomes (e.g., Coombs & Holladay, 2008; Eriksson, 2018). On the other hand, it provided implications for practice. The choice of both a crisis response message and the source to deliver this message was found to be crucial for crisis communication on social media. These results shed light on the significance of strategically matching the crisis
response message strategy with the source of information in order to improve crisis response. Specifically, when it comes to crisis response strategies, traditional crisis communication strategies were found to remain an effective tool for crisis communication online. Using rebuild as a strategy on social media was found to lead to a higher post-crisis reputation of organizations than using a denial. Besides, results showed a significant source effect on
post-crisis reputation. Refuting the “conventional wisdom” that the CEO is the best option for a crisis spokesperson, the findings showed that when the regular employee was the source of information on social media, the reputation of the organization was the most positive.
Considering the results, a regular employee, such as a lower, or a middle manager might be perceived as more credible and trustworthy on social media, as someone that does not communicate in favor of the organization, as someone similar to self, and depending on the situation, as someone that might have more first-hand knowledge or expertise than executives. Practically speaking, in an event of a crisis, the best choice for organizations would be to pair a regular employee with a carefully crafted accommodative message in which the organization shows compassion, concern or regret about the situation, and offers compensation to the parties concerned. In this way, organizations should be able to effectively reduce reputational damage caused by the crisis.
Limitations and Future Research
This study has some limitations that could be addressed in future research. Firstly, respondents in this study were presented with one type of organization, which was a fictitious, eliminating participants’ prior knowledge and perception of the organization. Consequently, respondents might have had difficulties in rating the reputation of a fictitious organization. After all, a reputation is a multidimensional construct, which develops over time (Fombrun & van Riel, 2004). Real companies have established reputations based on previous experiences with the company. On the one hand, not being familiar with the organizations did not allow for participants to bring any undesired associations with the organization, controlling for their attitudes and emotions towards the organization. On the other hand, using a fictitious
company had its weaknesses, such as lowering the external validity of results. Future research could consider other types of organizations and organizations with established reputations.
Secondly, a non-probability, convenient sample, was selected for this research because of time and budget constraints. One should bear in mind that a convenient sampling poses some limitations concerning the generalization of results to the broader population. Nevertheless, convenience samples are common in the field of organizational studies (Bryman, 2016). Usually, the response rate by convenient sampling is fairly high, allowing for studies to offer a solid foundation for future research in the field.
Next, this study employed just one crisis scenario (i.e., baby formula contamination). A rebuild strategy and a regular employee as a spokesperson were more efficient for this particular crisis type used in the experiment. Future research should continue to explore the efficiency of rebuild strategies and a regular employee as a source of information in different situations. This will demonstrate if the effects of crisis response strategies and source hold valid across contexts and might help to improve findings’ generalizability. Logically,
stakeholders might have found a Production Supervisor as better qualified to address the topic of food contamination than a CEO. However, in the event of a financial crisis, for instance, a chief financial officer (CFO) might be perceived as more suitable to address the issue. Also, in the event of a food contamination without any fatalities, a regular employee such a
production supervisor might have been more suitable, signalling that the scale of the crisis is not too large. However, during a more severe crises, a highly-ranked organizational employee such as a CEO, managing director, or a president may be preferable demonstrating
that the crisis is organization’s highest priority and that the organization is doing anything possible to deal with it. Future research needs to test these assumptions.
Moreover, in this study, the CEO and the production supervisor were disclosing crisis information from their private Facebook accounts. Therefore, two different Facebook
templates were used: a corporate Facebook template for the organization and private profile templates for the employees. This might have increased the external validity of results while,
it the same time, decreased their internal validity. To see if the final outcomes will remain the same and simultaneously increase the internal validity of the experiment, in future research, statements might be disclosed from the official company’s account on employees’ behalf, keeping the templates the same.
Furthermore, in this study, the source condition concerned the Facebook account of the organization, a CEO, and a Production Supervisor, who were all disclosing strategic crisis information. However, nowadays, with the advances in communication technologies,
organizations may not always be able to completely control who speaks in the media. By publishing a Facebook post about an issue, for instance, any employee may become an organizational spokesperson. Future research could focus on employees disclosing unofficial crisis information, or so to speak, engaged in whistleblowing. This could have critical
theoretical and practical implications.
To conclude, experimental studies like this one are vital for crisis communication research, bringing important insights into how stakeholders perceive communication
strategies (Tkalac Verčič et al., 2019). Despite the above-stated limitations, the current study provides some valuable suggestions for organizations of how to strategically match their crisis response message with a source to deliver that message to optimize their social media crisis response. Besides, since more questions await future investigations, the study serves as a solid foundation for potential research.
References
Arpan, L. (2002). When in Rome? The effects of spokesperson ethnicity on audience
evaluation of crisis communication. Journal of Business Communication, 39(3), 314– 339. doi.org/10.1177/002194360203900302
Arpan, L., & Roskos-Ewoldsen, D. (2005). Stealing thunder: Analysis of the effects of proactive disclosure of crisis information. Public Relations Review, 31(3), 425–433. doi.org/10.1016/j.pubrev.2005.05.003
Bakker, T., Trilling, D., de Vreese, C., Helfer, L., & Schönbach, K. (2013). The context of content: The impact of source and setting on the credibility of news. Recherches en Communication, 40, 151-168.
Bryman, A. (2016). Social research methods (5th ed.). Oxford: Oxford University Press. Callison, C. (2001). Do PR practitioners have a PR problem?: The effect of associating a source with public relations and client-negative news on audience perception of credibility. Journal of Public Relations Research, 13(3), 219–234.
doi.org/10.1207/S1532754XJPRR1303_2
Callison, C., & Zillmann, D. (2002). Company affiliation and communicative ability: How perceived organizational ties influence source persuasiveness in a company-negative news environment. Journal of Public Relations Research, 14(2), 85–102.
doi.org/10.1207/S1532754XJPRR1402_1
Carroll, C. E. (Ed.). (2016). The SAGE encyclopedia of corporate reputation. Sage Publications.
Claeys, A., & Cauberghe, V. (2014). What makes crisis response strategies work? The impact of crisis involvement and message framing. Journal of Business Research, 67(2), 182– 189. doi.org/10.1016/j.jbusres.2012.10.005
Claeys, A., & Opgenhaffen, M. (2016). Why practitioners do (not) apply crisis
communication theory in practice. Journal of Public Relations Research, 28(5–6), 232–247. doi.org/10.1080/1062726X.2016.1261703
Coombs, W.T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10(3), 163–176. doi.org/10.1057/palgrave.crr.1550049
Coombs, W.T. (2015). Ongoing crisis communication: Planning, managing, and responding. (4th ed.). University of Central Florida: Sage Publications.
Coombs, W.T., & Holladay, S. J. (2002). Helping crisis managers protect reputational assets: Initial tests of the situational crisis communication theory. Management
Communication Quarterly, 16(2), 165–186. doi.org/10.1177/089331802237233 Coombs, W.T., & Holladay, S. J. (2008). Comparing apology to equivalent crisis response
strategies: Clarifying apology’s role and value in crisis communication. Public Relations Review, 34(3), 252–257. doi.org/10.1016/j.pubrev.2008.04.001
Dardis, F., & Haigh, M. (2009). Prescribing versus describing: Testing image restoration strategies in a crisis situation. Corporate Communications: An International Journal, 14(1), 101–118. doi.org/10.1108/13563280910931108
Edelman Trust Barometer (2013). Employee engagement practice. Retrieved from
https://www.slideshare.net/EdelmanInsights/employee-engagement-insights-from-2013-edelman-trust-barometer-21279838
Eriksson, M. (2018). Lessons for crisis communication on social media: A systematic review of what research tells the practice. International Journal of Strategic
Communication, 12(5), 526–551. doi.org/10.1080/1553118X.2018.1510405 Fearn-Banks, K. (2011). Crisis communications: A Casebook Approach. (4th ed.). UK: Rutledge.
Flanagin, A. J., & Metzger, M. J. (2008). Digital media and youth: Unparalleled opportunity and unprecedented responsibility. In M. J. Metzger & A. J. Flanagin (Eds.), Digital media, youth, and credibility (pp. 5–28). Cambridge, MA: MIT Press.
Flanagin, A. J, & Metzger, M. J. (2013). Trusting expert- versus user-generated ratings online: The role of information volume, valence, and consumer characteristics.
Computers in Human Behavior, 29(4), 1626–1634. doi.org/10.1016/j.chb.2013.02.001 Fombrun, C. J., & van Riel, C. B. M. (2004). Fame & fortune: How successful companies
build winning reputations. New York: Prentice Hall Financial Times.
Freberg, K. (2012). Intention to comply with crisis messages communicated via social media. Public Relations Review, 38(3), 416–421.
doi.org/10.1016/j.pubrev.2012.01.008
Hellmueller, L., & Trilling, D. (2012). The credibility of credibility measures: A meta-analysis in leading communication journals, 1951 to 2011. In WAPOR Hong Kong 2012: Paper presentation. Lincoln, NE/Hong Kong: World Association for Public Opinion Research/Public Opinion Programme, The University of Hong Kong. Hovland, C., & Weiss, W. (1951). The influence of source credibility on communication
effectiveness. The Public Opinion Quarterly, 15, 635–650. doi.org/10.1086/266350 Hovland, C., Janis, I., & Kelley, H. (1953). Communication and persuasion: Psychological
studies of opinion change. New Haven: Yale University Press.
Jaques, T. (2012). Crisis leadership: A view from the executive suite. Journal of Public Affairs, 12(4), 366–372. doi.org/10.1002/pa.1422
Kiambi, D., & Shafer, A. (2016). Corporate crisis communication: Examining the interplay of reputation and crisis response strategies. Mass Communication and Society, 19(2), 127–148. hdoi.org/10.1080/15205436.2015.1066013
Kim, S., Avery, E., & Lariscy, R. (2009). Are crisis communicators practicing what we preach?: An evaluation of crisis response strategy analyzed in public relations research from 1991 to 2009. Public Relations Review, 35(4), 446–448.
doi.org/10.1016/j.pubrev.2009.08.002
Kim, Y., & Park, H. (2017). Is there still a PR problem online? Exploring the effects of different sources and crisis response strategies in online crisis communication via social media. Corporate Reputation Review, 20(1), 76–104. doi.org/10.1057/s41299-017-0016-5
Lee, B. (2005). Hong Kong consumers’ evaluation in an airline crash: A path model analysis. Journal of Public Relations Research, 17(4), 363–391.
doi.org/10.1207/s1532754xjprr1704_3
Lee, J., Kim, S., & Wertz, E. (2014). How spokesperson rank and selected media channels impact perceptions in crisis communication. Public Relations Journal, 8(2), 1–21. Retrieved from https://doaj.org/article/f790d114f59b49d08849489d3ced4bd
Liu, B., Austin, L., & Jin, Y. (2011). How publics respond to crisis communication strategies: The interplay of information form and source. Public Relations Review, 37(4), 345– 353. doi.org/10.1016/j.pubrev.2011.08.004
Ma, L., & Zhan, M. (2016). Effects of attributed responsibility and response strategies on organizational reputation: A meta-analysis of situational crisis communication theory research. Journal of Public Relations Research, 28(2), 102–119.
doi.org/10.1080/1062726X.2016.1166367
Metzger, M., & Flanagin, A. (2013). Credibility and trust of information in online
environments: The use of cognitive heuristics. Journal of Pragmatics, 59(PB), 210– 220. doi.org/10.1016/j.pragma.2013.07.012
Metzger, M., Flanagin, A., & Zwarun, L. (2003). College student web use, perceptions of information credibility, and verification behavior. Computers & Education, 41(3), 271–290. doi.org/10.1016/S0360-1315(03)00049-6
Metzger, M. J., Flanagin, A. J., Eyal, K., Lemus, D. R., & McCann, R. M. (2003). Credibility for the 21st century: Integrating perspectives on source, message, and media
credibility in the contemporary media environment. Annals of the International Communication Association, 27(1), 293–335.
doi.org/10.1080/23808985.2003.11679029
Miller, J. M., & Krosnick, J. A. (2000). News media impact on the ingredients of presidential evaluations: politically knowledgeable citizens are guided by a trusted
source. American Journal of Political Science, 44(2), 301–315. doi.org/10.2307/2669312
Miller, A., & Kurpius, D. (2010). A citizen-eye view of television news source credibility. American Behavioral Scientist, 54(2), 137–156. doi.org/10.1177/0002764210376315 Nan, X. (2009). The influence of source credibility on attitude certainty: Exploring the
moderating effects of timing of source identification and individual need for cognition. Psychology and Marketing, 26(4), 321–332. doi.org/10.1002/mar.20275 O’Keefe, D. J. (2002). Persuasion: Theory and research (2nd ed.). Thousand Oaks, CA: Sage.
Roh, S. (2017). Examining the paracrisis online: The effects of message source, response strategies and social vigilantism on public responses. Public Relations Review, 43(3), 587–596. doi.org/10.1016/j.pubrev.2017.03.004
Seeger, M., & Ulmer, R. (2001). Virtuous responses to organizational crisis: Aaron Feuerstein and Milt Colt. Journal of Business Ethics, 31(4), 369–376.
Shoemaker, P. J., & Vos, T. P. (2009). Media gatekeeping. In D. W. Stacks & M. B. Salwen (Eds.), An integrated approach to communication theory and research (2nd ed.). (pp. 75–89). New York: Routledge.
Snoeijers, E., Poels, K., Nicolay, C., Khan, G., Hoffman, M., & Misztur, T. (2014).
#universitycrisis: The impact of social media type, source, and information on student responses toward a university crisis. Social Science Computer Review, 32(5), 647– 661. doi.org/10.1177/0894439314525025
Stephens, K. K. & P. Malone. (2010). New media for crisis communication: Opportunities for technical translation, dialogue and stakeholder responses.” In W. T. Coombs & S. J. Holladay (Eds.), The handbook of crisis communication, (pp. 381–395). UK: Blackwell Publishing.
Tkalac Verčič, A., Verčič, D., & Coombs, W. (2019). Convergence of crisis response strategy and source credibility: Who can you trust? Journal of Contingencies and Crisis
Management, 27(1), 28–37. doi.org/10.1111/1468-5973.12229
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. doi.org/10.1016/j.pubrev.2012.09.010 van der Meer, T., & Verhoeven, J. W. M. (2014). Emotional crisis communication. Public
Relations Review, 40(3), 526-536. doi.org/10.1016/j.pubrev.2014.03.004 van Zoonen, W., & van der Meer, T. (2015). The importance of source and credibility
perception in times of crisis: Crisis communication in a socially mediated era. Journal of Public Relations Research, 27(5), 371–388.
Verhoeven, J. W. M., van Hoof, J. J., ter Keurs, H., & van Vuuren, H. A. (2012). Effects of apologies and crisis responsibility on corporate and spokesperson reputation. Public Relations Review, 38(3), 501–504. doi.org/10.1016/j.pubrev.2012.02.002
Westerman, D., Spence, P., & van der Heide, B. (2012). A social network as information: The effect of system generated reports of connectedness on credibility on
Twitter. Computers in Human Behavior, 28(1), 199–206. doi.org/10.1016/j.chb.2011.09.001
Xu, J., & Wu, Y. (2015). Using Twitter in crisis management for organizations bearing different country-of-origin perceptions. Journal of Communication Management, 19(3), 239–253. doi.org/10.1108/JCOM-06-2013-0050
Yang, S., Kang, M., & Johnson, P. (2010). Effects of narratives, openness to dialogic communication, and credibility on engagement in crisis communication through organizational blogs. Communication Research, 37(4), 473–497.
doi.org/10.1177/0093650210362682
Zaremba, A. (2010). Crisis communication: Theory and practice. New York: Routledge. Zheng, B., Liu, H., & Davison, R. (2018). Exploring the relationship between corporate
reputation and the public’s crisis communication on social media. Public Relations Review, 44(1), 56–64. doi.org/10.1016/j.pubrev.2017.12.006
Appendix A
Appendix B Table B1
Results of a two-way ANCOVA of the main effects and the interaction of crisis response strategy and source of information on reputation, when controlling for source credibility perception
Sum of Squares df Mean Square F p η2 Source Credibility 119.67 1 119.67 132.38 .000 .02 Crisis Response Strategy 35.35 1 35.35 39.11 .000 .01
Source of Information 7.83 2 3.92 4.33 .014 .001
Crisis Response Strategy* Source of Information
.36 2 .18 .20 .822
Error 291.08 322 .90
Total 6703.56 329
Figure B1. Non-significant interaction effect of crisis response strategy and source of crisis information on reputation, when controlling for source credibility perception
Table B2
Group means and Standard deviations for all conditions in the two-way ANCOVA
Crisis response strategy Source of information M SD N Denial Organization 3.66 1.31 57 CEO 3.55 1.20 53 Regular employee 4.03 1.11 53 Total 3.74 1.22 163 Rebuild Organization 4.95 .89 53 CEO 4.65 1.22 56 Regular employee 5.12 .97 57 Total 4.91 1.05 166 Total Organization 4.28 1.30 110 CEO 4.11 1.32 109 Regular employee 4.59 1.17 110 Total 4.33 1.28 329
Syntax * Encoding: UTF-8.
Encoding: UTF-8. * Cleaning data.
DATASET ACTIVATE DataSet1.
FREQUENCIES VARIABLES=Age Gender Edu Count FB /STATISTICS=STDDEV MEAN
/ORDER=ANALYSIS. * Define Variable Properties. *Age.
MISSING VALUES Age('2 '). *Count.
MISSING VALUES Count(' ', 'I prefer'). EXECUTE.
* Age 17 --> case excluded
*Recoding variable country 1) Making sure all countries are written the same, 2) automatic recode.
AUTORECODE VARIABLES=Count /INTO Country_rec
/PRINT.
*Frequencies Country, Facebook (yes or no), Gender, Age, Education FREQUENCIES VARIABLES=Country_rec FB Edu Gender Age /STATISTICS=STDDEV MEAN MEDIAN MODE
/ORDER=ANALYSIS.
*Computing condition variable.
*1 step: recode the conditions so that instead of 1 everywhere, one column has 2, other 3, etc. *2 step: compute a new experimental conditions variable using the mean.
RECODE FL_26_DO_Group1DenialxOrganization (1=1) INTO Group1_DenialxOrganization.
VARIABLE LABELS Group1_DenialxOrganization 'Group1_DenialxOrganization'. EXECUTE.
RECODE FL_26_DO_Group2RebuildxOrganization (1=2) INTO Group2_RebuildxOrganization.
VARIABLE LABELS Group2_RebuildxOrganization 'Group2_RebuildxOrganization'. EXECUTE.
RECODE FL_26_DO_Group3DenialxCEO (1=3) INTO Group3_DenialxCEO. VARIABLE LABELS Group3_DenialxCEO 'Group3_DenialxCEO'.
EXECUTE.
RECODE FL_26_DO_Group4RebuildxCEO (1=4) INTO Group4_RebuildxCEO. VARIABLE LABELS Group4_RebuildxCEO 'Group4_RebuildxCEO'.
