EXPLORING HOW TWO-PHASE CRISIS RESPONSE STRATEGIES INFLUENCE PUBLIC PERCEPTIONS OF AN ORGANIZATION IN A CORPORATE SOCIAL RESPONSIBILITY-RELATED CRISIS By Shu (Karen) Zhang deGategno A Capstone Project Presented to the Faculty of the School of Communication in Partial Fulfillment of the Requirements for the Degree of Masters of Arts in Strategic Communication Supervisor: Dr. Joseph Erba American University April 24, 2014 © COPYRIGHT Shu (Karen) Zhang deGategno 2014 ALL RIGHTS RESERVED ABSTRACT Theories of crisis communication have developed tremendously in the past ten years and have helped crisis communication practitioners rationally handle crisis situations. These existing theories largely focus on planning single-phase response strategies to crises. In practice, however, a crisis usually involves several, different phases due to changing events, responses and new public reactions over time. This study aims to test the crisis response strategies of the most dominant crisis communication theory, Situational Crisis Communication Theory’s (SCCT), in two phases of responses to a crisis scenario related to corporate social responsibility (CSR). Through an experiment with 97 participants, this study seeks to explore the relationship between an organization’s initial and later crisis responses and public perceptions of the organization in terms of perception of crisis response, customer treatment, organization and the CSR initiative. The results of the experiment indicate that SCCT’s different levels of acceptance are limited to measuring the acceptance of crisis responses and are not able to measure overall perceptions of an organization. Separating perceptions into four dimensions provided a more comprehensive picture of how the public perceives crisis responses. For instance, an excuse response can be the most effective response strategy in certain circumstances. A full apology response is not recommended in some situations. Crisis responses of no comment or denial are so publicly unacceptable that when an organization responds with either of these strategies, compensation has to be made in order to improve public perceptions of the organization. Despite some limitations, this study contributes to understanding public perceptions of organization during crises. Specifically, its main contribution is its experimental design using two-phase crisis responses. Participants perceive two different responses at two different times about the same organization during the same crisis. By taking different phases of crisis responses into account, identifying the changes of public perception and quantifying their effects, this study expands on existing crisis communication theories and adds practical knowledge for dealing with real-world crises. Contents Acknowledgement................................................................................................................................................................... ii List of Tables............................................................................................................................................................................. iii Introduction ............................................................................................................................................................................... 1 Literature Review .................................................................................................................................................................... 5 Defining Organizational Crisis....................................................................................................................................... 5 Responding to a Crisis ...................................................................................................................................................... 6 Corporate Social Responsibility.................................................................................................................................... 7 Relationship between CSR and Crisis ......................................................................................................................... 9 Situational Crisis Communication Theory.............................................................................................................. 12 Research Question and Hypotheses .............................................................................................................................. 16 Methods ..................................................................................................................................................................................... 18 Overview .............................................................................................................................................................................. 18 Participants ......................................................................................................................................................................... 18 Recruitment ........................................................................................................................................................................ 19 Experimental Procedures .............................................................................................................................................. 19 Measurement...................................................................................................................................................................... 21 Consent ................................................................................................................................................................................. 22 Privacy................................................................................................................................................................................... 22 Data storage ........................................................................................................................................................................ 22 Ethical concerns ................................................................................................................................................................ 23 Reliability & Validity........................................................................................................................................................ 23 Results ........................................................................................................................................................................................ 24 Discussion ................................................................................................................................................................................. 30 Conclusions .............................................................................................................................................................................. 33 Recommendations............................................................................................................................................................ 34 Implications ........................................................................................................................................................................ 36 Limitations........................................................................................................................................................................... 38 Future Research ................................................................................................................................................................ 39 References ................................................................................................................................................................................ 41 Appendices ............................................................................................................................................................................... 46 Stimuli ................................................................................................................................................................................... 46 News Story Questionnaire ............................................................................................................................................ 52 General Questionnaire .................................................................................................................................................... 54 i Acknowledgement I would like to thank my research professor and mentor, Dr. Joseph Erba, who was always available to exchange ideas and discuss my research with me. He walked me through the whole research process from the very beginning till the very end with great patience and understanding. I would like to thank my husband Patrick deGategno. His advice and suggestions were extremely valuable in developing and executing this study. I certainly cannot make this achievement without his full support, love, and care. Thanks to my mother Xiufang Tang, who has and continues to give me all the help and support she can. And finally, thanks to my daughter, Alexandra. She is the reason of why I do what I do. ii List of Tables Table 1: SCCT Crisis Response Strategies ................................................................................................. 13 Table 2: Crisis Response Strategies by Level of Responsibility Acceptance ............................... 14 Table 3: Four Conditions of Stimuli in This Study ................................................................................. 20 Table 4: Number of Participants in Four Conditions ........................................................................... 24 Table 5: Comparisons of Condition 2 (Denial – Excuse) ..................................................................... 27 Table 6: Comparisons of Condition 3 (Excuse – Full Apology) ........................................................ 27 Table 7: Comparisons of Condition 4 (Full Apology - Compensation) .......................................... 27 Table 8: Recommended Second Phase Responses based on Results of This Study ................. 36 iii Introduction Situational Crisis Communication Theory (SCCT) is considered a dominant organizational crisis theory (Liu, Austin & Jin, 2010). It takes an organization’s previous reputation into consideration during a crisis and gives crisis managers a convenient schema for identifying types of crises in order to craft response strategies intended to minimize the damage to the organization’s image and reputation (Coombs, 2007). SCCT’s schema though was developed according to a series of experimental studies of static crises, or crisis situations that do not change over time. SCCT works well theoretically, but it has only been tested under conditions of such static crises. Crises rarely remain static. The theoretical framework needs to be further explored to match the dynamic conditions of crisis situations, crisis responses and changing public perceptions. In practice, a crisis usually evolves into different phases because of a string of new events, organizational responses and public reactions. For example, Malaysian Airlines’ MH370 airplane disappeared on March 14, 2014 and still cannot be found at the time of this writing. This crisis for Malaysian Airlines has involved several phases, and could potentially involve more before it is resolved. In phase one, the airline and the Malaysian government did not respond immediately to the plane’s disappearance, which caused the public to speculate while waiting for a response. In phase two, the airline and the government began sharing a wide variety of information and publicly speculated on what might have happened to MH370. These responses caused the families and the public to accuse the airline and the government of hiding the truth, which further damaged their credibility. In phase three, Malaysian Airlines and the government judged without any evidence that “it had to be assumed ‘beyond reasonable doubt’ that the plane was lost and 1 there were no survivors” (BBC, 2014a). This “conclusion” was deemed by the global public to be unsubstantiated and irresponsible, severely damaging the already tarnished reputations of the airline and the Malaysian authorities (BBC, 2014b). Currently, the crisis has achieved a fourth phase in which international search teams including Australia, China, Japan, Malaysia, New Zealand, South Korea, the United Kingdom and the United States are still making their best efforts to search for the missing airplane (BBC, 2014b). This multinational search effort makes the previous Malaysian response appear even more irresponsible by comparison. In a dynamic crisis such as this which features multiple phases, organizations cannot rest upon a single response and must respond to each phase if they hope to salvage their public reputation. How should an organization craft its response strategies in each phase? How does public perception of an organization’s previous response affect subsequent perceptions of the organization in later phases? How should the organization respond in order to improve the public’s previous perception towards the organization? In the Malaysian Airlines case, what should the airline have done in response to the incident from the beginning, and how would a different initial crisis response have influenced the airline’s future response(s)? Existing crisis communication theories do not appear capable of answering these questions for the following four reasons: First, most crisis communication studies focus on the business side of the crisis asking how organizations craft a response that will alleviate the crisis and minimize the damage to operations (Bergman, 1994; Sturges, 1994). Crisis responses change public impressions of organizations, but these are not simply from bad to good or from concern to confidence. 2 Second, studies focusing on how the public reacts to an organization’s crisis response only measure public perceptions of the organization’s general response during a single phase of the crisis (Dean, 2004; Coombs, 2007). These studies do not seem to account for the reflexive communication inherent in the supply–demand relationships organizations have with their stakeholders and the public. These studies instead appear to assume that crisis responses are transmissions to the public, that crises remain static, and that an organization’s response will remain unchanged. In practice, however, like the Malaysian Airlines case, crises often have several phases, and organizations have to respond at each phase as the crisis – and the public’s perceptions of the organization – changes. Third, crisis case studies offer valuable insights for crisis communicators in terms of lessons learned and best practices, but crisis responses are highly contextual, no two crises are alike, and past experience is not uniformly applicable. There are case studies that reflect changes of stakeholder perceptions during a particular crisis (Graham & Greenberg, 2002; Klein & Dawar, 2004). Theories such as Image Restoration Strategies (Benoit, 1995, 1997) and Impression Management Strategies (Allen and Caillouet, 1994) are drawn from previous crisis incidents in which an organization in crisis either succeeded or failed in salvaging its reputation. These studies and theories, however, do not offer a framework for crisis practitioners to follow in order to either forecast public reactions or plan a dynamic organizational response widely applicable to a variety of crises. Lastly, SCCT, the most dynamic crisis communication theory available to practitioners, does not as yet appear to account for changes in public perceptions during multiphase crises. Because SCCT is an “audience-centered approach to crisis 3 communication” (p. 143) and relies on experimentation rather than case studies, it is a significant improvement to other theories. SCCT offers guidance to crisis practitioners for mapping crisis situations using three important considerations: initial crisis responsibility, crisis history and prior relational reputation (Coombs, 2007). None of these considerations, however, relates to the organization’s previous crisis response, and the theory itself seems more focused on single-phased crisis responses. There appears to be no research yet conducted that uses SCCT in the context of public reactions to organizations’ crisis responses during different crisis phases. Studies of this sort are badly needed in order to provide recommendations on situations like the Malaysian Airlines’ to crisis managers. This is the deficit in crisis communication that this research hopes to address. This study is designed to test SCCT response strategies in two phases of responses to a crisis scenario related to corporate social responsibility (CSR). The researcher seeks to explore the relationship between an organization’s initial crisis response and later crisis response and how they affect public perceptions of the organization. More specifically, this study explores how public perceptions of the organization are affected by an organization’s two-phase responses to a crisis scenario related to CSR. Through an experiment involving 97 participants, this study measures changes in public perceptions toward an organization across four variables, including perception of the organization’s crisis response, perception of the organization’s treatment of its customers, perception of the organization itself, and perception of the organization’s CSR initiative. 4 Literature Review Defining Organizational Crisis Researchers have developed multiple definitions of organizational crises. Coombs (2007b), however, has synthesized a definition which attempts to capture the common characteristics other authors have used to describe this type of crisis (p. 3). Coombs defined an organizational crisis as “the perception of an unpredictable event that threatens important expectancies of stakeholders and can seriously impact an organization’s performance and generate negative outcomes” (p. 2). A few unique characteristics distinguish organizational crises from other events. Coombs (2007b) notes for instance that a crisis is perceptual. Whether a company is in crisis depends on whether stakeholders perceive an event is a crisis. “If stakeholders believe an organization is in crisis, a crisis does exist, and stakeholders will react to the organization as if it is in crisis” (p. 3). Other studies have shown that the causes of crises are closely linked to the relationship between the organization and its stakeholders or the general public. Dionisopolous and Vibbert (1988) assert that organizations have a public persona and are generally perceived as individuals by their stakeholders and the public. In this sense, a crisis primarily embodies an attack on the character of the organization. Coombs (2007b) further notes, “Crises can violate expectations that stakeholders hold about how organizations should act” (p.3). In a case study, Ice (1991) expands on this idea, contending that organizations serve the expectations of a variety of stakeholders, not some monolithic, homogenous public having only one relationship with an organization. Instead, stakeholders have different interests, expectations, and relationships with the organization. Thus, the stakeholders’ perceptions of a crisis depend on their existing 5 expectations of the organization and their varying interests in the crisis. As such, different stakeholders will be affected in different ways by the crisis. Relatedly, Hobbs (1995) notes that a crisis can be anything that damages or breaks the relationship between the organization in crisis and its stakeholders. Hearit (1994, 1995) defines an organization’s relationship with its stakeholders in terms of social legitimacy, the match between an organization’s values and its stakeholders’ values. A crisis can be a threat to the organization’s social legitimacy, i.e. a disruption perceived by the stakeholders in the match between theirs and the organization’s values. This social legitimacy violation is a form of character attack which prompts an organization to defend itself and repair its character (Hearit 1994 and 1995). Responding to a Crisis Any crisis response has as its ultimate goal restoration of the company’s image (Allen and Caillouet 1994; Coombs 2010; Hobbs 1995). Companies use crisis response strategies to “repair the reputation, to reduce negative affect and to prevent negative behavioral intentions” (Coombs, 2007a. p.170). Case studies have served as the primary method researchers have used to develop crisis response theories. Different crisis communication researchers developed different crisis response approaches based on case studies. For example, Allen and Caillouet (1994) used impression management and account research – the research of messages that designed to explain the event and influence perceptions of offending organization and developed a list of 16 impression management strategies. Benoit (1995) similarly studied apologia and account to develop 14 image restoration strategies. Both of these two sets of 6 strategies aimed to be used by the organization to shape the public reputation by managing impressions of a crisis and repair reputational damage from a crisis (Coombs, 2007b). Coombs (2007b), however, developed SCCT, which moves beyond the typical case study approach to conduct experimental research and develop an “audience-centered approach to crisis communication” (p.143). As will be shown later in the literature review section, SCCT is designed to facilitate company efforts to determine its crisis responsibility by evaluating the reputational threat (i.e. the situation) it faces and create crisis response strategies based on the accountability it must assume in accordance with its crisis responsibility (Coombs, 2007b). There appears to be little experimental research conducted in support of crisis response theories. One crisis communication experiment, however, successfully linked how stakeholders’ perceptions of a firm in crisis are influenced to three factors: (1) the companies’ reputation for social responsibility prior to a crisis, what Hearit (1994, 1995)would refer to as its prior social legitimacy; (2) the company’s response to the crisis, its self-defense; and (3) their culpability for the crisis (Dean 2004). Dean found that each of these factors has a direct relationship to stakeholder perceptions of the organization in crisis, with reputation having the greatest effect and response having the least on the stakeholders’ perceptions of the firm (Dean 2004). These factors clearly reflect and reinforce Coombs’ research results, as well as his theory. Corporate Social Responsibility From its roots in business practices going back to the 1950s to the present, CSR as a concept has linked organizations’ impact on society to their public reputation and overall 7 business operations. Yet there remains no clear definition of the concept found in any scholarly study (Motwani, 2012). In the 1950s, CSR was initially referred to as social responsibility (SR) by authors such as Howard R. Bowen because as the concept of corporation had not yet fully developed. In the 1960s, authors such as William C. Frederick and Joseph W. McGuire defined SR as “…certain responsibilities to society which extend beyond…” a corporation’s economic and legal obligations (McGuire, 1963. p. 144). In the 1970s, scholars began to assert that companies could only be judged socially responsible on the basis of “the actual policies with which they were associated” (Heald, 1970. p. xi), or rather on the consequences of their business practices for society. Scholars in the 1970s also began to contend that social responsibility was a long-run profit maximizer: “The social responsibility of business is to increase its profits” (Friedman 1970). From then on until the 1990s, further empirical research was conducted on CSR. However, this research has not moved beyond the conclusions of scholars in the seventies. As Carroll’s (1991) research indicates, the concept of CSR links the success of the companies’ business operations to its public reputation by acknowledging that the company has certain social obligations it must meet in order to maintain a positive public image (p. 42). Regardless of how CSR is defined, it is widely believed to include a key element that is, CSR means going beyond what is required by law. It is more than a company’s own interest and legal requirement, but an approach of doing well by doing good (McWilliams, & Siegel, 2001). It is increasingly popular in public relations for corporations to publicly claim their business embodies corporate social responsibility (CSR) (Moir, 2001). Corporations that 8 accept these social responsibilities typically declare to the public that they promise to do so and commit themselves to various public initiatives that are intended to prove their sincerity to uphold these promises. By communicating good deeds, exerting considerable positive influence on public opinion, corporations are rewarded for doing well by embedding CSR in the business operation and enhancing people’s perceptions of the image, reputation and credibility (Pfau et al., 2008). For example, each focus area of Wal-mart’s CSR initiative can be linked to its business practices. Wal-mart’s CSR initiative focuses on “environmental sustainability”, “U.S. manufacturing”, “women’s economic empowerment”, “hunger & nutrition”, “diversity & inclusion”, “veterans & military families” and “responsible sourcing”. “Environmental sustainability” and “U.S. manufacturing” basically depict Wal-mart’s perspective on its manufacturing and production practices. “Hunger & nutrition” and “veterans & military families” addresses its retail business. And as the largest private employer in the world (CNN, 2012), “women’s economic empowerment” and “diversity & inclusion” constitute declarations of its human resources policies. By consistently practicing CSR initiatives, Walmart is “proud” of its efforts “…to become a more sustainable and more responsible business” (Wal-mart website). Relationship between CSR and Crisis From a public relations perspective, CSR, however it is defined, is clearly an effort by organizations to publicly project an ideal image of themselves as being socially responsible (Batruch, 2011; Klein & Dawar, 2004). CSR is thus one aspect of an organization’s image branding strategy, and like branding, is designed to establish and enhance an 9 organization’s public reputation (Pfau et al., 2008). CSR initiatives accomplish this ultimate goal by managing public impressions of the organization and its business in order to influence the public to perceive the organization’s business activities as good for society (Klein & Dawar, 2004). In cases such as Wal-mart’s, which Coombs and Holladay (2006) studied, CSR initiatives create a “halo effect” that “shield” companies from having stakeholders “automatically assume the worse and assign a high amount of crisis responsibility” in crisis situations (p. 134). In this sense, CSR seems to be a practical tool in crisis communication to reduce the organization’s crisis responsibility and ease the tension between the organization and its stakeholders via establishing a good organizational reputation. A CSR initiative appears to reduce the company’s exposure to risks of crises. Taking into consideration various social issues through the organization’s business operations enables the organization “to anticipate potential risks, put in place mitigating measures, and align itself with a growing number of stakeholders” (Batruch, 2011). Thus, CSR seems to be a measure undertaken by a company to minimize the company’s exposure to potential future crises that could disrupt business and hurt profits. Sometimes, however, CSR initiatives can become part of or even create a crisis. When businesses accept their social obligations and then are involved in some incident that public perceives has violated the business’ social obligations, CSR can backfire and threaten the business’ reputation. The possibility of this happening with a CSR initiative depends on how the business and the public jointly define the business’ social responsibilities. For example, CSR Europe, a membership-based organization of over 5,000 companies across the world aiming to 10 “bring the CSR agenda forward”, looks at these following areas in their reports and guides (CSReurope.org): Organizational accountability, including CSR strategy and corporate governance, reporting and disclosure, and product and services. Environment, including resource efficiency, energy and climate change, environment protection and biodiversity. Human rights, including supply chain and procurement, child and compulsory labor, non-discrimination, and freedom of association. Community/society, including community impact/engagement, stakeholder engagement, education, health and wellbeing, and poverty alleviation initiatives. Labor practices, including employment, training and education, diversity and equal opportunities, health and safety, and active ageing. Sustainable living in cities, including inclusive business models and sustainable lifestyle. Such a broad set of social responsibilities opens CSR Europe members to a wide array of possible crises that have a CSR component to them. A few examples of crises can be identified that violate the social obligations to which CSR Europe companies commit themselves: Environmental crisis: BP oil spill (Cherry & Sneirson, 2011). Employee and labor crisis: FoxConn employee suicide cases (Xu & Li, 2013). Organizational accountability: Toyota 2010-11 recalls (Andrews et al., 2011). 11 Situational Crisis Communication Theory Based on attribution theory, SCCT is intended to be an evidence-based framework for crisis managers can use during a crisis to identify factors influencing the crisis and how stakeholders perceive the situation (Coombs, 2007b). It is an “attempt to provide crisis managers with a resource for making informed decisions concerning ways to protect the organizational reputation during a crisis” (Coombs & Holladay, 2002. p. 182) “The crisis situation is the focal point of SCCT. The amount of reputational damage a crisis situation can inflict drives the selection of the crisis response strategy.” (Coombs, 2006. p.243) SCCT developed a crisis assessment system having two steps. First, assess the type of crisis the organization faces (Coombs, 2007b). SCCT divides crises into three groups: (1) victim, where the organization takes little responsibility and further, the organization itself is the victim of the crisis; (2) accidental, which involves minimum responsibility from the organization and is mostly contributed by other factors such as weather; (3) preventable crisis, which makes the stakeholders strongly believe that it is the organization’s responsibility for the crisis. These specific types of organizational crises produce “…specific and predictable levels of crisis responsibility – attributions of organizational responsibility for the crisis” (Coombs, 2007a. p.168). Second, assess the threat of the incident to the organization’s reputation based on two intensifying factors, which contribute to stakeholders’ perception of a crisis: the organization’s history of crisis and its prior negative reputation (Coombs, 2007b). If either of these two factors exists, people will attribute greater responsibility to the organization for the crisis. 12 With the results of this crisis assessment, the organization can select the crisis response strategy recommended by SCCT. These strategies are organized according to whether the intent of the strategy is to change public perceptions of the crisis or of the organization in crisis (Coombs, 2007b). The strategies are grouped into four postures described in Table 1. Table 1: SCCT Crisis Response Strategies Denial Posture Attack the accuser: Crisis manager confronts the person or group claiming something is wrong with the organization Denial: Crisis manager asserts that there is no crisis. Scapegoating: Crisis manager blames some person or group outside of the organization for the crisis. Diminish Posture Excuse: Justification: Rebuild Posture Compensation: Apology: Crisis manager minimizes organizational responsibility by denying intent to do harm and/or claiming inability to control the events that triggered the crisis. Crisis manager minimizes the perceived damage caused by the crisis. Crisis manager offers money or other gifts to victims. Crisis manager indicates the organization takes full responsibility for the crisis and asks stakeholders for forgiveness. Bolstering Posture Reminder: Tell stakeholders about the past good works of the organization. Ingratiation: Crisis manager praises stakeholders and/or reminds them of past good works by the organization. Victimage: Crisis managers remind stakeholders that the organization is a victim of the crisis too. Table Source: Coombs, 2007b. p.140 SCCT also categorizes response strategies to reflect the amount of responsibility each strategy is perceived to accept (Coombs, 2010). These are outlined in Table 2. 13 Table 2: Crisis Response Strategies by Level of Responsibility Acceptance 1. Full Apology: organization takes full responsibility for crisis Very High Acceptance and requests forgiveness from stakeholders. Can include compensation. 2. Correction Action: organization takes steps to repair the High Acceptance crisis damage and/or prevent a recurrence of the crisis. 3. Ingratiation: the organization reminds stakeholders of past Mild Acceptance good works by the organization or praises the stakeholders in some fashion. 4. Justification: organization tries to minimize perceived Mild Acceptance damage related to crisis. Includes claim that damage was minor or victim deserved it. 5. Excuse: organization tries to minimize its responsibility for Mild Acceptance the crisis. Includes denying intent or control over the crisis event. 6. Denial: organization maintains that no crisis occurred. No Acceptance Response may include efforts to explain why there was no crisis. 7. Attack the Accuser: organization confronts people or group No Acceptance who say that a crisis exists. Response may include a threat such as a lawsuit. Table Source: Coombs, 2010. p.159 This categorization explicitly indicates that “full apology” yields “high acceptance”. With more and more empirical research being conducted, researchers overly promote apology as the response (Coombs & Holladay, 2008). However, apology is the organization admitting full responsibility for the crisis. It can be used “as evidence in court to win lawsuits against the organization” (p. 253). With this in mind, Coombs and Holladay attempted to clarify apology’s role compared with responses of sympathy and compensation, in order to offer alternative response strategies to crisis managers to avoid expensive lawsuits. The result indicate no significant difference in the effects of apology, compensation and sympathy on public perceptions. 14 SCCT explains the dynamic between a company’s crisis response and stakeholder’s perception to the organization in theoretical detail. As such, SCCT is adopted as the theoretical framework for this capstone research. The present experiment is designed with two phases of stimuli, aiming to measure the change of perceptions according to different crisis responses. The first phase of stimuli will test whether or not the level of responsibility acceptance is valid in a CSR related crisis; the second phase of stimuli will be changed upon the levels of responsibility acceptance. The result is expected to show how the public’s perceptions are changed in this CSR related crisis considering the factor of the initial crisis response. 15 Research Question and Hypotheses This study seeks to answer how public perception of an organization changes according to the organization’s different responses to a crisis involving its CSR initiative. This study first will test participants’ perceptions of the organization after giving its firstphase crisis response. The crisis response will then change based on SCCT’s level of acceptance of different crisis response strategies after the first organization response, and then participants will state their perceptions of the organization after the second-phase crisis response. The changes between these two responses will be analyzed to answer the research question. SCCT predicts that excuse response strategy has higher level of acceptance than denial. While “denial” has acceptance level of “no acceptance”, “excuse” has the level of “mild acceptance” (Table 2). Hypothesis 1 is based on this conclusion. H1: Participants will display more favorable overall perceptions of the company after reading that the company presented its excuse for an incident than after reading that the company denied being involved in the incident. Hypothesis 2 is based upon the same level of acceptance table. Based on this table, full apology response strategy has higher level of acceptance than excuse. “Full apology’s” acceptance level is “high”, whereas “excuse” is “mild” (Table 2). H2: Participants will display more favorable overall perceptions of the company after reading that the company presented its full apology for an incident than after reading that the company presented its excuse for the incident. While SCCT predicts the level of acceptance of different crisis response strategies, relevant studies have also explored the different perceptions to responses in the same 16 acceptance level. For instance, Coombs & Holladay (2008) found that among all response strategies, full apology is popularly recommended by crisis communication researchers; however, there is no significant different effectiveness between full apology and compensation (Coombs & Holladay, 2008). The third hypothesis was made upon this finding. H3: Participants will display similar overall favorability towards the company after reading that the company provided compensation in an incident than after reading that the company presented its full apology for the incident. Since Table 2 lists that full apology yields the highest level of acceptance, it seems natural to make the following hypothesis: H4: After reading the first story, participants who read the response of full apology will display the highest acceptance comparing with participants who read the responses of giving no comment, denial or excuse. The results of this study are intended to provide crisis managers practical recommendations to change or improve the crisis response strategies based on the change of public perception to the organization, which is influenced by the organization’s previous crisis response. 17 Methods Overview This experiment study explores potential changes to public perceptions based on changes in a realistic, fictional company’s responses to a realistic, fictional CSR crisis. There are five different responses from the company to the crisis, which are arranged in different combinations to create four treatment conditions. One condition only has two phases of crisis responses (table 3). All crisis responses are based on the level of public acceptance to the responses (table 2) (Coombs, 2010). Participants were randomly assigned to one of four conditions. They were exposed to one (condition 1) or two stories (condition 2,3 and 4) about the same company responding in different ways to a crisis, and answered identical set of questions after having read each story. Comparing the changes of the answers to the questions is the metrics to measure the RQ. Participants The participants in this research belong to the public who will be affected by the CSR crisis. In this particular crisis scenario created, the potential participants must be able to see themselves as a local community member who consumes natural gas. There were 97 participants who completed the questionnaire. The age range is from 18 to 77 (M=34.49, SD=13.098). Within these participants, majority of them are Caucasian (72.2%, N =70), versus 11.3% Asian (N=11), 6.2% black (N=6), 4.1% Hispanic (N =4), and 6.2% reported “other” (N =6). Gender wise, 73.2% were female (N =71), 25.8% were male 18 (N =25), and 1% self-identified as “other” (N =1). Vast majority of participants (90.8%) held degrees of college and above including BA, BS, MA, MS, PhD, and professional graduate degrees (N=88), whereas 6.2% completed high school education (N=6), and 3.1% completed 3-year College (N=3). Recruitment The majority participants were recruited online via Facebook and researcher’s personal email contacts. Some initial participants “shared” the research on their own Facebook page, and some of them forwarded the research to their own email contacts. Recruitment also replied on a convenient sampling. Twenty students from American University’s Crisis Communication class instructed by Prof. Corinne Hoare hand=filled and completed questionnaires during class. In the consent form, the participants were told that the research was aimed at understanding their perception of one or two news stories without being told what specifically the news stories were about. The purpose of the study was debriefed to the participants upon the completion of the questionnaire. Experimental Procedures The fictitious story describing a hypothetical corporation is created based on a real company’s CSR practice information. The corporation is a gas company’s local branch office in the town of “Danville”, which can be seen as any town in the United States. The name of the business, GAS Corp is fabricated. Such a company was chosen because it is a utility company. Most adults can easily identify such a company with their own life experience 19 and could be affected by the actions of such a company. The scenario is written in a news article format. The company’s responses to the crisis in every phase are also written in news article formats. Participants were randomly assigned to one of four conditions in which they read one or two stories corresponding to different types of crisis responses, which are outlined in Table 3. Table 3: Four Conditions of Stimuli in This Study Story 1 Story 2 Condition 1 no comment N/A Condition 2 Denial Excuse Condition 3 Excuse Full apology Condition 4 Full apology Compensation After reading the news stories, participants were asked to answer questions about perceptions toward the responses, how the company treats their customers, the company itself, and its CSR initiative. These items are measured on a 7-point semantic scale ranging from 1 (strongly disagree) to 7 (strongly agree). Except for the first group as control group, all other groups were given a new stimulus as the “company” changed its response to the scenario. Participants in these three groups were asked the same questions again. All participants were asked general questions about how they generally perceive CSR and demographic questions. Questionnaire was available online via Qualtrics from March 20, 2014 to March 30, 2014. It took about 10-15 minutes for each participant to finish the questionnaire. Upon the completion, the participants were thanked and debriefed on the true purpose of this study. 20 Measurement The scale is adapted from previous studies measuring organizational reputation, crisis responsibility (Coombs & Holladay, 1996) and CSR (Tucker, 2009). Some items are adapted from an Organization-Personal Relationship (OPR) study (Kim, 2001). There are four themes as dependent variables. They are: perception of the responses, perception of customer treatment, perception of the company, and perception of the company’s CSR initiative. Control variable is that the company gives no comments. In the theme of perception of responses (3 items), sample items look like this: “I think GAS Corp’s response is honest.” (Coombs & Holladay, 1996) and “GAS Corp’s response can solve its problems effectively.” In the theme of perception of how the company treats its customers (5 items), sample items include: “GAS Corp is concerned with the well-being of its public.” (Coombs & Holladay, 1996) and “I feel customers are important to GAS Corp.” There are 2 items in the theme of perception of the company, they are “I feel GAS Corp has positive corporate culture.” and “GAS Corp represents a good business.” (Kim, 2001) And sample item in the theme of perception of the company’s CSR initiative (3 items) looks like this: “GAS Corp’s social responsibility campaigns tend to do good things for the community.” and “GAS Corp is committed to take social responsible actions.” (Tucker, 2009) In the general questionnaire, all participants were asked control questions (4 items) including “How much do you care to know about community news?” and “I understand the concept of Corporate Social Responsibility.” Also some questions about demographic (4 items) such as gender, ethnicity, age and education. 21 Consent There was a page of text soliciting the consent of each participant at the beginning of the questionnaire. The consent form included 10-15 minute time commitment, stating the data will be confidential because all data will be stored in a secure location and no information can be associated with individual respondents. The consent form also clearly stated that participants are free to withdraw at any time and providing contact information of the research, advisor and the IRB. By choosing to continue the questionnaire, participants agreed with all the information provided in the consent form. Privacy The data collection guarantees confidentiality because there were not any questions asking for names, contact information, employer or anything that might reveal the identity of the participants included. It was not possible to associate the answers with any specific participants. Data storage Data was entered in a personal computer owned by the researcher, and then transferred to a computer solely for analyzing data at American University by a researchonly flash drive. The computers are password-protected and the researcher only accesses the internet on secure network. Data is going to be deleted two months after the submission of the final paper. 22 Ethical concerns This study has no potential ethical concerns or could cause any psychological risks to the participants. Reliability & Validity Experimental research has been used repeatedly by both crisis communication researchers and CSR researchers to measure perceptions of crisis responses, organization, and CSR. In this research, several items are adapted from previous studies including Coombs & Holladay (1996), Kim (2001) and Tucker (2009). Cronbach’s alpha and correlations prove the reliability of all four themes: perception of crisis response (α = 0.78), perception of customer treatment (α = 0.93), perception of organization (correlations = 0.85) and perception of CSR (α = 0.91) The RQ aims to measure participants’ changes of perception towards the company based on different crisis response scenarios. The different phases of responses are based on the level of acceptance in SCCT (Coombs, 2010) including “denial”, “excuse”, “full responsibility” and “compensation”, with “no comment” as a control variable. The researcher ensures the company’s response in the news stories were based on Coombs’s (2010) definitions of each type of crisis response. Prof. Corinne Hoare, instructor of Crisis Communication class at American University confirmed the validity of the crisis responses. 23 Results This study aims to measure the changes of public perceptions of four variables crisis response, customer treatment, organization and CSR, based on the responses the organization (GAS Corp) presented through two news stories. The two phases of news stories for all four treatment groups are shown in Table 4. Table 4: Number of Participants in Four Conditions Company Response Story 1 Story 2 Treatment 1 (N =23) no comment N/A Treatment 2 (N =24) Denial Excuse Treatment 3 (N =26) Excuse Full apology Treatment 4 (N =22) Full apology Compensation Participants answered the same questions after reading each story. Facilitating data analysis, participants’ answers to the second set of questions were cut and pasted below the answers of the first set of questions. All the answers to the same questions were under one column. A variable was created to attribute a nominal value to each of the seven stories. This allowed for comparisons both within the three conditions with two stories, as well as among the seven stories. After testing the reliability of the items measuring each dependent variable, composite scores were created for the dependent variables. They are: perception of crisis response (3 items, α = .78), perception of customer treatment (5 items, α = .93), perception of organization (2 items, r = .85, p <.001) and perception of CSR initiative (3 items, α = .91). 24 One-way ANOVAs were conducted to compare the effect of the company’s responses on the four dependent variables – public perception of response, customer treatment, organization, and CSR initiative between the first and second story in the three two-story conditions. ANOVAs and Tukey’s post hoc tests revealed significant changes in participants’ perceptions in these three conditions between the two stories for all dependent variables. All significant results reported are based on Tukey's post-hoc tests. H1: Participants will display more favorable overall perceptions of the company after reading that the company presented its excuse for an incident than after reading that the company denied being involved in the incident. ANOVA results showed that there was significant change of participants’ perception of crisis response (F(6,160) = 15.16, p < .05). The significant change was also found for participants’ perception of customer treatment (F(6,160) = 9.44, p < .001), for participants’ perception of organization (F(6,161) = 3.36, p < .01), also for participants’ perception of CSR initiative based on individual story (F(6,160) = 3.71, p < .01). Post-hoc results revealed that for the first variable - perception of crisis response, participants reported statistically significant more favorable perceptions of the company after having read the second news story in which the company presented its excuse for the incident (M = 4.47, SD = 0.26) than after having read the first story in which the company denied being involved in the incident (M = 3.26, SD = 0.26). The change was significant (p < .05). For the other three dependent variables – perception of customer treatment, perception of organization and perception of CSR initiative, participants did not display more favorable perception toward the company after having read the second news story in 25 which the company presented its excuse for the incident than after having read the first story in which the company denied being involved in the incident. There was no significance found in these three dependent variables (p > .05). (See Table 5) The results proved that publics perceive a company’s excuse better than a denial in crisis response. However, between the company responding to an incident with an excuse and denying involvement with the incident, public perceptions toward how the company treats its customers, the company itself and the company’s CSR initiative do not change. Therefore, H1 was only partially supported. H2: Participants will display more favorable overall perceptions of the company after reading that the company presented its full apology for an incident than after reading that the company presented its excuse for the incident. Post-hoc results showed that for the first dependent variable – perception of crisis response, participants reported significantly more favorable perception (p < .001) toward the company after having read the second news story in which the company presented its full apology for the incident (M = 5.24, SD = .25) than after having read the first story in which the company presented an excuse for the incident (M = 3.89, SD = .25). The change was significant (p < .05). (See table 6) Participants also displayed more favorability (p < .05) toward the company in the second dependent variable of perception of customer treatment after after having read the second news story in which the company presented its full apology for the incident (M = 4.35, SD= .25) than after having read the first story in which the company presented an excuse for the incident (M = 3.17, SD = .25). 26 In the variables of perception of organization and perception of CSR initiative, there were no significant changes (p > .05) between given full apology and an excuse. The results proved that public perceives a company’s full apology better than excuse in the responses during a crisis, and the difference of these two responses make public perceive differently on how the company treats its customers. Specifically, public perceive that the company treats its customer better after being given a full apology than being given an excuse to an incident. However, between the company responding to an incident with full apology and an excuse, public perceptions toward how the company itself and the company’s CSR initiative still do not change. This hypothesis was also partially supported. Table 5: Comparisons of Condition 2 (Denial – Excuse) Dependent Variable 1: perception of crisis response 2: perception of customer treatment 3: perception of organization 4: perception of CSR initiative Denial M = 3.26, SD = .26 M = 2.61, SD = .26 M = 3.02, SD = .28 M = 3.13, SD = .28 Excuse M = 4.47, SD = .26 M = 3.63, SD = .26 M = 3.63, SD = .28 M = 3.49, SD = .28 p Full Apology M = 5.24, SD = .25 M = 4.35, SD = .25 M = 3.98, SD = .26 M = 4.19, SD = .26 p .020 .099 .711 .967 Table 6: Comparisons of Condition 3 (Excuse – Full Apology) Dependent Variable 1: perception of crisis response 2: perception of customer treatment 3: perception of organization 4: perception of CSR initiative Excuse M = 3.89, SD = .25 M = 3.17, SD = .25 M = 3.33, SD = .27 M = 3.36, SD = .26 .003 .020 .582 .285 Table 7: Comparisons of Condition 4 (Full Apology - Compensation) Dependent Variable 1: perception of crisis response 2: perception of customer treatment 3: perception of organization 4: perception of CSR initiative Full Apology M = 4.70, SD = .27 M = 4.00, SD = .27 M = 4.00, SD = .28 M = 3.77, SD = .28 27 Compensation M = 5.24, SD = .27 M = 4.69, SD = .28 M = 4.43, SD = .29 M = 4.48, SD = .28 p .788 .580 .934 .558 H3: Participants’ overall favorability towards the company after reading that the company provided compensation in an incident and after reading that the company presented its full apology for the incident will display no significant difference. Results indicated that participants’ favorability toward the company in all four dependent variables after having read the company compensated the victims and having read being given full apology were not significant enough (p > .05) to be different, even though the mean of compensation was slightly higher than full apology. Therefore, this hypothesis is validated. (See Table 7) H4: After reading the first story, participants who read the response of full apology will display the highest acceptance comparing with participants who read the responses of giving no comment, denial and excuse. Post-hoc results showed that participants displayed more favorable perception (p < .05) toward the company after having read the second story where the company presented full apology (M = 5.24, SD = .25) or compensation (M = 5.24, SD = .27) than having read the first story where the company made no comment (M = 2.54, SD = .27), denied (M = 3.26, SD = .26) or gave an excuse (M = 3.89, SD = .25) for the incident. Similar results were also found in the second dependent variable – perception of customer treatment. Participants also displayed more favorable perception (p < .05) toward the company after having read the second story where the company presented full apology (M = 4.35, SD = .25) or compensation (M = 4.69, SD = .28) than having read the first story where the company made no comment (M = 2.60, SD = .27), denied (M = 2.61, SD = .26) or gave an excuse (M = 4.00, SD = .27) for the incident. 28 However, full apology in the second story in both the variable of perception of organization (M = 3.98, SD = .26) and variable of perception of CSR initiative (M = 4.19, SD = .26) had insignificant effects after making no comment (M = 3.17, SD = .28 and M = 3.07, SD = .29), denial (M = 3.02, SD = .28 and M = 3.13, SD = .28) and giving an excuse (M = 3.33, SD = .27 and M = 3.36, SD = .26). Instead, in both of these two variables, participants’ perceptions showed significant changes (p < .05) only when the company compensated the victims in the second story (M = 4.43, SD = .29 and M = 4.48, SD = .28) or after having read that the company made no comment (M = 3.17, SD = .28 and M = 3.07, SD = .29) or denied (M = 3.02, SD = .28 and M = 3.13, SD = .28). Full apology in the second story apparently did not make significant changes in participants’ perceptions as expected. Also in both of these two dependent variables, neither full apology nor compensation had a significant impact in the second story after the company presented an excuse in the first story. 29 Discussion This study explores changes in public perceptions of an organization based on twophase responses given by the organization during a crisis. Both Dean (2004) and SCCT stress that three factors affect the reputational threat an organization faces during a crisis: crisis responsibility, reputation prior to the crisis and the organization’s history of crises. This study shows that public perceptions of an organization embroiled in a two-phase crisis are significantly affected by the organization’s previous response in the earlier phase of the crisis. This is a new factor affecting the reputational threat an organization faces that Dean (2004) and the SCCT do not address. Specifically, participants better accepted an excuse only after they read about a denial; participants better accepted a full apology only after they read about an excuse. Interestingly, when participants were given different responses simultaneously after reading about the incident, the response of full apology did not necessarily have the highest acceptance. In terms of measuring the level of public perception of crisis response and customer treatment, giving an excuse was just as sufficient as a full apology. This may have occurred because in this particular scenario, the company’s crisis responsibility is relatively low. Full apology also made a difference in comparison to any other response during the first phase of the crisis in terms of getting better public perception of the organization and its CSR initiative. When the company followed an initial response of denial by offering an excuse, participants seemed to accept the excuse better than the initial denial. This validates SCCT’s acceptance levels of response strategies between a denial and an excuse; excuse does have a higher level of acceptance than denial. 30 However, this study separated public perceptions into four variables, including perceptions of crisis response, of customer treatment, of the organization and of the CSR initiative. Excuse resulting in higher levels of acceptance than denial was only valid for the first variable, perceptions of crisis response. It did not have a significant effect on any other variables. In other words, even though the public seems to be more accepting of an excuse than of a denial, this higher acceptance only improves public perceptions of crisis response. It has no effect on public perceptions of the organization as a whole. Similarly, according to SCCT, the public is more accepting of a company’s full apology than of its excuse. This study’s results indicate that full apology yields a higher level of acceptance only with regard to the variables of perceptions of crisis response and perceptions of customer treatment. It did not affect the other two variables, perceptions of the organization and perceptions of its CSR initiative. In other words, compared with an excuse, full apology does lead the public to be more accepting of the company’s response, and it does make the public believe that they are treated better as customers by the organization. Full apology, however, does not improve public perceptions of the organization as a whole or its CSR initiative. Between responding with full apology and compensation, the experiment found that after the company responded with a full apology in the first phase of the crisis, responding in the second phase with compensation makes no further improvement to the public’s overall perceptions of the organization. This could have two interpretations. On the one hand, compensation in this particular scenario was specifically reimbursing victims’ medical expenses. This action might not have directly impacted study participants 31 themselves like full apology did. Therefore, participants may not “care” enough about whether or not the company made compensation to result in a significant change in their perceptions. On the other hand, this result may also suggest that publics strongly value organizations admitting culpability for the genesis of a crisis at the outset, and that spending money on compensating the public for the crisis may be an unnecessary further step in terms of improving public perceptions of the organization. The results also showed that, compared to the low acceptance responses of no comment and denial, the only response having a significant impact on perceptions was compensation. This finding is an interesting expansion on SCCT. While SCCT posits that the no comment and denial have the lowest level of acceptance, it does not suggest how organizations might change public perceptions after making these responses. The results in this study indicate that in order to significantly improve public perceptions of the organization or its CSR initiative after giving no comment or denial, an organization must let the public know that it took at least some responsibility for the crisis and compensated victims. The results expanded upon SCCT in one further way. SCCT studies generally focus on and measure public perceptions of crisis response strategies. As mentioned before, this study separated public perceptions into four themes. Each theme was measured for the organization’s different responses. The results of the experiment indicate that SCCT’s different levels of acceptance are limited to measuring the acceptance of crisis responses and are not able to measure overall perceptions of an organization. Separating perceptions into four dimensions provided a more comprehensive picture of how the public perceives crisis responses. 32 Conclusions This study provides strong evidence that SCCT response strategies are valid in multi-stage crises, but certain results also indicate important limitations to SCCT’s utility. First, when the organization enters a crisis with a previously good public reputation or is believed by the public to have had good intentions, responding to a crisis with an excuse can be as effective as a full apology or compensation in terms of maintaining positive public perceptions of the organization. While this may seem like a blinding flash of the obvious, the findings of this study imply that organizational crises do not happen in isolation. Rather, organizations’ public reputations prior to a crisis directly affect public perceptions of the organization during a crisis and influence the effectiveness of organizations’ crisis response strategies. By correlation, an organization can improve its chances of weathering a crisis by building and maintaining a positive public reputation when it is not in crisis. In this study, CSR may have represented the company’s positive intention and influenced participants’ presumption of the company’s good public reputation. Participants tended not to assume the worse of the company during a crisis due to its positive pre-crisis reputation, which is the “halo effect” CSR brought to the company (Coombs & Holladay, 2006). Second, SCCT holds that both no comment and denial responses are the least acceptable responses companies can offer the public. This study found that when the organization made no comment or denied involvement, the only response it could make in the second stage of the crisis that would alter public perceptions of it was compensation. Participants did not seem to accept a full apology if the company didn’t compensate the victims. This finding implies that no comment and denial response strategies can be potentially very costly gambits for organizations. These two strategies appear to be so 33 unacceptable to the public that companies using them early on could be left in later phases of a crisis with few options other than full acceptance of liability, which makes them vulnerable to costly litigation, and financial compensation. In other words, this finding provides a more concrete idea of the low level of acceptance of no comment and denial response strategies. Third, concerning public perceptions of an organization’s treatment of its customers, the effectiveness of responding with an excuse is as low as responses of no comment or denial. An excuse may be effective in terms of salvaging the organization’s general reputation in a crisis. When considering this type of response in light of how this organization treats its customers though, participants felt that an excuse indicated that the organization does not treat its customers well. Based on the crisis scenario in this experiment, the only way for an organization to positively influence participants’ perceptions of customer treatment remains for the organization to offer a full apology and/or compensate victims. Fourth, when a company responds with a full apology in the first phase of the crisis, compensation makes no further improvement to public perceptions towards the organization. Full apology directly affects the public, whereas compensation gives money to victims only and does not directly affect the general public. Recommendations The results of this study offer several interesting recommendations for crisis managers. First, if the organization has given a full apology to the crisis, then compensation 34 does not seem to play an important role in terms of improving the public’s overall perceptions toward the organization. Of note, in the scenario in this particular study, even though compensating the victims does not change the public’s perception, it does not necessarily mean that crisis managers should not cover the basic expenses of victims for legal, ethical or other reasons. Not doing so could needlessly prolong the crisis. Second, stakeholders and the public can be very forgiving to organizations in crisis when the organization is has a good public reputation prior to the crisis. For such organizations, the public would seem to accept a logical excuse as a reasonable crisis response. Therefore, the organization does not really need to apologize. Since apology is in part an admission of guilt for the incident and can be used as evidence against an organization in lawsuits, this finding can help crisis managers avoid litigation resulting from making unnecessary apologies. Third, not commenting on a crisis or denying its existence are strongly not recommended crisis response strategies. As Coombs (2007b) indicated, an organization is in a crisis if the public perceives an incident as a crisis. This means that the organization must somehow acknowledge that a crisis exists if it hopes to salvage its public image. Furthermore, based on the results of this study, giving these responses at the onset of a crisis will likely force an organization to make more costly responses than a simple full apology, such as giving compensation, in later stages to change the public’s perception. Finally, Table 8 outlines recommendations for second phase response strategies based on the results of this experiment concerning public perceptions of an organization’s initial crisis response. 35 Table 8: Recommended Second Phase Responses based on Results of This Study DV1: perception of crisis response DV2: perception of customer treatment DV3: perception of organization DV1: perception of CSR initiative Phase 1 Acceptance Recommended Response for Phase 2 No comment low Excuse Denial low Excuse Excuse mild Full Apology Full apology high N/A No comment low Compensation Denial low Compensation Excuse mild Full apology Full apology high N/A No comment low Compensation Denial low Compensation Excuse mild N/A Full apology high N/A No comment low Excuse Denial low Compensation Excuse mild N/A Full apology high N/A Implications As a rare experiment designed to measure changes in crises featuring two phases, this study considers realistic crises that commonly involve a series of evolving responses in order to examine the relationship between public perceptions of an organization and its 36 crisis responses. The results offer general support for Coombs’s levels of acceptance to organizations’ crisis responses (Coombs, 2010). The significant changes between stories in the variable of perception of crisis response show that the acceptance of an “excuse” is higher than a “denial”, a “full apology” is higher than an “excuse”. “No comment” and “denial” have the lowest acceptance, while there is no acceptant change between “full apology” and “compensation”. In certain circumstances brought on by the initial crisis response, however, these response strategies do not generate levels of public acceptability as described above. In this sense, the results of this study expand on SCCT, confirming or explaining some phenomena that SCCT research has not yet addressed. For example, SCCT explains that excuse yields mild acceptance by the public. This study found that the effectiveness of an excuse depends on the timing and purpose of the strategy. If the organization is believed to have good intentions going into a crisis, an excuse can be very effective in recuperating the organization’s reputation. If the organization initially responded with no comment or denied any involvement in the incident, then an excuse will not be effective. Also, according to SCCT, full apology and compensation have similar effects on public perceptions of an organization. This study found that there are differences in their effectiveness, but that these do not show until after an organization has initially offered the public crisis responses with the lowest levels of acceptance. When this occurs, compensation can significantly improve the public’s perception while full apology cannot. More broadly, this study proves that organizational crisis reflect the reflexive communication inherent in the supply–demand relationships organizations have with their 37 stakeholders and the public. In focusing designing organizational responses to static, single-phase crises, previous crisis communications studies and theories appear to assume that crisis responses are transmissions to the public, that crises remain static, and that an organization need only concoct a single response strategy and stick to it. In countless real world cases though, crises often have several phases, and organizations have to respond at each phase as the crisis – and the public’s perceptions of the organization – changes. Limitations In the responses to the given crisis scenario, this study entails two stages of changes in responses, which are designed to measure the participants’ change in perceptions. In reality, there might be several stages of changes of responses that they can cause the public change perceptions several times and eventually become unpredictable. This research provides a simple direction of how the public might change their perception according to the responses, but not necessarily concrete explanation of what the perception will be when the crisis is over. Another limitation of this study is that the company created in the scenario is fictional. The participants do not have any personal relationship or emotional ties to the company except those they imagine as a result of the fictional scenario. In this situation, participants are less likely to respond as they would if this were a real crisis having consequences for them or those close to them. A third limitation to this study is that the study does not measure the effects of time on participant responses. An actual crisis can often play out over a matter of days, weeks, or months, and the duration of the crisis may have its own effects on public perceptions of 38 the organization in crisis. By contrast, this study’s crisis scenario is played out over the length of time it takes a participant to read the stimuli and answer their questionnaire, ie 10-15 minutes. As a result, the responses of participants in this study are not affected by the duration of the crisis. Last limitation exits in the demographic of the participants of this research. Within all 97 participants who completed the questionnaire, there were 71 female (73.2%), 70 Caucasian (72.2%), and 88 participants held degree of college and above (90.8%). The large percentage of highly educated Caucasian female participants dominated the results. Therefore, the results of this study may be biased and more likely to represent the opinions of people in this particular demographic. More diverse participants are needed for future researches. Despite the above limitations, this study contributes to the understanding public perceptions of organization during crises. Specifically, its main contribution is its experimental design using two-phase crisis responses. Participants perceive two different responses at two different times about the same organization during the same crisis. By taking different phases of crisis responses into account, identifying the changes of public perception and quantifying their effects, this study expands on existing crisis communication theories and adds practical knowledge for dealing with real-world crises. Future Research The scenario designed in this study is a CSR-related crisis. CSR is perceived as the organization tends to do good things to the community (McWilliams & Siegel, 2001). The public could be influenced by this existing perception and become more “forgiving” to the 39 organization during the crisis. For future studies, researchers could design a scenario that is not CSR related to examine how the public perceives the organization through its crisis response strategies. One of this study’s limitations is that it does not measure the effects of time on participant responses. Future studies could reproduce more realistic crisis responses with a few days’ interval between responses. Finally, this study was designed to study public perceptions across a two-phase crisis during which organizations issued two responses. Future studies might consider adding further phases to their crisis scenarios to further enhance the realism of the experiment. 40 References Allen, M. W., and Caillouet, R. H. (1994) Legitimation Endeavours: Impression Management Strategies Used by an Organization in Crisis. Communication Monographs, 61, 44-62. Andrews, A. P., Simon, J., Tian, F., & Zhao, J. (2011). The Toyota crisis: an economic, operational and strategic analysis of the massive recall. Management Research Review, 34 (10), 1064-1077. Batruch, C. (2011) Does Corporate Social Responsibility Make a Difference? Global Governance, 17 (2), 155-159. BBC (March 24, 2014a). Flight MH370 'crashed in south Indian Ocean' - Malaysia PM. Retrieved from http://www.bbc.com/news/world-asia-26716572 BBC (March 25, 2014b). Malaysia Airlines MH370: Relatives in Beijing scuffles. Retrieved from http://www.bbc.com/news/world-asia-26728045 BBC (April 11, 2014c). The search for flight MH370. Retrieved from http://www.bbc.com/news/world-asia-26514556 Benoit, W. L. (1995). Sears' repair of its auto service image: Image restoration discourse in the corporate sector. Communication Studies, 46 (1-2), 89-105. Benoit, W. L. (1997). Image Repair Discourse and Crisis Communication. Public Relations Review, 23 (2), 177-186. Bergman, E (1994). Crisis? What Crisis? Communication World, 11 (4), 9-13. Bowen, H. R. (1953). Social responsibilities of the businessman. New York: Harper & Row. Carroll, A. B. (1991). The pyramid of corporate social responsibility: Toward the moral management of organizational stakeholders. Business Horizons, 34 (4), 39-48. 41 Cherry, M. A. & Sneirson, J. F. (2011) Beyong Profit: Rethinking Corporate Social Responsibility and Greenwashing after the BP Oil Disaster, Tulane Law Review, 85, 983-1038. CNN (2012). Top Companies: Biggest Employees. Retrieved from http://money.cnn.com/magazines/fortune/global500/2012/performers/companie s/biggest/ Coombs, W. T. & Holladay, S. J. (1996). Communication and Attributions in a Crisis: An Experimental Study in Crisis Communication. Journal of Public Relations Research, 8 (4), 279-295. 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. Coombs, W. T. (2006). The Protective Powers of Crisis Response Strategies: managing Reputational Assets during a Crisis. Journal of Promotion Management, 12 (3/4), 241-260). Coombs, W. T. & Holladay, S. J. (2006). Unpacking the halo effect: reputation and crisis management. Journal of Communication Management, 10 (2), 123-137. Coombs, W. T. (2007a) Protecting Organization Reputations during a Crisis: The Development and Application of Situational Crisis Communication Theory. Corporate Reputation Review, 10 (3), 163-176. Coombs, W. T. (2007b). Ongoing Crisis Communication: Planning, managing, and responding. (2nd ed.). Thousand Oaks, CA: Sage. 42 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. Coombs, W. T. (2010). Crisis Management: A Communicative Approach. In Botan C. & Hazleton V. (Ed.) Public Relations Theory 2 (pp. 171-197). Taylor & Francis. CSR Europe. Reports and Guides. Retrieved from http://www.csreurope.org/searchreports Dean, D. H., (2004). Consumer Reaction to Negative Publicity: Effets of Corporate Reputation, Response, and Responsibility for a Crisis Event. Journal of Business Communication, 41 (2), 192-211. Dionisopolous, G. N. and Vibbert, S. L. (1988). CBS vs Mobil Oil: Charges of Creative Bookkeeping. In H. R. Ryan (Ed.), Oratorical Encounters: Selected Studies and Sources of 20th Century Political Accusation and Apologies, 214-252. Frederick, W. C. (1960). The growing concern over business responsibility. California Management Review, 2, 54-61. Friedman, M. (1970). The Social Responsibility of Business is to Increase Its Profits. New York Times Magazine 32-33, 122-124. Graham, K. & Greenberg, J. (2002) Promotionalism and Subpolitics, Management Communication Quarterly 15 (4), 541-570. Heald, M. (1970). The social responsibilities of business: Company and community, 1900-1960. Cleveland, OH: Case Western Reserve University Press. Hearit, K. M. (1994) Apologies and Public Relations Crises at Chrysler, Toshiba, and Volvo. Public Relations Review, 20 (2), 113-125. 43 Hearit, K. M. (1995) From “We Didn’t Do It” to “It’s Not Our Fault”: The Use of Apologia in Public Relations Crises. In W. N. Elwood (Ed.), Public Relations Inquiry as Rhetorical Criticism: Case Studies of Corporate Discourse and Social Influence. Westport, CT: Praeger. Hobbs, J. D. (1995) Treachery by another Name: A Case Study of the Toshiba Public Relations Crisis. Management Communication Quarterly 8, 323-346. Ice, R. (1991) Corporate Publics and Rhetorical Strategies: The Case of Union Carbide’s Bhopal Crisis. Management Communication Quarterly 4, 341-362. Kim, Y. (2001). Search for the Organization-Public Relationship: A Valid and Reliable Instrument, Journalism and Mass Communication Quarterly 78 (4), 799-815. Klein, J. & Dawar, N. (2004). Corporate Social Responsibility and Consumers’ Attributions and Brand Evaluations in a Product-Harm Crisis, International Journal of Research in Marketing, 21 (3), 203-217. Jin, Y., Liu, B. F. & Austin, L. L. (2011). Examining the role of social media in effective crisis management: the effects of crisis origin, information form, and source on publics' crisis responses. Communication Research, 41 (1), 74-94. McGuire, J. W. (1963). Business and society. New York: McGraw-Hill. McWilliams, A., & Siegel, D. S. (2001). Corporate Social Responsibility: A Theory of the Firm Perspective. Academy of Management Review, 26 (1), 117-127. Moir, L. (2001). What do we mean by corporate social responsibility? Corporate Governance, 1 (2), 16-22. Motwani, S. (2012). Communicating CSR is More Challenging than Paying CSR, Institute of Management Studies 1 (1), 41-45. 44 Pfau, M., Haigh, M. M., Sims, J., & Wigley, S. (2008). The Influence of Corporate Social Responsibility Campaigns on Public Opinion. Corporate Reputation Review, 11 (2), 145-154. Sturges, D. L. (1994). Communicating Through Crisis: A Strategy for Organizational Survival, Management Communication Quarterly, 7 (3), 297-316. Tucker, D. (2009). Measuring Corporate Social Responsibility: A Scale Development Study, Journal of Business Ethics 85 (4), 411-427. Wal-Mart. Global Responsibility Report. Retrieved from http://corporate.walmart.com/global-responsibility/environmentsustainability/global-responsibility-report Xu, K. B. & Li, W. Q. (2013), An Ethical Stakeholder Approach to Crisis Communication: A Case Study of Foxconn’s 2010 Employee Suicide Crisis, Journal of Business Ethics 117 (2), 371-386. 45 Appendices Stimuli Condition 1 Since Sunday, 19 people, including children and seniors, have been hospitalized after drinking water that contained a flavoring agent called “Organic Water Flavoring Agent”, which is produced by Happy Earth Food Inc, but distributed freely to them by gas company GAS Corp during GAS Corp’s weekend outdoor event as part of the company’s Healthy and Sustainable Living Campaign. Common symptoms they experienced include severe stomach cramps, nausea, fever, hallucinations, and loss of consciousness. GAS Corp has been the primary gas supplier in the region of Danville for more than 50 years. A series of events of the company’s Healthy and Sustainable Living Campaign involve several local businesses as vendors and partners, including Happy Earth Food Inc. GAS Corp so far has not yet commented on the incident. Local newspaper Danville Today was unable to reach company representatives despite several efforts to make contact. On Tuesday, two more people were hospitalized after drinking tap water containing the flavoring agent bringing total of those hospitalized to 21. 46 Condition 2 (Story 1) Since Sunday, 19 people, including children and seniors, have been hospitalized after drinking water that contained a flavoring agent called “Organic Water Flavoring Agent”, which is produced by Happy Earth Food Inc, but distributed freely to them by gas company GAS Corp during GAS Corp’s weekend outdoor event as part of the company’s Healthy and Sustainable Living Campaign. Common symptoms they experienced include severe stomach cramps, nausea, fever, hallucinations, and loss of consciousness. GAS Corp has been the primary gas supplier in the region of Danville for more than 50 years. A series of events of the company’s Healthy and Sustainable Living Campaign involve several local businesses as vendors and partners, including Happy Earth Food Inc. “After an initial investigation, we found no direct relationship between the product we distributed and these people’s illness.” said Lisa Blackburn, a spokeswoman for GAS Corp, “The allegation is baseless and absurd.” On Tuesday, two more people were hospitalized after drinking water containing the flavoring agent bringing total of those hospitalized to 21. (Story 2) The toll of those who feel after drinking tap water containing Organic Water Flavoring Agent rose as of today to 25. 47 GAS Corp, which initially blamed their vendor with the incident, today responded publicly to these new developments. “Further investigation suggested that an ingredient contained in the water flavoring agent could have caused people’s sickness. It is not possible for any of us to anticipate an incident like this while we believed that we were choosing the right partner and the right product as part of our healthy and sustainable living campaign. GAS Corp was unaware, and we are as surprised and horrified as all members of our community.” Condition 3 (Story 1) Since Sunday, 19 people, including children and seniors, have been hospitalized after drinking water that contained a flavoring agent called “Organic Water Flavoring Agent”, which is produced by Happy Earth Food Inc, but distributed freely to them by gas company GAS Corp during GAS Corp’s weekend outdoor event as part of the company’s Healthy and Sustainable Living Campaign. Common symptoms they experienced include severe stomach cramps, nausea, fever, hallucinations, and loss of consciousness. GAS Corp has been the primary gas supplier in the region of Danville for more than 50 years. A series of events of the company’s Healthy and Sustainable Living Campaign involve several local businesses as vendors and partners, including Happy Earth Food Inc. 48 “Our investigation suggested that an ingredient contained in the water flavoring agent could have caused people’s sickness.” said Lisa Blackburn, a spokeswoman for GAS Corp, “It is not possible to anticipate an incident like this while we believed that we were choosing the right partner and the right product as part of our healthy and sustainable living campaign. GAS Corp is unaware, and we are as surprised and horrified as all members of our community.” On Tuesday, two more people were hospitalized after drinking tap water containing the flavoring agent bringing total of those hospitalized to 21. (Story 2) The toll of those who feel after drinking tap water containing Organic Water Flavoring Agent rose as of today to 25. GAS Corp, which initially claimed of also being “unaware”, today responded publicly to these new developments. “We take full responsibility for this incident and we apologize to the families involved and the public. GAS Corp is in contact with the families of those affected by our free samples to make individual arrangements for their medical needs. GAS Corp has discontinued its partnership with Happy Earth Food Inc. effective immediately. We are also doing everything in our power to prevent a further incident from happening.” 49 Condition 4 (Story 1) Since Sunday, 19 people, including children and seniors, have been hospitalized after drinking water that contained a flavoring agent called “Organic Water Flavoring Agent”, which is produced by Happy Earth Food Inc, but distributed freely to them by gas company GAS Corp during GAS Corp’s weekend outdoor event as part of the company’s Healthy and Sustainable Living Campaign. Common symptoms they experienced include severe stomach cramps, nausea, fever, hallucinations, and loss of consciousness. GAS Corp has been the primary gas supplier in the region of Danville for more than 50 years. A series of events of the company’s Healthy and Sustainable Living Campaign involve several local businesses as vendors and partners, including Happy Earth Food Inc. “After an initial investigation, toxic elements were found in the water flavoring agent,” said Lisa Blackburn, a spokesperson for GAS Corp “We take full responsibility for this incident and we apologize to the families involved and the public. GAS Corp is in contact with the families of those affected by our free samples to make individual arrangements for their medical needs. GAS Corp has discontinued its partnership with Happy Earth Food Inc. effective immediately. We are also doing everything in our power to prevent a further incident from happening.” On Tuesday, two more people were hospitalized after drinking tap water containing the flavoring agent bringing total of those hospitalized to 21. 50 (Story 2) The toll of those who feel after drinking tap water containing Organic Water Flavoring Agent rose as of today to 25. GAS Corp, which has taken full responsibility for the incident, today responded publicly to these new developments. “We once again sincerely apologize to these families and the public. GAS Corp has made arrangement with Town Hospital and the involved families to pay for all medical expenses of those stricken by this unfortunate incident. We are also in talks with the affected families to arrange for financial compensation to lessen the burden of their troubles. GAS Corp has never faltered before in its commitment to improving our community’s healthy and sustainable living. Going forward, we will strive to strengthen our commitment to health and well-being of our community.” 51 News Story Questionnaire (Perception of Response) 1. I think the message in GAS Corp’s response are clear. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 2. I think GAS Corp’s response is honest. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 3. GAS Corp’s response can solve its problems effectively. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree (Perception of Customer Treatment) 4. GAS Corp treats its customers fairly and justly. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 5. I can see myself becoming a customer of GAS Corp. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 6. GAS Corp is concerned with the well-being of its public. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 7. I feel customers are important to GAS Corp. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 8. If a customer has problem with GAS Corp, I am confident that GAS Corp will try its best to find solution. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree (Perception of Organization) 9. I feel GAS Corp has positive corporate culture. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 52 10. GAS Corp represents a good business. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree (Perception of Corporate Social Responsibility) 11. GAS Corp is committed to take social responsible actions. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 12. GAS Corp’s social responsibility campaigns tend to do good things for the community. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 13. GAS Corp plays an important role in building/maintaining the community. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree (Perception of the New Story) 14. Did you find the news article above was_______ Very difficult to read ○ ○ ○ ○ ○ ○ ○ Very easy to read 15. Did you understand the news article above? Didn’t understand at all ○ ○ ○ ○ ○ ○ ○ completely understood 53 General Questionnaire 1. Your gender: female; male; other 2. Your ethnicity: Asian/Pacific Islander; Caucasian; Black; Hispanic; Native American; Other:_______________ 3. Your age: ______________ 4. Your highest diploma achieved: High school; 3-year College; BA or BS; MA or MS; Professional graduate degree (Ex: JD, MBA); PhD 5. How much are you involved with your community volunteerism? Not involved at all ○ ○ ○ ○ ○ ○ ○ Very involved 6. How much do you care to know about community news? Do not care at all ○ ○ ○ ○ ○ ○ ○ Care very much 7. I understand the concept of “corporate social responsibility”. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 8. Corporate Social Responsibility is a good practice for a corporation. Strongly disagree ○ ○ ○ ○ ○ ○ ○ Strongly agree 54