Strategic Ambiguity and Crisis Apologia: The Impact of Audiences' Interpretations of Mixed Messages

ABSTRACT

Based on the perspectives of strategic ambiguity and organizational reputation, the current study examines the effects of mixed crisis response strategies, which adopt seemingly contradictory messages (i.e., apology and denial), through experiments. Consistent with the scope of strategic communication research, this study incorporates theoretical aspects of distinct areas of organizational communication to examine audience response to strategic messages, and makes recommendations for organizational communication strategies during crisis situations. The findings demonstrate that, instead of taking messages straightforwardly, people interpret the same messages divergently in their own ways, and these interpretations accordingly affect their attitudes and behavioral intentions. Findings indicate participants choose a dominant interpretation when given mixed messages, and subsequent responses are based on the initial interpretation, such that evaluating a mixed message as an apology yielded more positive outcomes than those who interpreted the message in other ways. In addition to people’s diverse interpretations, organizations’ crisis communication strategies and the business type also significantly influenced the outcomes. The apology-interpreters showed more positive outcomes than those who were exposed only to apology for an automaker’s crisis. On the other hand, for a nonprofit organization’s crisis, those who were exposed to a simple denial message showed more positive outcomes than those who perceived the mixed message as a denial. Based on these findings, this study offers practical recommendations on when to use single messages versus mixed messages, along with the explanation of how these divergent strategies work.

Notes

¹ These two dimensions have been proposed by many other authors using different terminology, such as economic performance and social responsibility (Etzioni, 1988), economic performance and social conduct (Chew, 1992), business competency and social conscience (Goldberg, 1999), organizational effectiveness and social performance (Riahi-Belkaoui & Pavlik, 1992), and business and social reputation (De Castro, Lopez, & Saez, 2006). In a similar vein, Mahon (2002) claimed that the marketplace of goods and services and the marketplace of ideas are two markets where “the interactions between the firm and its stakeholders take place” (p. 417). Even though the terms are slightly different, the core ideas do not differ that much: one trait points to a company’s competence in delivering its products and services and the other trait reflects the stakeholders’ sense of a firm’s social obligations.

² Before combining the data of Wave 1 and Wave 2 for the hypothesis tests, the two waves were compared on demographics, manipulation check and the outcome variables by utilizing SPSS v.22. Even though the two waves utilized different methods of data collection (i.e., pencil-and-paper versus computerized methods), accumulated evidence so far suggests that findings based on traditional methods are consistent with online-based findings on the effects of age and gender, personalities, etc. (Gosling, Vazire, Srivastava, & John, 2004). The two waves differed only in terms of age and political affiliations; but no difference was found on other demographic variables, nor on the manipulation checks (p > .05). In addition, the two waves were compared on the major outcome variables by using t-tests. A series of statistical tests found no meaningful differences between the two wave groups on the outcome variables (p > .05). Therefore, from the statistical front, the two groups are assumed to come from the same population as far as the current article’s scope is related. Therefore, the two groups are deemed not much different from each other and can be combined into a single group for the hypothesis tests. In addition, all the participants were randomly assigned to each of the treatment groups, thus, theoretically canceling out the influences of potential confounding factors. As expected, a series of random assignment tests detected no significant difference between treatment groups in terms of the proportion of the two waves, manipulation checks, message effects (i.e., perceived crisis severity and responsibility) and demographics (p > .05). This random assignment check supported the belief that all treatment groups had similar conditions for comparison.

³ These descriptions exclude missing values: one missing for gender; eight for race; and two for education.

⁴ The two organizations were selected as the manipulation because they are expected to represent the CA- and CSR-core businesses, respectively. From the organizations’ perspective, protecting their reputation for a high product quality is deemed more important than that for business ethics in nature for a technology-centered, heavy-industry manufacturers like automakers, whereas, as for a nonprofit organization, protecting its integrity and meeting the acceptable social norms are essential for fulfilling its missions.

⁵ The CA reputation was measured by asking “This organization would offer high quality products and services” and “This organization seems competent and effective in providing its products and services.” The CSR reputation was measured with the questions of whether the organization would “support good causes,” “be honest,” “be sincere and genuine,” and “behave ethically.”

⁶ The treatment groups were exposed to either apology or denial messages, whereas the PA (perceived-apology) or PD (perceived-denial) groups were exposed to mixed messages of both apology and denial, but interpreted them either apology or denial messages.

⁷ According to the cue-diagnosticity approach, a failure is attributable to many factors, including lack of motivation and fatigue, with ability categories. On the other hand, with morality categories, positive behaviors can be attributed to many factors, including conformity and ingratiation, making a piece of positive information less diagnostic. Such being the case, negative information is more indicative of morality, whereas positive information is more indicative of ability. For more information, refer to Wojciszke (1997, 2005).