The Analysis of Persuasive Messages in Crisis Communication
Master’s ThesisAuthor: Bianka Narwanova Student ID: 11093838 Supervisor: Dr. Hanneke Hendriks Master’s programme Communication Science
Word count: 7619
Academic year 2016-2017 Date: 03/02/2017
University of Amsterdam Graduate School of Communication
Nieuwe Achtergracht 166 1001 NG Amsterdam
2 Abstract
Corporate reputation is a collective judgement of company’s attributes (Barnett et al., 2006; Gray & Balmer, 1998). Companies need to establish and maintain a certain image among their customers. This brand image can be affected by corporate crises which involve sudden and unexpected events posing financial and reputational threats to the organization (Coombs, 2007). Crises require a form of strategic communication in order to restore the corporate reputation (Coombs & Holladay, 2002). The SCCT can be used to anticipate and predict the reactions of consumers towards a corporate crisis (Coombs, 2007). The purpose of this master’s thesis is to analyze the persuasive messages used by companies in press releases concerning corporate crisis. The aim is to investigate whether the occurrence and frequency of rational and emotional persuasive messages is dependent on the company industry type and the type of corporate crisis at hand. The research method used for this study is a content analysis of persuasive messages. The hypothesis testing did not fully confirm any of the expected effects for the industry and crisis types. The only significant differences were found in the frequency of rational messages for technological and financial industries which used more rational messages than fashion and food & beverage companies. Additional analyses showed significant differences in the frequency of emotional messages for the technological industry which used less emotional terms compared to the financial industry. The analysis also showed that companies use more emotional messages to address privacy crises compared to financial crises. Based on this research we can see that companies use different persuasive messages to address corporate crises.
3
Table of Contents
Introduction ... 4
Theoretical background ... 5
Crisis communication ... 7
Situational Crisis Communication Theory ... 10
Persuasion and persuasive messages ... 11
The present study ... 13
Hypotheses... 13 Sample ... 16 Method ... 14 Variables ... 16 Results ... 17 Hypothesis testing ... 18 Additional analyses... 23 Discussion ... 26 References... 31 Appendix 1. Codebook ... 33
Appendix 2. Coder instructions ... 35
Appendix 3. Frequency of persuasive terms... 36
4 Introduction
Organizations and corporations create and sell products and services to their customers. The customers form impressions about these companies based on their actions within the public domain (Gotsi & Wilson, 2001). Corporate reputation can be shaped by the information presented by the company and other sources, such as media channels (Coombs, 2007; Ihator, 2001). Good reputation is important to ensure customer preference and support (Coombs & Holladay, 2002). At times, companies can experience troubling events such as errors in manufacturing, causing pollution, or exploitation of employees. Consequently, such issues can turn into crises among consumers through media coverage (Coombs, 2007; Gotsi & Wilson, 2001). Crisis situations might be difficult to handle by organizations trying to avoid damage to their corporate reputation. The challenge is to act appropriately, take responsibility, and offer explanations that will satisfy the consumers (Coombs, 2007).
A lot of corporate crises are often broadcasted by news channels which point out the wrongdoings of the involved companies. Such negative information can harm the attitude of consumers toward the brand (Coombs, 2007). In order to maintain their image and dispute any negative opinions, companies need to react and formulate official statements delivered to the public. These statements are often published as press releases designed to inform the public about a particular crisis and to explain the situation from the company’s perspective. These documents contain a form of strategic communication used to influence the attitudes of consumers toward the brand and to maintain a positive image even in time of crisis (Rosselli et al., 1995). These persuasive strategies need to be examined in order to determine which types of messages are used by companies as a response to which types of scandals.
The Situational Crisis Communication Theory (SCCT) developed by Coombs in 2007 shows that there are different techniques and strategies used by companies to address a scandal. Companies can for instance focus on rational and informational messages which are
5
used to rationalize behavior (Albers-Miller & Royne Stafford, 1999) or they can use more emotional messages which lead consumers to sympathize with the company and
underestimate the seriousness of the matter (Benoit, 1997). There are different types of image restoration strategies and their use depends on the type of crisis. Most crises can be
categorized as environmental, financial, health, privacy or social scandals (Shrivastava et al., 1988). Each of these crisis impacts the consumers and the society in a different way.
Therefore, it is possible that companies use different strategies to address them. Rational messages might be more effective to address technological crises while emotional messages can be used to respond to more sensitive and personal issues among consumers (Cho & Gower, 2006; Choi & Lin, 2007).
The aim of this thesis is to investigate the different types of persuasive messages used by companies to address different types of crises. The persuasive messages (e.g. rational and emotional) used by various companies can be analyzed by breaking them down into the basic elements of strategic crisis communication. The main focus will be on whether companies from different industries focus on rational or emotional information when addressing different types of crisis. This will be examined through a content analysis of the occurrence and frequency of different types of persuasive messages. The role of industry type and crisis type in crisis communication will be analyzed for 10 different companies involved in 5 different crises types. In this way, this dissertation provides practical and systematic research that is useful for both PR professionals and consumer.
Theoretical background Corporate reputation
Corporate reputation is a collective judgement of a company’s attributes (Barnett et al., 2006; Gray & Balmer, 1998) which influences the way consumers interact with the brand. Therefore, companies need to establish and maintain a positive image among their customers. Corporate reputation is a dynamic concept which requires time to build up and manage (Gotsi
6
& Wilson, 2001). The company’s values are presented to the consumers who then form opinions about it in the form of corporate image. This image represents the immediate mental picture of an organization (Gray & Balmer, 1998) which gradually develops into corporate reputation through the bilateral exchange between companies and their customers.
Corporate reputation is heavily dependent on the perceptions and evaluations of consumers (Barnett et al., 2006; Gotsi & Wilson, 2001; Gray & Balmer, 1998). These evaluations are formed over time through direct personal experiences and the feedback from other consumers. Furthermore, impressions can also be based on reports from the media and advertising (Coombs, 2007; Gotsi & Wilson, 2001). In addition, consumers focus on the behavior of an organization and the extent to which it meets their expectations (Coombs, 2007). Positive reputation can attract more consumers, increase sale and investment interest, and create an advantage over competitors (Coombs, 2007).
Types of industries
Corporations and organizations can be categorized according to many features. Every company operates in a specific industry which is the basic determinant of its functioning and the products and services offered to the customers (Sweeney & Coughlan, 2008). Different types of industries also have different sets of values, goals and communication style.
Sweeney and Coughlan (2008) categorized companies into seven distinct industries namely; financial services, pharmaceutical - medical, pharmaceutical - health & beauty,
telecommunications, automobile, oil & gas, and retail. These industry types were
distinguished based on their reporting of corporate social responsibility. However, for this study the industry type was reduced to five categories of financial, technological, food and beverage, and fashion based on the original categories. This categorization will be useful for the analysis of strategic crisis communication since these companies use different
7
(Sweeney & Coughlan, 2008). Such differences can influence the use of persuasive messages therefore, industry type is expected to play an important role in crisis communication. Crisis communication - What is a crisis?
Corporate crises include many different situations that can affect the functioning of a company. Moreover, crisis situations involve sudden and unexpected events which pose financial and reputational threats to the organization (Coombs, 2007). Responsibility for a crisis is determined based on several contributing factors such as initial crisis responsibility, crisis history and prior relational reputation. An organization can be blamed for acts that they performed or permitted to occur (Benoit, 1997). The reputational threat to the company remains as long as the audience thinks the organization is responsible for the crisis (Benoit, 1997). Crises have the potential to harm consumers physically, emotionally or financially. These situations create uncertainty and stress for the consumers who react with an emotional response (Coombs, 2007). Hence, the organizations need to take responsibility for their actions and satisfy consumers’ need for information triggered by the crisis.
There are different types of crisis which can harm the reputation of a corporation. These corporate crises can be classified according to different factors such as accountability (see Coombs, 2007), reputational threat and impact. Many studies (see Benoit, 1997;
Coombs, 2007), mainly focus on the company’s accountability for the crisis. However, only a limited number of studies examine the impact of a corporate crisis on the consumers.
Therefore, this study aims to examine the role of crisis type in strategic communication based on the nature of the different crises.
The research on disaster communication suggests that different types of crises evoke different emotional reactions (Kim & Cameron, 2011). According to Nabi (2003), these emotional responses facilitate different degrees of message processing. Research by Choi and Lin (2007) showed that emotional messages are better remembered and have a stronger effect
8
on attitude change. Their finding also suggest that emotional messages lead to higher risk perception (Choi & Lin, 2007). These authors claim that consumers are more likely to accept emotional messages for corporate crisis of a more emotional nature. Examples include social and health crises, possibly affecting consumers’ lives through human victims or personal harm. The findings of Cho and Gower (2006) also suggest that companies who express human interest when addressing sensitive crises are more successful in positively influencing consumers’ attitudes. On the other hand, attitudes about cognitive issues benefit from rational messages in order to achieve attitude change (Kim & Cameron, 2011).
Based on these findings, we can expect companies to use different persuasive messages to address different types of crises. As such, a newly invented typology based on these findings will be used for the analysis. This typology will serve to examine an
understudied aspect of corporate crises: whether the nature of the crisis influences the type of persuasive messages used by companies. The next section will introduce the definitions of environmental, social, financial, health and privacy crises.
Types of crisis
According to Shrivastava and colleagues (1988), there are several triggering events that can identify the type of a corporate crisis. The authors claim that these sudden events can arise in the production, consumption or production systems. The production and post-production crises can be triggered by human accidents, failure of technology or misuse of harmful substances.
Environmental crisis. The first crisis type stemming from these triggering events is the environmental crisis. These crises include events which cause potential or real damage to human life and the environment (Shrivastava et al., 1988). The crisis causes environmental damage by pollution of air or water, destruction of vegetation and endangerment of species.
9
Lerbinger (2012) covers these types of crisis in his book on crisis management. Corporate crises do not only affect people outside of the company, but also people working for it. Organizations can be under financial pressure due to competitive markets or lack of required resources for steady performance. These companies might resort to questionable behavior and even illegal practices influencing their employees in order to decrease their functional costs and maximize their profit. Such harmful practices involve imposing poor working conditions on workers, making them work long hours or even employing child labor.
Financial crisis. The following type of financial crisis impacts the functioning of the company. The financial and resource pressures mentioned earlier can cause a financial crisis for the company striving to survive. Companies under such financial pressures either failed to achieve sufficient amount of sales or accumulated a large debt leading to bankruptcy.
According to Müller (1985) these companies need to take action to reduce costs in all departments such as marketing, design, purchasing, accounting and finance in order to bring their performance back on track.
Health crisis. Some crisis types can harm the consumers and other parties involved in the production process. Harmful and defective products can impact the health of the
consumers. Klein and Dawar (2004) discuss the product-harm crises which concern manufacturing of faulty or low-quality products and products containing dangerous
substances. Such events can be categorized as health crisis since these products can have a negative effect on consumer’s health. The consumers affected by these product-harm crises can form a negative opinion about the company. The poor quality of products such as food served in fast food chains, also belongs to this category even though it does not have a serious impact on health. Animal testing of products by the cosmetic industry can have a similar effect on consumer’s attitude. Therefore, for the purpose of this study these issues will also be included in the health crisis category.
10
Privacy crisis. The privacy crisis type is not distinctly recognized by research but it will be considered by the present study. This category includes the misuse of consumer’s information for tracking and targeting or other marketing activities. The sharing of client’s personal information with third parties also falls under this category. These types of scandals are important to study because they have a high potential to hurt the trust relationship
between the company and its customers. Mishandling of personal information is a sensitive issue for a majority of people. Hence, we can assume that it poses a reputational threat for the accused organization. This type of crisis is expected to provoke the same emotional responses and uncertainty as the previously mentioned crisis situations.
All the mentioned crises have a different impact on the consumers and consequently the corporate reputation of the responsible companies. In order to restore and maintain a positive reputation, companies need to employ a certain form of strategic communication to address the corporate crisis within the public. The following section will present theory on crisis communication and image restoration strategies used by companies.
Situational Crisis Communication Theory
Strategic crisis communication can be used to protect the reputational assets during a corporate crisis. Coombs (2007) developed the Situation Crisis Communication Theory (SCCT) based on the existing attribution theory. The SCCT provides an explanation of how consumers form attributions during a crisis. The main premise of SCCT is that consumers experience an emotional reaction as a result of a corporate crisis which reflects on their evaluations of the organization (Coombs, 2007). Post-crisis communication can be used by companies to repair the reputational damage caused by the crisis. If no action is taken by the responsible party, there is a high threat of reputational damage.
There are several different strategies that can be used by companies to improve their corporate reputation during a crisis. The most commonly used image restoration strategies
11
are: denial, evasion of responsibility, reducing offensiveness of event, corrective action and mortification. Many research studies often focus on these strategies when investigating the crisis communication of companies, (e.g. Benoit, 1997, 2005, image repair theory; Coombs, 2007, SCCT). However, the present study will rather focus on the deeper content of crisis communication by investigating the role of persuasive messages. According to Albers-Miller and Royne Stafford (1999), rational or emotional messages can be used to effectively
persuade consumers and change their attitudes towards a brand or a company. The companies in crisis also need to change the consumers’ attitudes in order to restore their corporate reputation. Therefore, the findings from this body of marketing research will be used as a broader framework to investigate the role of persuasive messages in strategic crisis
communication. The following section will introduce the definition of persuasive messages and outline their role in crisis communication.
Persuasion and persuasive messages
Persuasive messages are the building blocks of the persuasion process. There are many types of messages that can be used based on their content. The research by Rosselli and colleagues (1995) categorized these messages into rational and emotional. In crisis
communication, rational information can be used to rationalize certain actions of the
company. The company representatives can present the functional benefits of their actions as a form of persuasion. They can provide logical explanations or argue that their actions serve the good of the society. On the other hand, emotional messages can be used to trigger implicit emotional responses by consumers. Such emotional reactions can be used to shift the
consumer’s focus away from the crisis at hand.
Albers-Miller and Royne Stafford (1999) have studied the difference between rational and emotional messages. These persuasive messages differ in content, tone of voice, and persuasive intent. The rational messages used in advertising are designed to promote the
12
logical and rational decision-making of consumers. Such messages focus on the product benefits such as the quality, value and performance (Kotler and Armstrong, 1994). Rational appeals can reduce uncertainty during purchase and motivate consumers to make decisions based on the factual information about the product and its attributes. The persuasive messages containing these rational appeals have been shown to have a direct effect on consumer
attitudes. Research by Rosselli and colleagues (1995) showed that subjects who received rational arguments experienced greater attitude change and more positive cognitive
elaborations. This body of research shows that strong rational arguments can change people’s perceptions in line with the persuasive message.
The emotional messages function on a different basis than rational arguments. These message are focusing on the experiential side of consumers. Their aim is to evoke positive or negative emotions to motivate buying behaviour and to make consumers feel good about the product (Kotler & Armstrong, 1994). This type of persuasive message can be used to retrieve emotional experiences associated with an object or a situation. Using emotional words such as freedom, family and enjoyment (see others in Table 3.) can generate positive reactions from the consumers (Albers-Miller & Royne Stafford, 1999). Research showed that emotional messages can disrupt the usual information processing. Rosselli and colleagues (1995) showed that participants produce fewer cognitive responses to emotional appeals. They fail to carefully analyze the message content when faced with anxiety-provoking messages. This means that emotional arguments can distract consumers and reduce the attentional resources normally used to evaluate the quality of the message (Rosselli et al., 1995).
The presented research suggests that persuasive messages can be used to change the attitude of consumers through communication. However, the role of persuasive messages in crisis communication is still an insufficiently studied topic even though rational and
13
emotional messages can be used for successful corporate image restoration. Therefore, this study will combine the findings of marketing and advertising studies with the existing literature on crisis communication in order to investigate the role of persuasive messages corporate crisis communication.
The present study
This study aims to investigate the number and type of persuasive messages used by companies from different industries when addressing different types of crisis. First, the goal is to analyze the use of persuasive messages across industry type. The second step is to investigate the differences across different types of crisis. These differences will be analyzed in terms of occurrence and frequency of rational and emotional messages used by the selected companies. The aim is to investigate whether the use of persuasive messages depends on industry type and crisis type. Based on these independent variables, the study will investigate the following research questions:
RQ1: Which type of persuasive messages (rational or emotional) are most frequently used by companies across different industry types?
RQ2: What is the difference in the use of rational and emotional messages for different types of crisis?
The persuasive messages will be categorized into two types; the rational messages and the emotional messages. Research on the effect of these messages on cognitive processes of consumers showed that there is a differential route of processing used for rational and emotional messages (Nabi, 2003). Therefore, these messages will be used to analyze the differences in crisis communication strategy between industry type and crisis type. Hypotheses
Based on the categorization by Sweeney and Coughlan (2008) and the informational and functional nature of rational messages, we can expect financial and technological
14
companies to use more rational messages than companies in the food and fashion industry. The first hypothesis concerning the industry type is as follows:
H1: Financial and technological companies use more rational messages than companies in food & beverage and fashion industries.
H2: Companies in food & beverage and fashion industry use more emotional messages than financial and technological companies.
Based on the research by Shrivastava et al. (1988) and other authors (Cho & Gower, 2006; Choi & Lin, 2007), the crisis type is also expected to play an important role in the occurrence and frequency of persuasive messages due to the different nature of each crisis type and impact on the society. Thus, companies are expected to use different types of strategic communication when addressing these crises. Therefore, the hypotheses regarding the effect of crisis type on the occurrence and frequency of rational and emotional messages are as follows:
H3: Companies use more rational messages to address financial and privacy scandals and more emotional messages to address health and social scandals.
H4: Companies use a combination of rational and emotional messages to address environmental scandals.
Method
The research method used for this study is a content analysis of persuasive messages found in corporate press releases. The main focus is on the occurrence and frequency of rational and emotional messages found in these documents. The press releases were gathered from the media and press section of the official corporate websites. The first step of selection was to pick only press releases covering the specific crisis type for each company. Corporate scandals outside of the predefined crisis typology were excluded from the analysis (see Table 2.1). The aim was to analyze an average of 30 press releases per industry type and crisis type.
15
The industry type contains four different categories with ten different companies listed in Table 1.1. Each industry category contains two companies which are paired together based on their similarity. The company pairing will be kept as a constant in order to control for differences between specific companies. Therefore, the comparisons between industry types will be focused on between-groups not within.
Table 1.1 The company categorization according to industry type
The crisis type variable contains five main categories of scandals listed below in Table 2.1 together with their corresponding companies. These crisis types were selected based on the typology developed by Shrivastava et al. (1988). The selected companies were paired together with the crisis types based on the type of scandals they are mostly involved in.
Table 2.1 The crisis type categorization
Type of crisis Company name
Environmental Shell P&B oil
Financial ING Rabobank
Health McDonald’s Nestle
Type of industry Company name
Financial ING Rabobank Technological Shell P&B oil Facebook
Food & beverage McDonald’s
Nestle
Fashion
H&M C&A Target
16
Privacy Facebook Target
Social H&M C&A
The main coding categories concern rational and emotional arguments used to
persuade the public when addressing a company scandal. The unit of analysis for this study is a single press release per industry and per crisis. The occurrence and frequency of persuasive messages used in each press release was coded and analyzed in a form of two separate dependent variables. The coding was carried out by a single coder following the instructions and coding categories enclosed in the codebook. The coding only concerns terms found in-text and hence excludes titles, graphs, footnotes, and tables.
Sample
The study material chosen for this research study consists of press releases covering a corporate crisis. The study focused on the analysis of 10 different companies (see Table 1.1). These companies were selected based on their well-established corporate reputation, amount of public scandals and the coverage by media. These companies are considered to be suitable candidates for the analysis also because of the high amount of published press releases. The press releases were selected from the corporate website of each company. The selection was based on the main topic (e.g. crisis type) of the press releases. Only press releases covering the corresponding crisis type per industry were selected, documents with other topics were excluded from the analysis. A total of 146 corporate press releases were coded in order to reach an average number of 30 press releases per industry and crisis category
Variables
The main coding variables used for this analysis consist of two independent variables (industry type, crisis type) and two dependent variables (rational and emotional terms). The ‘industry type’ variable contains categories of different industries namely the financial, technological, food and beverage, and fashion. The crisis type variable consists of
17
environmental, financial, health, privacy and social scandals. In total, there are ten company names, four industry types and five crisis types (see Table 2.1).
The dependent variables measured in this analysis are the occurrence and frequency of rational and emotional terms. The coding book used for this research study was created based on the previous research by Albers-Miller and Stafford (1999). The final list used for this analysis was decreased from 42 terms to 20 persuasive terms most relevant to corporate crises. Ten additional terms were selected for the study in order to create a complete and exhaustive list of terms (see Appendix 1.).
The rational messages for instance include all functional, numerical1 and practical
information mentioned by the company when addressing a crisis. A list of 15 rational terms was selected for this coding category (see Table 3.2). The general frequency of these terms is shown in Appendix 3. The occurrence of rational terms was coded in form of a binary
variable (‘yes’, ‘no’) and the frequency of these terms was coded as a numerical variable representing the total frequency.
The emotional messages consist of emotional words which can be used to create sympathy within the audience and improve their evaluation of the company even in a time of crisis. The complete list of 15 emotional terms is shown in Table 3.2. The general frequency of these terms is shown in Appendix 3. The occurrence of emotional terms was again coded in form of a binary variable and the frequency as a numerical value. The neutral messages were excluded from the analysis.
Results Descriptive statistics
Tables 4.1 and 4.2 show the distribution of coded press releases among crisis and industry types. The goal was to analyze an average of 30 press releases per industry type and
18 crisis type in order to achieve a uniform distribution. Table 4.1. Frequency of industry types
Industry type Frequency
Financial 27 Technological 44
Food & beverage 24
Fashion 51 Total 146
Table 4.2. Frequency of crisis types
Crisis type Frequency
Environmental 30 Financial 27 Privacy 28 Health 30 Social 31 Total 146 Hypothesis testing
In order to test the hypotheses, the industry and crisis types were re-coded into different variables according to the paired categories. The first comparisons were made with regard to the occurrence of rational within these categories. According to the first hypothesis, financial and technological companies were expected to use more rational messages than companies in food & beverage and fashion industries. Therefore, the financial and
technological industries were recoded as value ‘1’ and fashion and food & beverage industry were recoded as value ‘0’. This new binary variable (from now on referred to as ‘paired industry type’) was used to test the occurrence and frequency of rational terms.
19 Hypothesis 1
Occurrence
The following table 5.1 shows the occurrence of rational and emotional terms for technological and financial industries versus fashion and food and beverage industries. A Chi-square was conducted to test the occurrence with paired industry type as an independent variable and rational messages as dependent variable. The Chi-square showed a
non-significant difference (p = .694) in occurrence of rational terms between these industries. Table 5.1. Occurrence of persuasive terms across paired industry categories
Occurrence
Rational terms Emotional terms
Industry type Yes No Yes No
Technological & Financial 69 2 36 35
Fashion & Food and beverage 72 3 43 32
Frequency
The second set of analyses compared the frequency of rational messages used by specified industry types. Table 5.2 shows the frequency of rational terms used by
technological and financial industries versus fashion and food and beverage industries. A one-way ANOVA was conducted to test the frequency of rational terms as the dependent variable and paired industry type as the independent variable. The ANOVA test showed a significant difference between the frequency of rational messages (F (1,145) = 5.650, p = .019) used by technological and financial industries versus fashion and food & beverage companies. This suggests that technological and financial industries use more rational messages than fashion and food & beverage companies. Even though the previous test for occurrence of rational messages showed no significant effect, the frequency of these
20
result, we can assume that industry type has an effect on the frequency of rational messages. Table 5.2. Frequency of persuasive terms across paired industry categories
Frequency
Rational terms Emotional terms
Industry type M SD M SD
Technological & Financial 17.83 27.37 1.42 2.70 Fashion & Food and beverage 9.73 10.72 1.29 1.61
Hypothesis 2 Occurrence
The second hypothesis concerns the use of emotional messages for specific industry types. The companies in food & beverage and fashion industry were expected to use more emotional messages than financial and technological companies. Table 5.1 shows the occurrence of emotional terms across paired industry types. Another Chi-square was conducted to test the occurrence of emotional terms as the dependent variable and paired industry type as the independent variable. The Chi-square test showed a non-significant difference (p = .422) in the occurrence of emotional terms between the paired industry categories.
Frequency
In order to fully test the second hypothesis, an ANOVA analysis was conducted to investigate the differences in frequency of emotional messages. Paired industry type was used as the independent variable and emotional terms as the dependent variable. The ANOVA test showed a non-significant effect of industry type on the frequency of emotional messages (p = .724). Based on the non-significant result for occurrence and frequency of emotional
messages across paired industry types, H2 can be rejected. Hence, it can be concluded that industry type has no effect on the use of emotional messages.
21 Hypothesis 3
The third hypothesis stated that companies use more rational messages to address financial and privacy scandals and more emotional messages to address health and social scandals. In order to test this hypothesis, the crisis type variable was re-coded into paired categories of privacy and financial crisis as ‘1’ and social and health crisis as ‘0’.
Occurrence
Two Chi-square tests were conducted to test the occurrence of persuasive terms across the paired crisis types. Table 6.1 shows the occurrence of rational and emotional messages for privacy and financial crisis and social and health crises. For the first Chi-square test, the paired crisis type was used as an independent variable and the rational messages as dependent variable. This test showed a non-significant difference in the occurrence of rational terms between privacy and financial crises versus social and health crises (p = .916). Thus, we can assume that the paired industry types have no effect on the occurrence of rational terms.
The second Chi-square used the paired crisis type as an independent variable and the emotional messages as dependent variable. This test also showed a non-significant result for the occurrence of emotional terms between the paired crisis type categories (p = .199). Thus, we can see that the paired industry types have no effect on the occurrence of emotional terms. Table 6.1. Occurrence of persuasive terms across paired crisis categories
Occurrence
Rational terms Emotional terms
Crisis type Yes No Yes No
Privacy & financial 53 2 25 30
Social & health 59 2 35 26
Frequency
22
versus social and health crises were tested by two separate ANOVAs. Table 6.2 shows the frequency of rational and emotional terms across paired crisis types. The first ANOVA used the paired crisis type as an independent variable and the rational messages as dependent variable. This test showed a non-significant effect of paired crisis type on the frequency of rational messages (p = .104).
The second ANOVA used the paired crisis type as an independent variable and the emotional messages as dependent variable. This test also showed a non-significant difference in the frequency of emotional messages across paired crisis types (p = .849). The
non-significant results for both occurrence and frequency of rational and emotional messages used for privacy and financial crises versus social and health crises suggest that H3 can be fully rejected. This shows that the paired crisis type has no effect on the use of rational and emotional terms.
Table 6.2. Frequency of persuasive terms across paired crisis categories Frequency
Rational terms Emotional terms
Crisis type M SD M SD
Privacy & financial 14.31 15.39 1.18 2.04
Social & health 10.21 11.43 1.25 1.58
Hypothesis 4
The last hypothesis stated that companies use a combination of rational and emotional messages to address environmental scandals therefore, there was no expected difference in their frequency. A paired-sample T-test was carried out to test the difference in frequency of rational and emotional messages used to address environmental crises. This test showed a significant difference (t (29) = 2,77, p = .010) between the frequency of rational messages (M = 19.53, SD =37.57) and emotional messages (M = 1.90, SD =3.30) used to address
23
environmental crises. This shows that companies use significantly more rational than
emotional messages to address environmental crisis. However, according to H4 there was no expected difference therefore, this hypothesis can be rejected.
Additional analyses
The hypothesis testing for occurrence and frequency of persuasive terms did not fully confirm any of the expected effects. Therefore, additional analyses were carried out to explore other potential differences between the paired industry and crisis categories. These additional tests used the more specific industry types (financial, technological, food and beverage, fashion) and crisis types (environmental, social, financial, health and privacy) in order to investigate the differences in occurrence and frequency not found in the previous analyses.
The first set of additional analyses was carried out to show the occurrence of
persuasive messages across all industry types as shown in table 7.1. Two separate Chi-square tests were conducted to test the difference in occurrence of rational and emotional terms between industry types. The first Chi-square used the four industry types as the independent variable and the emotional messages as dependent variable. This test showed a significant difference in the occurrence of emotional terms per industry type (χ2 (3) = 18,964, p = .000). Based on table 7.1. we can see that financial industry shows a low occurrence of emotional terms of only 6.3% of the total occurrence (N = 5). The highest occurrence of 70.5% of emotional terms was found for the technological industry (N = 31).
The second Chi-square used the industry type as independent variable and the rational messages as dependent variable. This test showed no significant effect of industry type on the occurrence of rational terms (χ2 (3) = 2,907, p = .406).
Table 7.1. Occurrence of persuasive terms across different industry types Industry type Emotional terms Rational terms
24
Yes No Yes No
Financial 5 22 27 0
Technological 31 13 42 2
Food & beverage 13 11 24 0
Fashion 30 21 48 3
Table 8.1 shows the occurrence of persuasive terms across different crises types. Again, two separate Chi-square tests were conducted to test the difference in occurrence of rational and emotional terms within crisis types. The first Chi-square with the crisis type as independent variable and rational messages as dependent variable showed no significant difference in occurrence of rational messages (χ2 (4) = 4,051, p = .399). The second Chi-square used crisis type as independent variable and the emotional messages as dependent variable. This test showed a significant effect of crisis type on the occurrence of emotional terms (χ2 (4) = 19,740, p = .001). Table 8.1 shows the occurrence of emotional terms for all crisis types. Similarly to the occurrence across industry types, financial crisis showed the lowest occurrence of emotional terms (N = 5). The highest occurrence was found for privacy crisis (N = 20).
Table 8.1. Occurrence of persuasive terms across different crisis type
Crisis type Emotional terms Rational terms
Yes No Yes No Environmental 19 11 29 1 Financial 5 22 27 0 Privacy 20 8 26 2 Health 15 15 30 0 Social 20 11 29 2
Additional analyses were also carried out to explore the effect of industry type and crisis type on the frequency of persuasive terms. These difference were tested by separate
25
one-way ANOVAs. The first ANOVA used industry type as independent variable and the frequency of rational messages as dependent variable. This test showed marginally significant differences between industry types (F (3,142) = 2.505, p = .062). The Bonferroni post-hoc test suggests that technological industries use more rational terms (M = 9,90, SD = 4.23) compared to the fashion industry (M = -9,90, SD = 4.23). The same test showed another marginal difference between the financial industry (M = 10,50, SD = 4.90) and the fashion industry (M = -10,95, SD = 4.90).
The second ANOVA tested the frequency of emotional terms as the dependent
variable and industry type as the independent variable. This test showed a significant effect of industry type on the frequency of emotional messages (F (3,142) = 4,470, p = .005). The post-hoc using the Bonferroni test revealed that there is a significant difference (p = .003) between companies in technological (M = 1,877, SD = 0.52) and financial industries (M = -1,877, SD = 0.52). This shows that technological companies use more emotional terms than companies in the financial industry. No significant differences were found between the other industry types.
The last set of analyses was used to test the effects of crisis type on the frequency of persuasive terms. First ANOVA used the five crisis types as the independent variable and the frequency of emotional messages as the dependent variable. This ANOVA showed a
significant difference in the frequency of emotional messages (F (4,141) = 3.084, p = .018). The post-hoc test showed a significant difference (p = .044) between financial and
environmental crises (M = 1,641, SD = 0.57). This shows that companies use more emotional messages to address environmental crises than financial crises. There was also a significant difference (p = .021) between the use of emotional messages to address financial (M = -1,812, SD = 0.58) compared to privacy crises (M = -1,812, SD = 0.58). This means that companies use more emotional messages to address privacy crises compared to financial
26 crises.
The last ANOVA used the crisis type as the independent variable and the frequency of rational messages as the dependent variable but showed no significant effect of crisis type on the frequency of rational messages (p = .089). In conclusion, the additional analyses enabled insight into the relationship between the more specific industry and crisis types. These results suggest that the pairing of categories lead to the omission of certain significant differences between the groups within one category. Therefore, the specific industry types and crisis types should be studies separately in order to examine the direct effects on the use of persuasive messages by different companies.
Discussion
The present study was carried out to study the persuasive strategies used by companies addressing their corporate crisis. The main aim was to investigate the use of persuasive messages in crisis communication of corporations. These companies were
expected to use different types of messages based on industry type and crisis type. Therefore, the persuasive messages were analyzed in two ways. First, the use of rational and emotional messages was compared according to industry type. Then, the occurrence and frequency of these messages was compared across crisis type. Due to the lack of research on the use of persuasive messages in crisis communication, the findings of this study were compared to the results presented by marketing and advertising studies.
The descriptive results showed that companies generally use more rational than emotional messages in their crisis communication. The frequency of persuasive messages differed only between paired industry types, more specifically between technological and financial industries versus fashion and food & beverage companies. The latter group used significantly less rational messages compared to the first group. The findings presented by Albers-Miller and Royne Stafford (1999) suggest that service providers use more rational
27
messages and informational ads than affective ads. In this case, we can consider the
technological and financial as service providers and fashion and food & beverage companies as goods producers. Therefore, the findings of this study are in line with the previous research which indicated that service providers use less emotional messages and more rational
information in their ads.
Many of the expected effects for the occurrence were not confirmed through hypothesis testing therefore, additional analyses were carried out to explore other possible differences between the more specific industry and crisis categories. The first significant difference was found in occurrence of emotional terms across different industry types. Financial companies used significantly less emotional messages than other industries. According to Mattila (1999) service advertisements contain more informational and factual cues than product advertisements. The present study did not support the argument for
financial industry since this category showed the lowest occurrence of rational messages. The lowest occurrence of emotional messages was also found for the financial crisis. This is an unexpected result since the financial industry together with the technological industry were expected to use more rational messages than the other industries. This study found very low values across the financial category therefore, it can be considered an outlier in the data set which influenced the final results.
Another significant difference revealed by the additional analysis was in the
frequency of emotional messages. The financial industry again showed a lower frequency of emotional messages compared to the technological industry. These differences were not found in the original analysis since these two industry types were paired together. This finding is not in line with the previous research or the predicted effect by this study which means that the financial and technological companies should not be considered as similar industry types.
28
Further differences were found between the frequency of emotional messages used to address privacy and financial crises. The analysis showed that companies use more emotional messages to address privacy crises compared to financial crises. These two crisis types were paired together during hypothesis testing therefore, there were no significant differences observed. However, the result is in line with the expectation since privacy crisis are more sensitive to consumers than financial crisis. According to research by Choi and Lin (2007), sensitive issues benefit from matching corporate responses. Therefore, we can see that companies use more emotional messages to address privacy issues rather than financial issues.
Even though the present study did not fully confirm any hypothesis, the results confirmed that companies from different industries do use different persuasive messages to address crises. This further develops the research of Benoit (1997) on image restoration strategies by introducing the role of persuasive messages. The analysis showed that companies mostly use a combination of rational and emotional messages in their press releases. None of the companies used exclusively rational or emotional messages therefore, only few major distinctions were found across industry or crisis types.
Results of this research have implications for PR professionals and also consumers. This study showed that different companies use rational and emotional terms as a form of crisis communication. The consumers should be aware that these persuasive messages can used by companies to persuade them. Based on these findings, consumers can learn which types of messages are used by companies to address which crises types. On the other hand, PR professionals should be able to employ these persuasive strategies when formulating a corporate press release in order to successfully persuade consumers even in the time of crisis. For instance, this study showed that companies use more emotional messages to address privacy crises which might concern more sensitive issues compared to financial crises.
29
Previous research by Kim & Cameron (2011), suggests that companies should focus on expressing concern for victims and showing human interest when addressing sensitive crises. Such situations can provoke emotional responses which are most effectively addressed by matching emotional messages. As we can see, this body of research provides valuable findings about the role of persuasive messages in crisis communication which can be effectively used by PR professionals and by companies experiencing corporate crisis. Limitations
The present study used a quantitative content analysis which has certain limitations. The coding categories of crisis types and persuasive terms created for this study were limited in number. All crises were coded into five categories which do not provide an exhaustive list of corporate crises. The list of persuasive terms only consisted of ten items per each category of message type. Therefore, the coding process might have missed terms outside of the list which can still be considered a part of the rational or emotional message. The numerical value for each persuasive terms also might not be a relevant measure since messages can compose of more terms. In order to capture the whole persuasive message, a more content-focused method should be employed. Furthermore, the coding should be carried out by multiple independent coders in order to avoid subjective bias.
The type of material used for the study also has its limitations. Press releases have different lengths based on the company or the type of issue therefore, some of them might contain more persuasive terms. Companies with longer press releases can show a higher number of rational and emotional terms. These differences were observed even between similar companies from the same industry (e.g. financial and technological industry). This could have influenced the final results for instance in the case of the financial industry type which showed only low frequency of rational terms compared to the other industries.
30 Conclusion
The purpose of the current study was to advance the literature on crisis
communication with the focus on the role of persuasive messages. A content analysis method was used to investigate the use of rational and emotional messages in corporate crisis
communication. The study revealed several interesting differences across industry and crises types. The analysis showed that companies do not exclusively use rational or emotional messages in crisis communication but rather a combination of both types of persuasive messages. The results showed significant differences between industry and crisis categories but in different directions than expected. These findings show that companies from the technological and financial industry use more rational messages than the fashion and food & beverage companies. Findings of this research have both practical implications for consumers and professional implications for PR professionals and crisis managers. Looking at these results, consumers can learn which terms are used by different companies to address specific crises. Furthermore, the PR professionals can understand which persuasive messages are most frequently used by companies to address different crisis types and utilize these findings when formulating their own press releases.
31 References
Albers-Miller, N. D., & Royne Stafford, M. (1999). An international analysis of emotional and rational appeals in services vs goods advertising. Journal of Consumer Marketing, 16(1), 42-57.
Benoit, W. L. (1997). Image repair discourse and crisis communication. Public relations review, 23(2), 177-186.
Choi, Y., & Lin, Y. (2007, May 23). Communicating risk: The effects of message appeal and individual difference on risk message processing. Paper presented at the annual meeting of the International Communication Association, San Francisco, CA.
Cho, S., & Gower, K. (2006). Framing effect of the public’s response to crisis: Human interest frame and crisis type influencing responsibility and blame. Public Relations Review, 32, 420-422.
Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate reputation review, 10(3), 163-176.
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.
Gotsi M., Wilson A. M., (2001), Corporate reputation: seeking a definition. Corporate Communications: An International Journal, 6(1), 24 – 30.
Ihator, A. S. (2001). Communication style in the information age. Corporate Communications: An International Journal, 6(4), 199-204.
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.
32
Kim, H. J., & Cameron, G. T. (2011). Emotions matter in crisis: The role of anger and sadness in the publics’ response to crisis news framing and corporate crisis
response. Communication Research, 38(6), 826-855.
Kotler, P. and Armstrong, G. (1994), Principles of Marketing, 6th ed., Prentice-Hall, Englewood Cliffs, NJ.
Lerbinger, O. (2012). The crisis manager. Routledge
Mattila, A. S., (1999), Do emotional appeals work for services?, International Journal of Service Industry Management, 10(3), 292 - 307
Müller, R. (1985). Corporate crisis management. Long Range Planning, 18(5), 38-48. Nabi, R. L. (2003). Exploring the framing effects of emotion: Do discrete emotions differentially influence information accessibility, information seeking, and policy preference? Communication Research, 30, 224-247.
Rosselli, F., Skelly, J. J., & Mackie, D. M. (1995). Processing rational and emotional messages: The cognitive and affective mediation of persuasion. Journal of Experimental Social Psychology, 31(2), 163-190.
Shrivastava, P., Mitroff I., Miller D., Miglani A. (1988) Understanding Industrial Crises In: Journal of Management Studies. 25(4), 285-303.
Sweeney, L., & Coughlan, J. (2008). Do different industries report corporate social responsibility differently? An investigation through the lens of stakeholder theory. Journal of Marketing Communications, 14(2), 113-124.
33 Appendix 1. Codebook Type of document a) press release b) news article c) report d) other ………... Source of document a) company website b) other ………... Year of publication: ………... Length of the article (word count):
………...
1. Does the press release cover a company scandal? (self-designed item) a. Yes
b. No (if yes, end the analysis)
2. Which of the target companies is mentioned in the press release? (self-designed item) a. Shell b. P&B oil c. ING d. Rabobank e. Facebook f. Target g. McDonald’s h. Nestle i. H&M j. C&A
k. None (if yes, end the analysis)
3. Which type of scandal is addressed in the press release? (self-designed item) a. Environment
b. Financial c. Privacy d. Health
34 e. Social
f. Other (if yes, end the analysis)
Fill in the following table: (designed by Albers-Miller and Stafford, 1999) List of persuasive terms
Emotional terms Frequency Rational terms Frequency
Freedom Effective Family Convenient Dear Wisdom/knowledge Morality Productivity Maturity Modern Community Technological Nurturance Safety Enjoyment Durable Traditional Cheap Humility/modesty Independence Feel/feeling Numbers Sorry Function Apologize Benefit/beneficial Sympathy/sympathize Solution TOTAL TOTAL
35 Appendix 2. Coder instructions
Please follow the enclosed instructions when coding for the different types of crisis.
Use the following examples of company crisis when determining the different types of crisis: Environmental: pollution and endangerment of species
Financial: bankruptcy and debt
Health: animal testing and food quality
Privacy: sharing of personal information and tracking Social: poor working conditions and child labor
36 Appendix 3. Frequency of persuasive terms
The following tables show the total frequency of emotional terms across all coding categories. Table 3.1 shows that the mostly used emotional terms were ‘family’ (N = 7), ‘community’ (N = 5) and ‘believe’ (N = 5). The least used terms were ‘morality’, ‘nurturance’ and ‘sympathy’.
Table 3.1. List of emotional terms based on Albers-Miller and Stafford (1999) Frequency Mean Std. Deviation
Freedom 3 0.05 .31 Family 7 0.27 .89 Dear 1 0.03 .16 Morality 0 0.00 .00 Maturity 1 0.06 .23 Community 5 0.30 .84 Nurturance 0 0.00 .00 Enjoyment 1 0.05 .21 Traditional 3 0.06 .32 Humility/modesty 2 0.02 .18 Feel/feeling 2 0.07 .28 Sorry 2 0.03 .22 Apologize 4 0.07 .38 Sympathy/sympathize 0 0.00 .00 Believe/belief 5 0.37 .80 Total 15 1.36 2.20
The next table 3.2 shows that the most frequently used rational terms were ‘growth’ (N = 65), numbers (N = 60) and ‘safety’ (N = 43). The least used terms were ‘durable, ‘modern’ and ‘cheap’. As we can see from the two tables, emotional terms were generally used less frequently (N = 15) than the rational terms (N = 205).
37
Table 3.2. List of rational terms based on Albers-Miller and Stafford (1999) Frequency Mean Std. Deviation
Effective 15 0.48 1.49 Convenient/convenience 2 0.07 .32 Wisdom/knowledge 2 0.04 .23 Productivity/production 29 2.11 3.99 Modern 1 0.01 .12 Technological/technologies 36 0.61 3.24 Safety 43 1.23 4.03 Durable 0 0.00 .00 Cheap 1 0.01 .12 Independence 5 0.23 .60 Numbers 60 6.51 9.65 Function 2 0.05 .28 Benefit/beneficial 25 0.42 2.15 Solution/resolve 20 0.69 2.56 Grow/growth 65 1.21 5.70 Total 205 13.67 20.90
