Communicate your way out of a reputation crisis: An experimental research about crisis response types and their influence on organisation reputation

(1)

Communicate your way out of a reputation crisis

An experimental research about crisis response types and their influence on organisation reputation

Institute for Security & Global Affairs

Leiden University – Faculty of Governance and Global Affairs Master Thesis Crisis and Security Management

Thesis Supervisor: Dr. Honorata Mazepus, Assistant Professor at Leiden University Second Reader: Dr. Sanneke Kuipers, Director of Education ad interim and Associate Professor at Leiden University

Program: Master Crisis & Security Management Student: Wouter Ypma, BSc

Student number: s2353113 Date of admission: 2019-08-04

Subject: Effects of crisis communication styles on organisation reputation Word count: 14353 words (Excluding References and Appendices)

(2)

2 ‘Keep calm and finish your thesis’

These words on the first page of the master thesis manual seem very simple, yet I believe I never met a master student who combined keeping calm with finishing a master thesis in one sentence. Although writing a master thesis is extremely challenging, I can now confirm that it does not mean you cannot keep calm during the thesis process.

Signing up for the elective course Crisis Communication has been one of my best decisions within my master’s degree. From this point onwards, I developed a strong interest in crisis communication, and I am satisfied that I have been able to convert this interest into a master thesis.

I would like to thank dr. Sanneke Kuipers for her feedback throughout my thesis process and with special gratitude, I would like to thank dr. Honorata Mazepus for her support during our meeting sessions, her critical feedback on my progress and her positive attitude towards the completion of my thesis.

(3)

3

1. Introduction

1.1 Intro

Imagine you are a client of a new health insurance organisation. Determined to

become a leading organisation in their field, the manager of this insurer has drawn up a clear vision for the organisation. General impression, preferability, trust and having a positive feeling are keywords of the manager which he considers elements of high importance for the success of this insurer. Furthermore, the manager states that the organisation should be well-managed, hire competent employees and have a high customer treating standard, thus clients as you are likely to remain stakeholders of the organisation.

Just when the organisation started to penetrate the market, an organisational crisis occurs. Due to an error in the security system, the health insurer leaked the data of its clients to private companies. This puts the new health insurer immediately at risk concerning its organisation reputation. Therefore, to respond adequately to possible reputational risks, a press conference is scheduled in a couple of minutes for an official crisis response from the manager of the organisation. What do you want to hear in this crisis response as a client of this health insurance organisation?

1.2 Problem outline

The situation sketched above appeals to the imagination, as possible ways to use crisis communication during the press conference must be passing your head. At the same time, answers to the stated question will have its implications for the reputation of this health insurance organisation… Thus, what is the best possible way to respond?

Although organisational crises occur regularly since the establishment of

organisations, the academic field of crisis communication is relatively new (Dionisopolous & Vibbert, 1988; Ice, 1991). At the start of its development, crisis communication research was merely focused on specific case-studies of crises and their organisational reactions to them. Due to the absence of a theoretical framework in crisis communication, firm conclusions about best practices in crisis response could not be made.

This changed with the emergence of the leading Image Repair Theory by Benoit (2005) and the elaboration on this framework by Coombs (2006) in the form of Situational

(6)

6

Crisis Communication Theory. These theories have been tested multiple times in the form of case studies. Nowadays, crisis communication research has begun moving beyond case studies towards experiments, assessing systematically how people perceive a crisis (Coombs & Holladay, 2008, p. 252).

Since there are not many experimental studies completed in crisis communication, there is a gap between applying the theoretical framework in practice. Furthermore,

perceptions of multiple stakeholders which constitute organisation reputation have not been tested substantially in different contexts yet. Therefore, the problem consists of testing multiple perceptions of stakeholders experimentally to derive suitable crisis responses.

1.3 Academic relevance

Now the problem for this thesis has been outlined, possible ways to solve this problem should be investigated. There are multiple factors which influence organisation reputation. Apart from the various crisis communication strategies Benoit (2005) and Coombs (2006) drafted, which are outlined more detailed in Chapter 2: Literature review of this study, previous organisational reputation status and an organisational brand influence reputation management (Piranfar & Matthews, 2008). This obstructs best crisis response practices to be composed within organisations, as these two factors always have an impact on organisation reputation. Therefore, the academical challenge consists of taking these factors out to determine whether a more objective best crisis response practice should be constituted when an organisational crisis happens.

To act on this academical challenge, this study presents a novel operationalisation to measure organisation reputation. Previous reputation measurement tools as the Fortune AMAC questionnaire have been primarily focused on the perceptions of the manager

concerning organisation reputation and evaluated reputation as financial performance (Chun, 2005). In contrast to the Fortune AMAC questionnaire, some operationalisations of

organisation reputation, for example the one from the Reputation Quotient, have been centred around multiple perceptions of stakeholders (Fombrun & Foss, 2001). However, this one is only applicable for corporate businesses and not for newly established organisations whereas their organisation reputation level had still been unknown (Fombrun & Foss, 2001).

(7)

7

aiming to bridge the gap between the public and private domain and attempting to derive best crisis response practices.

1.4 Societal relevance

Besides the academic gap this paper attempts to bridge, societal challenges concerning crisis response practices which influence organisation reputation positively are present in the year of 2019 as organisations often struggle with crisis response practices. Two recent examples show the complexity of using crisis communication when organisational crises occur.

In the Volkswagen emissions scandal of 2015, Volkswagen manipulated engine controls to avoid high emission outcomes in emission tests (Mansouri, 2011, p. 211). Volkswagen responded to this organisational crisis with contradictory statements, not being able to provide a solid and honest explanation for the scandal (Zhang, Veijalainen, & Kotkov, 2016, p. 177). This way, their clients valued the organisation as dishonest, doubting their crisis responses and detesting the way the clients were misled (Zhang et al., 2016, pp. 177-179). This split the way the organisation related to their customers, as clients are more inclined to distrust the organisation (Zhang et al., 2016, pp. 177-179).

A different organisational crisis happened at private company Target and concerned a huge data breach. Due to a hack in their internal system, Target exposed the data of 110 million customers (Conner, Brokaw & McManus, 2014, p. 55). Target offered their apologies for the leakage, only they provided an apology before they understood the scope of the crisis and the cause of the problem (Conner, et al., 2014, pp. 59-65). This way, they needed to adapt their statements multiple times. This resulted in Target being valued as unprofessional and suspicious by their clients, which embarrassed the legitimacy of the organisation. (Conner et al., 2014, pp. 59-65).

As these two organisational crises show how various crisis response practices are perceived as incomplete responses by their audiences, best crisis response practices remain a utopia. Compared to these organisational crises which occurred in private organisations, crisis response practices in public organisations have not been researched that extensively with a specific crisis communication’ framework. However, many public organisations experience difficulties with their communication strategies while forming crisis response practices during

(8)

8

a crisis (Olsson, 2014, pp. 113-116). They face problems preparing, communicating and managing large-scale societal crises (Olsson, 2014, p. 113).

Organisations exist which share both public and private characteristics. Health care insurance companies in the Netherlands are an example of this category. They have been privatised over the past years, but still share a public nature since all citizens over eighteen are obliged to take one out (Rijksoverheid, 2019). This type of organisations, which operate on the verge of the public-private domain, should be researched more extensively regarding crisis response practices, aiming to find out which response type influences organisation reputation most in this specific context. To help their reputation management, the societal challenge focuses on determining which best practices in crisis communications should be applied in practice for private organisations with public characteristics.

1.5 Research question

To address both these academic and societal knowledge gaps, I pose the following research question:

To what extent do types of crisis response influence organisation reputation?

In this research I will focus on multiple perceptions that stakeholders, who evaluate an organisation’s reputation, have of crisis responses when an organisational crisis occurs. This organisational crisis will happen in a fictional organisation that is both a public and a private organisation, which will be a non-existing health insurer named ‘Mijn Zorg’. Within this organisation, a hypothetical scenario of an organisational crisis will occur, which will be tested experimentally with different conditions to investigate the influence of crisis responses on organisation reputation.

Furthermore, to make this research more explicit, I will concentrate on crisis response types that range from the level of responsibility an organisation takes for the perceived organisational crisis. In my case, this means an organisation will either deny or take

responsibility when an organisational crisis hit. Besides, I will examine what effect denying responsibility has when at a later stage, an organisation takes responsibility for the perceived

(9)

9

crisis. Last, I will explore the effectiveness that providing an excuse or offering compensation as crisis response types have on organisation reputation.

1.6 Reading guide

First, a literature review will be constructed to outline the leading academic

perspectives on organisation reputation. In this review, also the field of crisis communication will be examined. At the end of this chapter, I will make choices regarding crisis response types I will test in this study and I will formulate hypotheses which are derived from the literature.

Second, the methodology used in this study will be explained, which includes

outlining my experimental design. Furthermore, both the independent variable crisis response types and the dependent variable organisation reputation will be clarified. To conclude this chapter, I will state which processes had been followed for acquiring and analysing the data used in this thesis.

Third, the results of this study will be presented, which will begin with explaining the sample characteristics. After the manipulation checks are outlined, a factor analysis will show whether the selected items in this study measured organisation reputation. Subsequently, I will test the hypotheses of this thesis, and I will end this chapter with the mean scores of the researched conditions on organisation reputation.

Fourth, a discussion and conclusion will be drawn up to put the results of this thesis into perspective. This includes discussing the results in relation to the literature and giving the limitations of this research. Thereafter, I will present the main findings and an answer to the central research question. In the end, I will conclude with a practical recommendation and future research opportunities.

(10)

10

2. Literature review

In this chapter, I will start with defining key concepts which are used in this thesis. These concepts are ‘organisation reputation’ and ‘stakeholders’. Furthermore, I will outline how organisation reputation is obtained, evaluated and maintained.

Thereafter, I will switch to the field of crisis communication by explaining crisis responsibility and outlining crisis communication theories. This includes the Image Repair Theory of William Benoit (1995) and the Situational Crisis Communication Theory of Timothy Coombs (2006). Additionally, I will show how these theories are put into practice.

Afterwards, I will continue with explicating the choices for the compassion and

apology crisis response types. Also, I will state to what extent I will research the staged retreat crisis response. In the end of the chapter, after having outlined a few studies of crisis

communication which have been executed experimentally, I will present hypotheses which will be tested in this thesis.

2.1 Organisation reputation

Organisation reputation has been defined in different ways. There is consensus over the fact that organisation reputation refers to the general perception of an organisation across stakeholders over a certain period (Waeraas & Byrkjeflot, 2012, p. 189). This way,

organisation reputation is the understanding of an organisation as it exists in the minds of people who are interested in an organisation (Fombrun, 1996, p. 12). Despite this is the beginning to define its concept, organisation reputation remains an intangible phenomenon which needs more clarification.

To further deepen out its definition, Lange, Lee and Dai (2011) conceptualised the term organisation reputation into three central themes. They derived these central themes from forty-three articles which attempted to define organisation reputation from 1999 until their publication.

The first conceptualisation of organisation reputation according to Lange et al. (2011) is being known, meaning organisation reputation is the general awareness or visibility of the organisation (p. 155). This means organisation reputation consists of the prominence of an

(11)

11

organisation in the collective perception (Bromley, 2000; Shamsie, 2003; Rindova et al., 2005).

The second conceptualisation of organisation reputation is being known for something, which Lange et al. (2011) outline as the perceived predictability of the outcomes and the performance of an organisation (p. 155). This means organisation reputation entails the sum of relevant concepts about the outcomes and performance according to their audience interests (Rindova et al., 2005; Rhee and Haunschild, 2006; Fisher and Reuber, 2007; Love and Kraatz, 2009).

The third conceptualisation of organisation reputation derived from those forty-three articles, is generalised favourability, meaning that organisation reputation entails the

perceptions or judgments of the organisation as good, attractive and appropriate (Lange et al., 2011, p. 155). This means organisation reputation embodies an assessment of the

organisation’s favourability based on organisational attributes which their audience considers of vital importance for their reputation (Roberts and Dowling, 2002; Fisher and Reuber, 2007; Love and Kraatz, 2009).

These three conceptualisations share the common denominator that organisation reputation is ‘’a form of altitude based on how well an organisation does or does not meet certain criteria or expectations stakeholders have for an organisation’’ (Coombs & Holladay, 2011, p. 58). Therefore, I will use this definition which catches the common thread of all three conceptualisations of organisation reputation.

2.2 Stakeholders

Now the definition of organisation reputation has been outlined, the importance of the perceptions of stakeholders should be explored to acquire a better understanding of this definition. Therefore, the term stakeholders needs to be defined in order to understand the perceptions of whom an organisation’s reputation has been built.

Stakeholders are ‘’any group or individual who can affect or is affected by the

achievements of the organisations objectives’’ (Freeman, 1984, p. 46). Thus, stakeholders are for example clients, governments or organisations, however, stakeholders are also employees,

(12)

12

employers and the media. This broad scope of stakeholders enables organisations to have different reputations amongst different stakeholders at the same time (Bromley, 2000).

Apart from having different reputations amongst different groups of stakeholders, the perceptions of organisation reputation could also vary within stakeholder groups. Therefore, organisations can improve their organisation reputation by analysing the different

characteristics within one stakeholder group (Chun, 2005). This will better its organisation reputation amongst a certain stakeholder group which has at least one characteristic, for example age or the fact that all stakeholders are customers of an organisation, in common (Freeman, 1984, p. 24). Since being a client of an organisation is a singular characteristic of a stakeholder group, I define stakeholders as clients of an organisation from whom their

perceptions influence organisation reputation in this thesis.

2.3 Composing organisation reputation

Defining organisation reputation and outlining the way I use the term stakeholders is essential to understand the meaning of these two concepts. However, this does not cover how organisation reputation is obtained. Thus, this paragraph contains insights in the following question: how does an organisation achieve organisation reputation?

Normally, an organisation follows three steps in order to achieve a leading reputation (Waeraas & Byrkjeflot, 2012, p. 190). First, a strategy is formed based on an idealised vision of how the organisation should be perceived in the future (Waeraas & Byrkjeflot, 2012, pp. 190-191). Second, based on the evaluation of those perceptions, a program for organisational self-presentation should be developed (Waeraas & Byrkjeflot, 2012, pp. 191-192). Third, an organisation must conduct measurements in order to measure audience perceptions (Waeraas & Byrkjeflot, 2012, p. 192).

In order to optimise achieving a leading organisation reputation, organisations should always be aware of the context in which an organisation builds up a reputation, which is of major importance to achieve a solid reputation (Waeraas & Byrkjeflot, 2012, pp. 192-193). Perceptions of stakeholders regarding organisation reputation will be influenced by changing contexts, which requires organisations to adapt to new inventions in their field (Waeraas & Byrkjeflot, 2012, pp. 192-193). Therefore, the challenge in reputation management consists of

(13)

13

managing the perceptions that form organisation reputation continuously in order to stabilise or better organisation reputation (Elsbach, 2006).

2.4 Measuring organisation reputation

Knowing how organisation reputation should be composed is helpful for organisations striving for positive organisation reputation outputs. However, this does not comprise how organisation reputation is evaluated. Thus, this paragraph contains insights in the following question: how does an organisation measure organisation reputation?

Multiple tools had been developed to evaluate organisational performance over the past decades. Due to the absence of a leading rating system to capture organisation reputation in the twentieth century, organisation reputation studies have varied frequently in measuring organisation reputation.

This changed with the establishment of the measurement tool Fortune AMAC in 1984. In this tool of magazine Fortune, CEOs and managers of Americas Most Admired Companies (AMAC) are asked to rank competitive organisations (Chun, 2005, pp. 94-99). Based on eight factors of organisation reputation which they constituted themselves, Fortune AMAC tries to capture organisation reputation (Chun, 2005, p. 99). These eight factors, outlined in Table 1, are as follows:

Table 1

Factors of the Fortune AMAC Factors

1. Financial soundness

2. Long-term investment value 3. Use of corporate assets 4. Innovativeness

5. Quality of the company’s management 6. Quality of its products and services

7. Ability to attract, develop and keep talented people 8. Acknowledgement of social responsibility

(14)

14

As CEOs and managers are the actors who rank these attributes, the key audience of the Fortune AMAC tool are single stakeholders (Chun, 2005, p. 94). Furthermore, Fortune AMAC is mainly established at managing financial performance of an organisation (Chun, 2005, pp. 94-99). In 2001, Charles Fombrun and Christopher Foss developed their tool to measure organisation reputation by elaborating on the Fortune AMAC measurement tool. Instead of focusing on single stakeholders, they state that multiple stakeholders define organisation reputation (Chun, 2005, p. 94). Furthermore, their measurement was not composed solely to measure financial performance (Chun, 2005, p. 94).

Fombrun and Foss (2001) argue that corporate organisation reputation is measured best with their measurement tool which they named the Reputation Quotient. By questioning people which organisations they liked and why they liked these organisations, they tried to capture people’s perceptions towards organisations (Fombrun & Foss, 2001, p. 1). They concluded that there are twenty items from which organisation reputation consist of (Fombrun & Foss, 2001, p. 1). These twenty items were listed under six factors which they believed summarised the twenty items (Fombrun & Foss, 2001, p. 1). These six factors and twenty items, outlined in Table 2, are as follows:

Table 2

Factors and items of the Reputation Quotient

Factors Items

1. Emotional appeal I have a good feeling about the company I admire and respect the company I trust this company

2. Product and services This company stands behind its products and services This company develops innovative products and services

This company offers high quality products and services This company offers products and services that are good value for money

(15)

15

3. Vision and leadership This company has excellent leadership This company has a clear vision for its future

This company recognises and takes advantage of market opportunities

4. Workplace environment This company is well managed

This company looks like a good company to work for This company looks like a company that would have good employees

5. Social and environmental responsibility

This company supports good causes

This company is an environmentally responsible company

This company maintains a high standard in the way it treats people

6. Financial performance This company has a strong record of profitability This company looks like a low risk investment This company tends to outperform its competitors This company looks like a company with strong prospects for future growth

(Chun, 2005, p. 94).

2.5 Reputation maintenance

Composing and measuring organisation reputation assist in exploring how organisations should act in practice to obtain a positive organisation reputation status. However, organisation reputation is not only measured once: it is constantly changing

(Argenti & Druckenmiller, 2004). Therefore, organisation reputation should be measured over multiple periods of time to make organisation reputation representative (Argenti &

Druckenmiller, 2004). Organisations should respond to possible risks such as financial recessions to maintain a solid reputation (Argenti & Druckenmiller, 2004).

One of the dangers for an organisation reputation status is an organisational crisis. Organisations try to prevent organisational crises to happen, however this is not always feasible. If an organisational crisis does occur, crisis response practices are tuned to preserve organisational reputation (Heath, 2004). This is because any crisis threatens an organisation’s reputation (Barton, 2001). Therefore, crisis management is deployed to minimise

(16)

16

organisation’s reputation damage and to focus on maintaining the same level of organisation reputation (Barton, 1993).

2.6 Crisis communication

Crisis communication is valued as the essence of crisis management (Coombs & Holladay, 2011, p. 25). Crisis communication deals with the relationship between the

occurrence of crisis situations, communication strategies and crisis perceptions of recipients. Therefore, this field is qualified as highly interdependent and very dynamic (Coombs & Holladay, 2011, p. 25).

Crisis communication is used during organisational crises to manage the effects of crises. Seeger, Sellnow and Ulmer (1998) identify an organisational crisis as ‘’a specific, unexpected and non-routine event or series of events that create high levels of uncertainty and threaten, or are perceived to threaten, an organisations’ high priority and goals’’ (p. 232).

In addition to this definition, an organisational crisis could possibly split the social structure (Seeger et al., 1998). This means that an organisational crisis could split the way an organisation relates to its customers and vice versa (Seeger et al., 1998). This results in different interaction approaches between an organisation and its stakeholders (Dowling, 2002). An organisational crisis could therefore damage an organisations’ valuable reputation and its legitimacy (Fombrun, 1996).

2.7 Crisis responsibility

When an organisational crisis occurs, the topic of discussion is whether organisations should be hold responsible for the organisational crisis. This topic is valid for the

Volkswagen- and Target cases which were outlined in the introduction chapter of this study, see paragraph 1.4 Societal relevance. In these cases, the types of crisis response were not adequate, which can be explained by the fact that these organisations had not clearly understood their crisis responsibility regarding the reactions to their crisis responses from stakeholders.

This shows that crisis communication has a lot to do with crisis responsibility an organisation takes or should take for actions which caused the initial crisis (Fombrun, 1996).

(17)

17

Perceptions of stakeholders determine to a large extent whether an organisation should be hold responsible for the organisational crisis (Fombrun, 1996). Sometimes it occurs that an external factor plays a significant role in causing a crisis, but an organisation is still held responsible for handling the crisis (Fombrun, 1996). If to a certain extent crisis responsibility could be proven, organisations should choose crisis response strategies to minimise their organisation reputation damage (Fombrun, 1996).

2.8 Crisis communication theories

In the process of formulating responses, organisations should weigh which of the two basic crisis communication strategies is most suitable to use for crisis responses. First, an organisation could select crisis knowledge management, focusing more on identifying

sources, collecting and analysing information, and sharing knowledge with their stakeholders (Coombs & Holladay, 2011, p. 25). Second, an organisation could prefer stakeholder reaction management, which involves communicative efforts to influence how stakeholders perceive the crisis, the organisation in crisis and the organisations’ crisis response (Coombs & Holladay, 2011, p. 25).

These two basic crisis communication strategies are not mutually exclusive, instead they are interchangeable (Coombs & Holladay, 2011, p. 25). This means an organisation could choose to use both strategies at the same time. Furthermore, there are no differences regarding the audience towards whom those two basis crisis communication strategies are communicated (Coombs & Holladay, 2011, pp. 25-42). Either crisis knowledge management or stakeholder reaction management could be used to address both internal and external audiences (Coombs & Holladay, 2011, pp. 25-26). This entails that possible stakeholders could be employees of the organisation in case of an internal audience; however, stakeholders could also be clients from the organisation in case of an external audience. In this thesis, I will focus on stakeholder reaction management, since I will examine crisis response types which influence the perceptions of stakeholders.

2.9 Stakeholder reaction management: 1

Within the stakeholder reaction management strategy, distinctions should be made in type of crisis communication during a crisis. William Benoit developed the first leading crisis communication theory in 1995: Image Repair Theory (IRT). His framework is based on two

(18)

18

assumptions, 1) that an offensive act has been done to the organisation and 2) that an accusation of responsibility for the act is designated (Benoit, 2005).

The first condition includes that an organisation values image and organisation

reputation as important, thus any act which tries to undermine an organisation’s reputation, is regarded as offensive (Benoit, 1995, pp. 10-14). The second condition does not indicate that one’s responsibility for the act is proven, however it explains that the accusation of

responsibility for the act is based on the perceptions of stakeholders (Benoit, 1995, pp. 17-22). This way, stakeholders hold an organisation responsible for the act, instead of objectively state that an organisation is responsible for an act (Benoit, 1995, pp. 21-22). If these two conditions are satisfied, IRT proposes crisis communication to defend organisation reputation (Benoit, 2005).

2.10 Image Repair Theory in practice

William Benoit (1997) designed five major strategies as crisis response practices to restore organisation reputation. In Table 3, I outlined these strategies and their key

characteristics:

Table 3

Major strategies Image Repair Theory

Crisis response strategies Meaning of strategies

Denial Management states that they did not do it or shift the blame

towards another party

Evasion of responsibility Management says they responded to another act, the act was an incident, or they meant well in the act

Reducing offensiveness of event

Management minimises the act, declares that there are more important considerations, reduce credibility of the accuser or offers compensation

Corrective action Management plants to solve or prevent the problem

Mortification Management apologises for the act

(19)

19

There are a couple more characteristics which define the theoretical framework of Benoit (1997). In case of IRT, the focus to develop a crisis communication strategy is the active communicator of the crisis response (Benoit, 1997, p. 2). Furthermore, IRT is primarily centred around crisis communication for corporate organisations, which makes IRT more focused on private businesses (Benoit, 1997, p. 1). This makes IRT less applicable for the public sector as the public sector differs in norms, values and reputation building from the private sector (Benoit, 1997, pp. 1-4).

2.11 Stakeholder reaction management: 2

Building on the framework of Benoit (2005), Timothy Coombs (2006) examined with his theoretical model Situational Crisis Communication Theory (SCCT) how crisis

response could assist in reputation management. Coombs (2006) explored how crisis

responses came across to the recipients of the message, which moves the focus away from the active communicator of the crisis response to the audience which receives the crisis response (Coombs, 2006). This part of stakeholder reaction management is called Attribution Theory

(AT), which had been founded by Bernard Weiner (1986) to assist in corporate crisis

communication (Coombs, 2011, p. 592).

Coombs (1998) defines the concept of crisis responsibility as ‘’the degree to which stakeholders blame the organisation for the crisis’’ (p. 180). This way, the audience

determines the level of crisis responsibility from the organisation, which indicates potential crisis damage to an organisation’s reputation (Coombs & Schmidt, 2000). Weiner (1995) states that the attribution of responsibility for a crisis affects people’s affective, behavioural and cognitive reactions towards the organisation, which determine their willingness to interact with the organisation in the future. This way, the attribution of responsibility will influence to what level stakeholders like, know and relate to the organisation (Weiner, 1995, pp. 10-12).

2.12 Situational Crisis Communication Theory in practice

To test various crisis response strategies, Coombs (2006) designed ten crisis response strategies on recipients, varying from the range of responsibility a crisis response implied. Just as his theoretical framework, his crisis response strategies were also built on the theory of Benoit (2005), believing that these ten crisis response practices covered the reactions in which most stakeholders will find themselves. In addition, Coombs (2006) executed his study

(20)

20

experimentally, putting his respondents in the position to rate his ten crisis response strategies, providing his theory with empirical evidence. These ten crisis response strategies are outlined in Table 4:

Table 4

Crisis response strategies in Situational Crisis Communication Theory Crisis response strategies Meaning of strategies

Denial Management claims there is no crisis

Scapegoat Management blames some outside entity for the crisis

Attack the Accuser Management confronts the group or person claiming that something is wrong

Excuse Management attempts to minimise crisis responsibility by

claiming lack of control over the event or lack of intent to do harm

Justification Management attempts to minimise the perceived damage

caused by the crisis

Ingratiation Management praises other stakeholders and/or reminds

people of past good works by the organisation

Concern Management expresses concern for victims

Compassion Management offers money or other gifts to victims

Regret Management indicates they feel badly about the crisis

Apology Management accepts full responsibility for the crisis and

asks stakeholders for forgiveness

(Coombs, 2006).

2.13 Situational Crisis Communication Theory – accepting responsibility

When an organisation accepts responsibility for a crisis due to the perceived

responsibility by the audience, four strategies remain present to deal with a crisis and help the victims of this crisis (Coombs & Holladay, 2011, p. 36). These four crisis response strategies are concern, compassion, regret and apology. If, on the other hand, no perception of crisis is present or no responsibility for a crisis has been taken, an organisation does not worry about victims (Coombs, 1999, p. 129). Therefore, the condition of responsibility is necessary to include crisis response strategies which are directed to victims of a perceived crisis.

(21)

21

It is hard to fully differentiate between these four crisis response strategies and to categorise crisis responses solely into one of these four crisis response practices. In addition to this, an organisation can choose to complement their crisis response strategies, meaning that an organisation can use two or more crisis response strategies in their original crisis response. For example, the concern strategy could possibly contain elements of the compassion strategy as they both centralise the concept of victims in their response strategy (Coombs & Holladay, 2011, p. 36). This corresponds with the regret and apology response strategies concerning a fine line between feeling bad about a crisis and asking for forgiveness not being entirely present. Therefore, I will not research the differences between either concern and compassion responses or regret and apology responses.

2.14 Compassion and apology

Since the distinction between either concern or compassion and either regret or apology is clearer, I will use one of both pairs of crisis response strategies in this study. Most of the research in crisis communication has been contributed to investigate the compassion crisis response strategy (Coombs, 1999, p. 128). This has been done by evaluating actual crises and comparing crisis responses which were used in various cases, such as the study of Marcus and Goodman (1991) for example. They measure the effects of different crisis response strategies on stock returns in actual crises, founding that by offering any form of compassion, stocks were most likely to return to their previous level (Marcus and Goodman, 1991).

Apology has also been frequently used as a crisis response. It has been proven to be an effective strategy, however, only when apology has been compared to crisis strategies

whereby no- or less responsibility for the crisis had been taken, has apology been proven to be an effective strategy (Coombs & Holladay, 2008, p. 253). There is not much research about the effectiveness of apology in relation to other crisis response strategies where an

organisation takes responsibility for the perceived crisis (Coombs & Holladay, 2008, p. 253). Thus, to research apology in relation to other crisis response practices where an organisation takes responsibility for the perceived crisis, I choose the crisis response strategies compassion and apology in this thesis.

(22)

22 2.15 Staged retreat crisis response

Organisations do supplement or reconsider their crisis response strategies. It occurs that organisations take responsibility at a later stage for a crisis. This is because blame avoidance is perceived as the stage where as little harm as possible could be done to an organisation’s reputation (Hood et al., 2009). In a staged retreat crisis response,

organisations aim to minimise reputation damage by avoiding blame during a crisis (Hood et al., 2009). This strategy has three steps, varying in level of defensiveness as well as duration of crisis (Hood et al., 2009, p. 4).

First, in staged retreat, the management of an organisation is expected to deny the problem of the perceived crisis to minimise reputation damage (Hood et al., 2009, p. 4). Second, at a later stage, if it is not tenable to deny the problem of the crisis any longer,

organisations should admit that there is a problem, only they should not take responsibility for the crisis (Hood et al., 2009, pp. 4-5). Third, again one stage later, if it is not feasible to deny any form of responsibility for the crisis, an organisation should, besides admitting the

problem, also take responsibility for the crisis (Hood et al., 2009, pp. 4-5).

These different levels of responsibility an organisation takes in a staged retreat approach, have implications for its reputation level (Dean, 2004). In this study, the first stage will not be tested, since it will be clear in this design that a problem affected an organisation. To test the effects of denying or taking responsibility for the perceived crisis at different stages of an organisational crisis, I choose the second and the third stage of the staged retreat crisis response in this thesis.

2.16 Experimental studies of crisis communication

Knowing the response strategies which will be tested in this thesis, the spectre of experimental studies of crisis communication should be explored to view what has already been researched experimentally concerning crisis communication studies. As the field of crisis communication developed during the 1990s, increasingly more research has been devoted to experimental studies of crisis communication. The following studies indicate the variety of experimental studies which have been executed already.

Arpan and Roskos-Ewoldsen (2005) tested experimentally the influence of organisations providing people first with information about crises before the media does.

(23)

23

They found a strong positive relation between an organisation providing information first over the media in terms of control over the crisis and having a more positive image about the organisation (Arpan & Roskos-Ewoldsen, 2005).

Huang, Lin and Su (2005) tested crisis communication styles in Taiwan by letting managers of Taiwans’ top-500 organisations rank actual experiences of handling crises. This way, they built a model of effective crisis communication styles valid for the Taiwanese context (Huang, Lin, & Su, 2005).

Dean (2004) investigated experimentally the role of responsibility within crisis responses compared to crisis responses whereby only the response itself was visible. His results state that organisations taking responsibility were ranked with higher organisation reputation, regardless of their crisis response. Dean (2004) argues that this means the actual crisis response is subordinate to the role of responsibility within crisis responses.

These experimental studies regarding information transfers of organisations in relation to the media (Arpan & Roskos-Ewoldsen, 2005), crisis communication styles in Taiwan (Huang et al., 2005) and the role of responsibility within crisis responses (Dean, 2004) show the variety of experimental crisis communication studies which have been executed already.

The experimental study in this thesis will focus on testing multiple perceptions of

stakeholders on crisis response strategies compassion and apology. Furthermore, the second and the third stage of the staged retreat crisis response will be tested. This I will do when an organisational crisis unfolds in the public-private context of health insurance companies, which will be my contribution to the experimental studies of crisis communication.

2.17 Hypotheses

In this paragraph, the hypotheses for this thesis will be outlined, which are derived from the literature review. Starting with Marcus and Goodman (1991), who put crisis communication strategies into two main groups, respectively accommodative and defensive crisis responses. Accommodative crisis responses are characterised as responses whereby an organisation takes responsibility and/or further action, whereas defensive crisis responses are defined as responses which deny there is a problem and/or deny taking responsibility for the perceived crisis (Marcus & Goodman, 1991).

(24)

24

Regardless whether crisis responsibility is denied or taken, both crisis response strategies can be effective compared to not providing stakeholders with any form of crisis response (Coombs, 2006). However, defensive crisis response strategies are less effective if the responsibility for a crisis becomes (more) visible, making defensive crisis responses a risk compared to accommodative crisis responses (Benoit, 1995). To research these statements, my first and second hypotheses will be:

H1: Crisis response which includes taking responsibility will lead to higher organisation reputation than no crisis response

H2: Crisis response where an organisation denies responsibility, will lead to lower organisation reputation than crisis response taking responsibility

Scholars in crisis communication often emphasise that crisis response strategies should be more accommodative in order to meet crisis responsibility (Coombs & Schmidt, 2000). Accommodative crisis response strategies mean that the communication strategy should stress helping victims (Coombs, 1999, p. 129). Compassion is a crisis response strategy which prioritises helping victims (Coombs, 2006). The experimental study of Siomkos and

Shrivastava (1993) finds that having compassion for victims in crisis response strategies has a positive outcome on organisation reputation. Therefore, my third hypothesis will be:

H3: Crisis response which includes taking responsibility together with compassion will lead to higher organisation reputation than crisis response which includes taking responsibility

Coombs & Schmidt (2000) argue that helping victims is done most effectively when, besides focusing primarily on the needs of victims, the communicator takes full responsibility and focuses primarily on the needs of victims (Coombs & Schmidt, 2000). The apology crisis response practice is the only crisis response strategy where full responsibility for the crisis is accepted by its communicator (Coombs, 2006).

The experimental study of Bradford & Garrett (1995) on organisation reputation management assumes that the acceptance of full responsibility in apologies leads to higher organisation reputation. Following Dean (2004), who stated that actual crisis responses are

(25)

25

subordinate to the role of responsibility within crisis responses, this implies that apology would have the highest positive relation to organisation reputation. To test whether these assumptions are valid, my fourth and fifth hypotheses will be:

H4: Crisis response which includes taking responsibility together with apology will lead to higher organisation reputation than crisis response which includes taking responsibility

H5: Crisis response which includes taking responsibility together with apology will have a larger positive effect on organisation reputation than crisis response containing taking responsibility together with compassion

Changing your crisis communication strategy at various stages of the crisis could influence the level of organisation reputation. In different stages of a crisis, stakeholders require different approaches (Sturges, 1994). Therefore, an organisation should customise their crisis communication to the needs of their stakeholders as a crisis unfolds (Sturges, 1994, pp. 297-298).

On the other hand, a change in crisis communication strategy puts the legitimacy of an organisation in danger, questioning reasons for an organisation to change previous crisis communication strategies (Coombs & Holladay, 2011). As shown in the two examples in paragraph 1.4 Societal relevance, switching crisis responses had a negative influence on how stakeholders perceived the organisation. Therefore, my sixth, seventh and eighth hypotheses will be:

H6: Crisis response where an organisation denies responsibility first before taking responsibility, will lead to lower organisation reputation than crisis response taking responsibility

H7: Crisis response where an organisation denies responsibility first before taking responsibility, will lead to lower organisation reputation than crisis response denying responsibility

(26)

26

H8: Crisis response which includes taking responsibility together with apology will have a larger positive effect on organisation reputation than crisis response containing taking

responsibility together with compassion, even when in both crisis responses the responsibility was denied first

(27)

27

3. Method

In this chapter, I will explain the experimental design I used in this thesis. After having outlined the general advantages of this design, I will analyse which contributions this research makes in terms of reliability and validity. After that, I will show the relationship between the variables used in this study, which I will display in two figures.

When this relationship is presented, I will move on with outlining the independent variable crisis response type. In this paragraph, I will state which conditions I worked with and how the hypothetical scenario has been designed in this thesis. Thereafter, I will continue with clarifying what is understood by the dependent variable organisation reputation. In this paragraph, I will demonstrate which dimensions and indicators I used to operationalise organisation reputation.

In the end of this chapter, I will state the process that has been followed for the data collection. This includes mentioning which requirements for participants of the experiment I maintained. Furthermore, I will address which survey tool I used and how I distributed this survey. I will conclude by arguing which program I utilised for analysing the obtained data and which tests I executed with this program.

3.1 Experimental design

In this thesis, I used an experimental design to investigate the effects of crisis response types on organisation reputation. One of the advantages of experimental designs is that

experimenters are given high experimental control over the measurement and recruitment, which enables the experimenter to measure the causal effects of the independent variable on the dependent variable (McDermott, 2002, p. 39).

Another advantage of experimental design is randomisation. This means that participants are randomly assigned to the experimental group and the control group, which minimises other factors which may influence the experiment (Bryman, 2012, pp. 51-52). This way, experiments allow confidence regarding causal inferences about relationships among the variables of interest, thus experiments provide the ability to derive causal inferences

(28)

28

manipulated in this study, the crucial element of standardisation in experimental design has been met (McDermott, 2002, p. 33).

Experimental research is suitable to study the field where crisis communication’ case-studies vary in their characteristics (McDermott, 2002). Furthermore, evidence is needed to support strong theoretical causal claims, however the traditional methods of inquiry have provided us with incomplete coverage (McDermott, 2002) of the effect crisis communication could have on organisation reputation up so far. This means that best practices in crisis response are still to be found within the field of crisis communication, whereas multiple case-studies have already been examined. Thus, it has been an opportunity for experimental design, being a relatively new design in the field of crisis communication, to examine best crisis response practices.

3.1.1 Reliability and validity

In social sciences, the best way to value research is to evaluate reliability and validity of the research (Bryman, 2012, p. 46). Reliability questions whether research is repeatable (Bryman, 2012, p. 46). Due to the clear procedures of experimental design in this research, replication to validate the results could be done easily compared to other designs, therefore this study has a high degree of reliability.

In terms of validity, a distinction could be made between internal validity and

generalizability (Bryman, 2012, p. 43). In case of internal validity, it could be argued that this experimental research is valid in terms of causal inference due to the control group (Bryman, 2012, p. 47). Because of this, causality between the independent and dependent variable could be measured thoroughly. Furthermore, compared to non-experimental methods, survey

experiments allowed that causality in hypotheses was measured with large internal validity (McDermott, 2011, pp. 38-39). Last, a factor analysis was executed to investigate the

dependent variable organisation reputation, see paragraph 4.3 Factor analysis. This resulted in the factor analysis with Cronbach’s α of .92 which was used in this study, indicating that the reliability of the concept organisation reputation is high. This contributed positively to the internal validity of this study.

In case of generalisability, one advantage of this study is that this experiment is based on a hypothetical scenario which aims to derive best crisis response practices, see paragraph

(29)

29

3.2.1 Hypothetical scenario. Without having a previous organisational reputation status and without a bias towards organisational brand in case of this hypothetical organisation, which are two elements which normally influence reputation management (Piranfar & Matthews, 2008), the effects of this study could influence other sectors more than experiments which are based on existing companies. Apart from those other sectors, this hypothetical scenario makes this study applicable within the field of health insurance companies, which has a positive effect on its generalisability.

3.1.2 Variables

In this project, the effects of one independent variable were tested on one dependent variable. The independent variable in this study has been crisis response type. The dependent variable which was studied has been organisation reputation. Figure 1 shows a simplified overview of the relationship between the independent variable crisis response type and the dependent variable organisation reputation.

Figure 1. Relationship between independent and dependent variable.

To outline a more detailed overview of the independent variable, it can be set out to provide insights into the different levels of this variable. These levels correspond to the different conditions that were tested in this study, see Table 5.

The arrows in Figure 2, which go up and down from the independent variable crisis response type, indicate that in this study, a list of factors together lead to one condition. When for example an apology is offered for an organisational crisis as in the lower left of Figure 2, then taking responsibility for the organisational crisis preceded the apology. This expands the simplified overview of Figure 1, which results in Figure 2:

(30)

30

Figure 2. Different levels of the independent variable and its relationship with the dependent variable.

3.2 Crisis response type – independent variable

Now the different levels of the independent variable have been described, I will

continue outlining the choices I made for setting up an experiment. I embedded an experiment in a survey, in which participants evaluated hypothetical scenarios of an organisational crisis. In this experiment, participants were assigned to one of the conditions in which they read a vignette with a particular scenario followed by questions. Vignettes are texts that are

presented to the respondents, manipulating different levels of the independent variable (Evans et al., 2015, p. 163). In this study, the independent variable crisis response type had multiple levels representing different types of responses, see Table 5. Therefore, participants of this experiment were presented with one of the vignettes which entailed one of the levels of the independent variable.

For this experiment, eight vignettes of crisis response types were used to measure the effects on the dependent variable organisation reputation. This way, the independent variable crisis response type was split into eight scenarios of crisis response. These eight vignettes differed in their characteristics. The first vignette had served as a control condition without

Apology

Responsibility

Denial

Crisis response type

Responsibility

Apology Compassion

Organisation reputation Compassion

(31)

31

any experimental manipulation. This way, it had been possible to measure the other

conditions in relation to the control condition (Bryman, 2012, p. 51), which allowed making substantiated statements about the effects of crisis response types on organisation reputation.

Vignettes two up to and including eight all consisted of an element of crisis responsibility, see Table 5. In some vignettes, the responsibility for the crisis was taken (condition 2; condition 3; condition 4), whereas in other vignettes, the responsibility for the crisis was denied (condition 5; condition 6; condition 7; condition 8). Furthermore, in three vignettes, the responsibility was initially denied, however taken at a later stage (condition 6; condition 7; condition 8). Finally, some vignettes either contained an element of apology (condition 3; condition 7) or compassion (condition 4; condition 8). Table 5 provides an overview of the conditions used in this study:

Table 5

Conditions of the experiment

Conditions Crisis response types Condition 1 Control (no elements) Condition 2 Taking responsibility

Condition 3 Taking responsibility + apology Condition 4 Taking responsibility + compassion

Condition 5 Denial

Condition 6 Initial denial, then taking responsibility

Condition 7 Initial denial, then taking responsibility + apology Condition 8 Initial denial, then taking responsibility + compassion

This study was conducted with a between-subjects design, a design where participants of the survey experiment were randomly assigned to one of the conditions (Evans et al., 2015, p. 163). This way, the level of perceived organisation reputation was compared between the conditions. Randomisation of assigning conditions is therefore helpful to eliminate sources of systematic variation, which ensures that systematic variation between the experimental conditions is due to the manipulation of the independent variable (Field, 2013, p. 18).

(32)

32 3.2.1 Hypothetical scenario

In this research, I used a hypothetical scenario of an organisational crisis in a fictive organisation. Due to the absence of an organisations’ reputation level and previous crisis responses practices in case of a hypothetical organisation, respondents’bias towards this organisation was minimalised. Furthermore, after a general introduction of the experiment, all respondents evaluated on the same case, regardless of the condition they were randomly assigned to.

In this case, a health insurance company named Mijn Zorg, leaked the data of 40.000 individual customers to private healthcare companies, such as hospitals, dentists and

specialists. Respondents were asked to imagine that Mijn Zorg was their own insurance company. While an organisational crisis unfolded, all private customer information became available to third parties which were not directly related to the organisation. To deal with this organisational crisis, crisis responses were presented, aiming to have a positive influence on organisation reputation.

From this point onwards, the vignettes started to differ, see Table 5. When the organisation admitted responsibility, the participants read that: ‘’In a first reaction, health insurer Mijn Zorg claims responsibility for the data breach that took place within the organisation’’, which was the first sentence that was presented in some vignettes (condition 2; condition 3; condition 4). When the organisation denied responsibility, the participants read that: ‘’In a first reaction, health insurer Mijn Zorg denies that they are responsible for the data breach that took place within the organisation’’, which was the first sentence in other vignettes (condition 5; condition 6; condition 7; condition 8).

Most of these vignettes were followed by: ‘’Over time, Mijn Zorg claims

responsibility for the data breach’’ (condition 6; condition 7; condition 8). Furthermore, two vignettes ended with: ‘’Mijn Zorg asks everyone who has been hit by the data breach for forgiveness’’ (condition 3; condition 7), whereas two other vignettes concluded: ‘’Mijn Zorg offers compensation to anyone who is affected by the data breach’’ (condition 4; condition 8). For a full overview of all the vignettes, see Appendix 7.1 Crisis in organisaties – Qualtrics Survey.

(33)

33 3.3 Organisation reputation – dependent variable

Measuring the beliefs and perceptions of an audience after providing standardised, controlled and carefully constructed context, is considered one of the biggest advantages of a survey (Finch, 1987, pp. 105-106). The survey in this study consisted of sixteen questions in total, of which the first eight questions measured the dependent variable organisation

reputation. On a 7-point Likert-type scale, ranging from fully disagree (1) to fully agree (7) with a neutral (4) option, participants were asked to rate eight dimensions which contribute to organisation reputation. For a full overview of these dimensions, see Table 6.

3.3.1 Operationalisation

The outlined dimensions are measured by indicators. Indicators are non-abstract phenomena which are in this study present in the form of questions (Bryman, 2012, pp. 164-165), attempting to capture the perceived organisation reputation of stakeholders. This way, the abstract concept of organisation reputation is made tangible in this thesis, which enables the concept of organisation reputation to be operationalised. Operationalisation of the concept organisation reputation means ‘’devising measures of the concept in which the researcher is interested’’ (Bryman, 2012, p. 161). In Table 6, the operationalisation of the concept

organisation reputation is outlined, showing the dimensions and indicators of the concept:

Table 6

Dimensions and indicators which are part of organisation reputation

Dimensions Indicators

1. General impression In general, I have a positive impression of this organisation

2. Preferability I would prefer an organisation such as health insurance company Mijn Zorg 3. Trust I would trust an organisation such as health insurance company Mijn Zorg 4. Positive feeling I have a good feeling about health insurance company Mijn Zorg

5. Good managing I would argue that health insurance company Mijn Zorg is well managed 6. Competent

employees

I would argue that health insurance company Mijn Zorg is an organisation that has competent employees

7. High customer treating service

I would argue that health insurance company Mijn Zorg has a high standard in the way they treat their clients

(34)

34 3.4 Data collection

Now the experimental design and both the independent and dependent variables are outlined, multiple phases of data collection will be explained. In general, there were two requirements for respondents to participate in this survey. First, since the survey was in Dutch, they needed to understand the Dutch language in order to answer the questions. Second, the minimum age for filling in this survey had been eighteen, since this corresponds with the minimum age people are obligated to have a health insurance in some countries, for example in the Netherlands. These requirements were mentioned when the survey was distributed, thus participants of this survey were aware of these two requirements before starting the experiment.

For the distribution of this survey, survey tool Qualtrics was used. For a full overview of the complete survey, see Appendix 7.1 Crisis in organisaties – Qualtrics Survey. Since this survey was distributed in Dutch to reach more participants, the Qualtrics survey is added as its original. Due to the short amount of time a participant needed for this survey, no incentives were provided to participants of this experiment. In the end of the experiment, all participants had the opportunity to leave additional questions and remarks about the experiment, only nobody had made use of this option. Hereafter, respondents were thanked for their cooperation and completion of the survey.

To analyse the obtained data, data analysis program SPSS was used. In SPSS, multiple tests were executed to measure the difference in effects of crisis response types on

organisation reputation. This way, the eight hypotheses based on the literature review were tested, which were either verified or falsified. Thus, to do that, two one-way between groups analyses of variance (ANOVA) and two independent samples t-tests were performed in SPSS.

(35)

35

4. Results

In this chapter, I will begin with outlining information about the sample characteristics of this thesis. I asked four demographic questions in my survey, which will be presented with its limitations. Furthermore, I executed four manipulation checks, whose significance together with an overview of the conditions that differed significantly from each other I will show.

To evaluate whether the items that were tested in this study measured organisation reputation, I drafted a factor analysis. Thereafter, I will demonstrate the results of the hypotheses tests, ending the paragraph with an overview of the hypotheses and whether support for them was found. I will conclude the results with mean scores of all conditions on organisation reputation to find out which crisis response type scores the highest on

organisation reputation.

4.1 Sample characteristics

In this paragraph, I will provide background information of respondents including their sex, age, education level and current work situation to minimise subjectivity of this research (Bryman, 2012, pp. 45-47). The distribution of this survey had been executed

through various channels to obtain the highest response rate. This resulted in a total sample of 295 respondents. One respondent was excluded from the analysis due to giving only nonsense answers and therefore this respondent’s request to be excluded from the study. This resulted in a total sample with N = 294, which was used for analysis.

To start with, respondents were recruited via personal links (n=215). These were respondents who were recruited via WhatsApp, Blackboard or face-to-face conversations. Furthermore, respondents were recruited via social media (n=75). These included participants recruited via Facebook and LinkedIn. Last, respondents were recruited via e-mail (n=4). These consisted of a group of people who requested this survey via mailing.

In this survey, I asked four questions about the demographics of the participants. Looking at these demographics, the answers of 76 participants were missing. Since the demographics were the last part of this survey, people could have left after partially filling in the questionnaire. This could also mean they found the survey to hard or they misunderstood the questions, causing respondents to leave prematurely. Although the survey only took less

(36)

36

than five minutes on average, the estimated completion time of the survey could have possibly been a burden for respondents to complete this survey.

Because of the 76 missing values, N = 218 were the respondents who filled in the

demographics. Out of these respondents, 80 male-, 137 female- and 1 other sex participant(s) took part in this survey. Having substantially more female participants than male participants in this study, causes one of the limitations of this research. The age of the participants ranged from 18 to 63, M = 27.92, SD = 10.81. Most of the respondents were 30 years or younger (83.9%), M = 23.6, SD = 2.29.

This causes another limitation of this study due to a relatively young sample. This young sample could be explained since I recruited respondents within the reach of my own network, consisting mostly of peers with the same age.

This outcome is reflected in the next chart about current work situation of the respondents too. Most of the

respondents were students (56.4%), followed by fulltime workers (25.7%) and parttime workers (11.5%). This

represented 123 students, 56 fulltime workers and 25 parttime workers respectively.

Not many respondents were independent workers (N = 6), job seekers (N = 6), retired (N = 1) or had another current work situation status than available in the answer categories (N = 1). This marks a limitation of this study, having many students in its sample size, which does not reflect the demographics of the Netherlands for example.

Figure 3. Gender of participants.

(37)

37

Concerning the highest obtained education level, most of the respondents were highly educated in this study, respectively N = 147 (67.4%) for WO and N = 42 (19.3%) for HBO.

Furthermore, other education level categories consisted of N = 14 (MBO), N = 13 (HAVO/VWO) and N = 2 (Other education level). Therefore, since highly educated participants are more

represented, the outcomes of this study are possibly biased.

4.2 Manipulation checks

One of the advantages of a survey experiment is doing manipulation checks to verify whether the manipulations were effective (Sniderman, 2011, pp. 263-263). In this study, I asked four statements to check if participants understood the text as intended, which were measured on a 7-point Likert-type scale (see Table 7). Thereafter, I tested each manipulation by a one-way between groups analysis of variance (ANOVA).

Table 7 shows an overview of the manipulation checks used in this study and the significant differences of the ANOVA. In Table 7, statistically significant differences for all four manipulations are visible, indicating that all four manipulation checks were valid.

Table 7

Manipulation checks

Statements Significance of ANOVA

1. Mijn Zorg immediately claimed responsibility for the data breach that took place within the organisation

F (7, 211) = 93.796, p <.001, R2 = .757

2. In an initial response, Mijn Zorg denied responsibility for the data breach that took place within the organisation

F (7, 211) = 48.481, p <.001, R2 = .617

Communicate your way out of a reputation crisis: An experimental research about crisis response types and their influence on organisation reputation