Master’s Thesis
The Effects of Journalistic News Frames on Public Attitudes
towards Whistleblowing
University of Amsterdam
Graduate School of Communication
Master’s Programme Communication Science
Author: Vilma Nurmela Student ID: 11181397
Supervisor: Andreas R.T. Schuck Date of Completion: June 30th 2017
1 Abstract
Research shows that what we see, read and hear on the news is always handpicked by someone, and constructed frames shape the way people form their opinions and build attitudes towards a variety of topics. The topic of whistleblowing is particularly relevant as it has been prominent in the media in recent years, predominantly due to highly publicized cases such as those of Edward Snowden and WikiLeaks, but there is no prior evidence as to how news frames regarding
whistleblowing affect public attitudes. This thesis fills this gap as it offers empirical insights into these effects through the use of an online experiment (N=162). Based on framing theory, this thesis provides evidence that journalistic news frames, namely those of Courage and Self-interest, affect the attitudes that the public have towards whistleblowing. Findings discovered
that people exposed to the Courage frame had more positive opinions and attitudes towards whistleblowing, and were more likely to act with regard to the issue, than people exposed to the Self-interest frame. Additionally, while self-perceived morality and political knowledge did not
moderate this relationship, positive emotions did indeed function as a significant mediator. Exposure to the Courage frame increased positive emotions which in turn resulted in more positive opinions and attitudes as well as higher behavioural intentions and actual behaviour.
2 Introduction
Complete objectivity in news reporting is unreachable (Landert & Miscione, 2017). Journalists are responsible for providing an overall context for a variety of issues and they are both consciously and inadvertently faced with selection processes whenever reporting
information (Schuck & de Vreese, 2006). Particularly in the current media environment, in which media representations shape people’s perceptions on issues (Boyle, Schmierbach,
Armstrong et al, 2006), and when keeping or omitting information can lead to a certain bias, it is crucial to understand the effects of such selection processes (Landert & Miscione, 2017).
Nelson, Oxley & Clawson (1997, p.221) define this procedure of framing as “the process by which a communication source constructs and defines a social or political issue for its
audience”. Framing can portray the same topic in numerous ways, and as it places certain aspects of reality in the forefront while pushing others to the background, certain judgements are implied (Lecheler, Schuck, & de Vreese, 2013). News frames have a major impact on shaping public opinion and people’s understanding of an array of issues (Schuck & de Vreese, 2006). While studies into public opinion have recognized framing as an important factor, the exact effects of news frames are not well understood as few have discovered how frames influence public opinion (Nelson, Oxley, & Clawson, 1997). Therefore, further research into the effects of framing is essential (Landert & Miscione, 2017).
One relevant issue that has barely been touched upon within framing theory is the topic of whistleblowing (Wahl-Jorgensen & Hunt, 2012), which refers to “the action of reporting perceived organizational wrongdoing, typically from within the organization” (Heumann, Friedes, Redlawsk, Cassak, & Kesari, 2016). The topic is highly relevant as it has been gaining
3
increased media attention, particularly in the wake of WikiLeaks (Di Salvo & Negro, 2015) and the intense debate surrounding Edward Snowden (Heumann et al., 2016). Reports show that people are split in their opinions on the actions of such whistleblowers (Public Concern at Work, 2011; Tevani & Grodzinsky, 2014), but despite the issue being continuously contested, apart from the occasional opinion poll, “little is known about how the public perceives the
whistleblower” (Heumann et al., 2016, p.7). Furthermore, academic studies about
whistleblowing in the political sphere are practically non-existent and no evidence exists as to how the aforementioned news frames affect such varying public attitudes (Harwood, 2016, Wahl-Jorgesen & Hunt, 2012).
To fill the wide gap in research, this study is the first to investigate whether journalistic news frames have an effect on public attitudes about whistleblowing. An experiment was conducted that compares dominant issue frames specific to whistleblowing, namely the positive Courage frame and negative Self-interest frame, which were discovered in a content analysis of
newspaper coverage related specifically to whistleblowing (Vogele & Baudermann, 2016). As it is unlikely that frames will have a universal effect on all people (Entman, 1993), this study also investigates the roles of possible moderators and a mediator.
The structure of this thesis is as follows. First, a theoretical framework is provided with focus on the central concepts of framing theory and whistleblowing, and the research question is introduced along with the hypotheses. Subsequently, an explanation of the methodological approach followed by results are provided as well as a discussion of the central findings. The thesis finishes with concluding remarks, an analysis of the contributions and limitations of the study and suggestions for further research.
4 Framing
Mass media have been the subject of substantial research in the field of communication (Nelson, Oxley & Clawson, 1997). Framing, priming and agenda-setting are often examined under one broad category and studied with similar approaches, but as key theoretical differences exist, it’s important to distinguish between the three (Scheufele, 2000). Agenda-setting looks at why certain situations come to command attention, while priming focuses on how making certain characteristics of a person or an issue more noticeable than others affects our judgement (Chong & Druckman, 2007). Framing instead refers to the procedure of constructing an event or an issue in a particular way for an audience (Valkenburg, Semetko, & de Vreese, 1999), and hence
focuses on how changes in precise media content rather than mere coverage affect public opinion (Nelson & Oxley, 1999). All in all, while each theory looks into media effects, agenda-setting and priming are concerned with telling audiences what issue or characteristics to think about, whereas framing assumes that subtle changes in content can affect people’s interpretation, thus implying how to think about issues (Chong & Druckman, 2007; Scheufele, 2000).
This thesis focuses on a single issue, so interest lies only with framing which, essentially, involves selecting certain aspects of an issue or an event and making them more meaningful or memorable to audiences (Chong & Druckman, 2007). This can be achieved, for instance, by repetition, by placing certain information strategically, or by associating certain aspects with culturally familiar symbols (Entman, 1993). Even a single sentence can have great consequences and the probability of an issue being stored in people’s memories is heightened (Entman, 1993).
Research shows that the choices journalists make when covering a story can result in considerably different portrayals of exactly the same events, which in turn usually affect the way
5
people weigh issues (Nelson, Clawson, & Oxley, 1997). This process isn’t always intentional but often results in differing views nonetheless (Lee, McLeod, & Shah, 2008). Still, it is crucial to acknowledge that the way audiences interpret news media, and symbolic devices within them, depends on their pre-existing memories and attitudes (Pan & Kosicki, 1993). News frames can’t therefore be reduced merely to the information they provide but, by activating information already stored in people’s memories, framing allows for shortcuts to news consumption (Pan & Kosicki, 1993), thus having great potential to influence public consciousness (Entman, 1993).
Traditionally, because “framing has important implications for political communication” (Entman, 1993, p.55), numerous interest parties ranging from journalists to interest groups and lobbyists try to get their agenda covered, thus making the competition for dominant frames in news media fierce (Nelson & Oxley, 1999). Journalistic news frames work as bridges between official elite discourse about any problem and public understanding of the matter in question, which is why those with power dedicate large amounts of time and effort toward influencing the way information is presented in the media, instead of just what information gets printed (Nelson,
Oxley et al., 1997). Ultimately, however, journalists and news organizations choose what to include and what to leave out, giving them ultimate power to compose selective versions of complex problems (Feinholdt, Schuck, Lecheler, & de Vreese, 2016).
Whistleblowing
A whistleblower refers to an insider in an organization who has reasonable belief to challenge its policy or practice in accordance with an ethical mandate with the intention of taking action to address perceived wrongdoing (Heumann et al., 2013-14; Brown, Lewis, Moberly, & Vandekerckhove, 2014). Whistleblowing is crucial in keeping the powerful accountable for their
6
actions (Brown et al., 2014), but while numerous laws have been passed to protect
whistleblowers from retaliation, those taking action rarely succeed (Heumann et al., 2013-14). The whistleblowing charity Public Concern at Work (2016) reported overall negative outcomes since 4 in 5 whistleblowers were victimized, dismissed, bullied or forced to resign for their actions, which could frighten even the most moral citizens to act (Park, Blenkinsopp, & Park, 2014). This being said, the term whistleblowing itself is inherently merely descriptive without normative evaluations, but having gained increased media attention, the different frames surrounding whistleblowing stories are what give it various angles (Brown et al., 2014). Research on whistleblowing tends to focus only on specific cases, but two content analyses that have looked at newspaper coverage of general whistleblowing frames have
discovered mixed results (Di Salvo & Negro, 2015). Wahl-Jorgesen & Hunt (2012) found that, in contrast to negative coverage in the American press, most British newspapers frame
whistleblowing in a neutral or positive way, however, coverage is increasingly personalized and focuses only on those cases that fit prevailing news trends. Vogele & Baudermann (2016), on the other hand, discovered two dominant yet opposing frames: one highlighting the whistleblowers’ actions as courageous and another focusing on the whistleblowers self-interest.
When it comes to specific cases, WikiLeaks has been a key figure in shifting discussion on the topic of whistleblowing, and its impact on framing has been significant (Lynch, 2013). Framing of WikiLeaks tends to be mixed as some news organizations praise it for having national interests at heart when exposing wrongdoing to citizens while others say it harms national security (Hindman & Thomas, 2014). Right wing organizations as well as mainstream press, particularly in the US, tend to frame WikiLeaks and its face, Julian Assange, as threats whereas the left and smaller publications, particularly in Europe, frame them as heroes (Handley
7
& Rutigliano, 2012; Landert & Miscione, 2017; Ottosen, 2012). Another punctuating phase in the debate over whistleblowing was when Edward Snowden became a global phenomenon by revealing details of the National Security Agency’s mass surveillance program (Chadwick & Collister, 2014). Interestingly, social media discussions often praise Snowden’s actions as being heroic, whereas legacy media tend to frame him as a traitor (Qin, 2015). There are similarities between the cases of Assange and Snowden as focus is often on their personal lives instead of the leaks themselves (Chadwick & Collister, 2014). Coverage of Assange can be positive but ranges to very negative when descriptions such as “volatile” or “insufferable” are used (Luther & Radovic, 2014). Similarly, Snowden is at times referred to simply as a whistleblower, but the terms hero, victim and villain also make appearances (Di Salvo & Negro, 2015).
In addition to these content analyses on framing, a handful of surveys have looked into public perceptions about whistleblowing. There is widespread disagreement as to whether whistleblowing is unethical or beneficial for public interest (Harwood, 2016). While some see whistleblowers as courageous employees who ought to be praised for their actions, others view them as traitors or spies (Tavani & Grodzinsky, 2014). Even though recent studies show positive views are increasing, the majority of data indicates that over half of people see whistleblowers as criminals or mischief makers (Public Concern at Work, 2011) or as having acted because of holding a grudge, to protect their own company or for wanting personal gain (Heumann et al., 2016). When it comes to WikiLeaks and Snowden, attitudes are relatively negative overall. In total, 60% of the people who know about WikiLeaks releasing classified U.S. documents say it causes harm (Pew Research Center, 2010), and while 66% of citizens say Assange is voluntarily avoiding lawful arrest, only 16% say he should be allowed to leave his asylum (YouGov, 2016).
8
With Snowden, over half of the U.S. population think his leak harmed public interest, while only a third believe it served it (Pew Research Center, 2013).
These studies don’t give a conclusive picture on whistleblowing as most are conducted merely on individual cases (Heumann et al., 2016). Whereas overall framing results are rather mixed and opinion polls suggest rather negative views, no studies have looked into the
connection between such news frames and public opinion, hence conducting more research into the formation of people’s attitudes is crucial (Park et al., 2014). This thesis makes its
contribution by aiming to answer the research question:
In what ways do journalistic news frames affect public attitudes towards whistleblowing?
News frames on whistleblowing
Scholars make a distinction between so-called generic and issue-specific frames
(Feinholdt et al., 2016). By definition, generic frames exceed thematic or cultural limitations and are applied to a wide set of topics (Valkenburg et al., 1999), whereas issue-specific frames are only relevant for certain events and issues (de Vreese & Boomgaarden). This thesis looks at the latter and focuses on frames specific to whistleblowing that Vogele and Baudermann (2016) identified in their content analysis of whistleblowing coverage in German language media. The most dominant whistleblowing frame, that of Courage, was used in 42% of news coverage. It portrays the whistleblower as a victim who acts in the best interest of the public, while coverage of people hiding the secret is negative. The second most dominant frame, that of Self-interest, fundamentally opposes the Courage frame as it evaluates the whistleblower in a negative way while secret keepers are portrayed as victims instead. A third frame, that of Wrongdoing, was
9
also discovered but not used in this study as it places the least focus on the whistleblower, and is rarely used in news coverage.
In addition to differences in generic and issue-specific frames, framing studies also make a distinction between neutral frames and ones that are valenced (de Vreese & Boomgaarden, 2003). Valence refers to frames that are implicitly good or bad, or carry opposing positive and negative elements, and by doing so aim to gain support for or against an issue (Lecheler, Schuck et al., 2013). The two frames described above, those of Courage and Self-interest, are classified as valenced as they inherently oppose one another, and based on previous studies into the effects of opposing news frames, public attitudes can be expected to differ depending on which frame people are exposed to. Nelson, Clawson et al. (1997), for instance, discovered that people’s tolerance on the Ku Klux Klan depended on frame exposure: those subjected to a free speech frame portrayed higher approval ratings whereas those exposed to a frame that described the KKK as disrupting public order expressed less approval. Similarly, multiple studies have
discovered that exposure to positive frames regarding different EU-related issues leads to higher support for it than exposure to negative frames (de Vreese & Boomgaarden, 2003; Lecheler, Schuck et al., 2013; Schuck & de Vreese, 2006).
This study also differentiates between opinions and attitudes. Opinions refer to certain facts presented, and in this thesis relate directly to the stimuli, whereas attitudes refer to whether or not people approve of certain kinds of behaviour which relate to more firmly existing
predispositions they hold towards a topic (Domfeh & Bawole, 2011). In line with previous findings on framing introduced above, this thesis expects participants’ opinions and attitudes towards whistleblowing to differ based on which of the two frames, Courage or Self-interest, they are exposed to. Hence, the following hypotheses were formulated:
10 H1a: People exposed to the Courage frame have more positive opinions and attitudes
towards whistleblowing than people exposed to the Self-interest frame.
H1b: People exposed to the Self-Interest frame have more negative opinions and attitudes
towards whistleblowing than people exposed to the Courage frame.
Additionally, research shows that exposure to controversial content doesn’t only affect attitudes, but increases the likeliness to engage in topic-related discussion and behaviour (Chen & Berger, 2013). While some studies have found negative information to increase participation, this study follows in the footsteps of Berger & Milkman (2012) who discovered that exposure to positive content in fact increases likeliness to engage in actual behaviour and sharing of
information. Hence, the following hypothesis was formulated:
H1c: People exposed to the Courage frame express higher behavioural intentions and
engage in actual behaviour regarding whistleblowing more than people exposed to the Self-interest frame.
Moderators
This thesis also looks into indirect effects since by including both moderators and a mediator in the same analysis, prominence is given to their dynamic roles (Baron & Kenny, 1986). Frames often operate indirectly via moderators by activating characteristics already at people’s disposal, and such qualities often influence the direction or strength of an effect (Nelson, Oxley et al., 1997).
11
Framing usually depicts political issues in a way that clashes specifically with one’s existing moral values (Lee et al., 2008). Morality here refers to a code of conduct that an individual accepts, which can affect how various political issues are understood (Giammarco, 2016). With whistleblowing, one’s moral orientation is a crucial determinant when judging the person blowing the whistle (Park et al., 2014), and observers consider it most important to release information for the right moral reasons (Domfeh & Kawole, 2011). This thesis expects people with higher morality to be more affected by the news frames, thus the following
hypotheses were formulated:
H2a: People with higher self-perceived morality who are exposed to the Courage frame
have more positive opinions and attitudes towards whistleblowing than people with lower self-perceived morality.
H2b: People with higher self-perceived morality who are exposed to the Self-interest
frame have more negative opinions and attitudes towards whistleblowing than people with lower self-perceived morality.
H2c: People with higher self-perceived morality who are exposed to the Courage frame
express higher behavioural intentions and engage in actual behaviour regarding whistleblowing more than people with lower self-perceived morality.
Political knowledge
The concept of political knowledge is central to the field of political communication and has emerged as a significant moderator of people’s susceptibility to framing effects (Lecheler & de Vreese, 2012). Previous studies have discovered that it influences people’s reactions to various issues (de Vreese & Boomgaarden, 2003), yet results as to who are most affected are
12
mixed (Lecheler & de Vreese, 2012). Some argue that people with less knowledge are more vulnerable when exposed to frames due to an inability to counterargue (Schuck & de Vreese, 2006) while others claim that people with higher knowledge are more affected as they’re better equipped to process a message (Druckman & Nelson, 2003; Nelson, Oxley et al., 1997). This study expects the latter, hence the following hypotheses were formulated:
H3a: People with higher political knowledge who are exposed to the Courage frame have
more positive opinions and attitudes towards whistleblowing than people with lower political knowledge.
H3b: People with higher political knowledge who are exposed to the Self-interest frame
have more negative opinions and attitudes towards whistleblowing than people with lower political knowledge.
H3c: People with higher political knowledge who are exposed to the Courage frame
express higher behavioural intentions and engage in actual behaviour regarding whistleblowing more than people with lower political knowledge.
Mediator
It is not only the existence of an effect that is interesting, or its conditionality on
moderating factors, but growing amounts of studies have looked into the psychological processes that mediate such effects (Lecheler & de Vreese, 2012). Exploring a mediator refers to looking at “the intermediary causal mechanisms by which an independent variable influences the dependent variable” (Lecheler & de Vreese, 2012, p.186), or as Baron and Kenny (1986) put it more
simply, a mediator explains why certain effects occur. Emotions
13
Discrete emotions are a big part of framing as people access emotions from information they’re exposed to, which in turn influence their attitudes and behaviour (Nabu, 1999). While emotions can act in distinct ways individually, they can also be summarized into dimensions that they have in common, such as valence (Lecheler, Schuck et al., 2013). Empirical evidence shows that discrete positive emotions positively mediate attitudes and behaviour whereas discrete negative emotions are significant mediators of negative effects (Berger & Milkman, 2012; Lecheler et al., 2013). Since this thesis too uses frames with valence, the following hypotheses were formulated:
H4a: Being exposed to the Courage frame increases positive emotions which positively
affect opinions and attitudes on whistleblowing.
H4b: Being exposed to the Self-interest frame increases negative emotions which
negatively affect opinions and attitudes on whistleblowing.
H4c: Being exposed to the Courage frame increases positive emotions which positively
affect behavioural intentions and actual behaviour regarding whistleblowing.
Method
Procedure
To test the research question, an experiment was conducted. The method was chosen as many factors can be controlled for and it is thus “the superior way to establish causality in media effects research” (de Vreese & Boomgaarden, 2003, p.368). The experiment was conducted online between the 21st and 24th of May 2017 using the survey software Qualtrics. Participants
14
who provided informed consent were directed to the online questionnaire, which began with scales for moderating variables followed by random allocation to one of three articles created solely for the purpose of this study: one with a Courage frame, one with a Self-interest frame and a control group article. This was followed by manipulation check items, and scales for testing the mediator and dependent variables. The questionnaire finished with a debriefing where the
purpose of the study was explained further. See Appendix A for the complete questionnaire.
Sample
The population under study included adults above 18 years old. Participants were recruited online via email and Facebook primarily using the snowballing sampling method, and no incentive was provided for participating. In total, 242 people took part in the experiment, however 79 participants were excluded due to incomplete responses and one for being under the age of 18, resulting in a final sample of 162. Participants were randomly assigned to one of three articles: Courage frame condition (N=53), Self-interest frame condition (N=52) and the control group (N=57).
Participants represented 27 nationalities with the majority of people being Finnish (40.1%) and British (27.8%). There were more female participants (69.1%) than males (30.9%), and the majority (88.8%) had completed higher education with a minimum of a Bachelor’s degree. Most participants (42.6%) were in the age group 25-34, however age varied from 18-24 (29.0%) to 85 and older (0.6%). In terms of political views, people were slightly left-leaning (M=4.04, SD=2.24). Randomization was successful since there were no systematic differences between the conditions in terms of gender (²=3.18 p=.204), education (F(2,159)=1.40, p=.250), age (F(2,159)=.29, p=.746) or self-perceived left-right placement (F(2,159)=.14, p=.874).
15 Design
The design of this study was a one factorial post-test only, between-subjects experimental design with random assignment to one of three groups. Each participant was exposed to one news article related to a whistleblowing issue regarding the Canadian Foreign Ministry. See Appendix B for the three different versions of the stimuli.
Stimulus material
Three different news articles regarding whistleblowing were constructed, which allows
more control over details than when using already published texts (Feinholdt et al., 2016). All articles focused on a fictional Canadian civil servant working for the Foreign Ministry who released confidential information he possessed related to an arms deal between Canada and Saudi Arabia. First, a control group article was constructed which didn’t include quotes or explicit opinions on the event as the aim was for it to be as neutral and ‘fact-centred’ as possible. The articles constructed for the two experimental conditions used this same article as the core body of the texts, however, both also included text elements to mirror the conceptual dimensions of both frames, i.e. courage and self-interest, and thus evaluated the actions of the civil servant accordingly. In the Courage frame condition the article described the civil servants’ actions as heroic and having been done in the best interest of the public, whereas in the Self-interest frame article the actions were described as selfish with interest in personal gain. Additionally, a quote included in the Courage frame article expressed support for the whistleblowers’ actions, whereas a quote in the Self-interest frame article expressed criticism for his behaviour. The layout
between the articles was kept identical and the sentence-structure as similar as possible with emphasis being on different choices in wording and the tone adopted.
16 Manipulation checks
In order to check whether the manipulation had worked, two manipulation checks, both measured on a seven-point semantic differential scale, were created to assess participants’ perceptions on the focus of the articles. The primary question, which was expected to reflect the manipulation more closely, asked whether participants believed that the article focused on the courage or self-interest of the actors involved. The second question focused on valence by asking whether participants perceived the tone of the article as positive or negative. The expectation was that participants exposed to the Courage frame would perceive the article as focusing on courage the most and as being the most positive, and those exposed to the Self-interest frame the least so. The control group’s perceptions were expected to lie in between the two experimental frames.
Measures
In order to measure participants’ gender, age, education and nationality, single item questions were used. People’s self-perceived political orientation was measured using an eleven-point scale from left (0) to right (10). For descriptives see above.
Moderators
Morality
Self-perceived morality was measured by using a scale adapted from Graham, Haidt & Nosek’s (2009) Moral Foundation scale. Their four original items, such as “In the fight against terrorism, some people’s rights will have to be violated”, were used along with two additional statements1. Morality was therefore measured with six items on a six-point Likert scale
1 The items “It is our responsibility to help those who flee their unsafe homelands” and “Torture is acceptable if it
17
(1=strongly disagree - 6=strongly agree). In order to improve reliability, one item2 was excluded
from further analysis and as the remaining five showed relatively good reliability, they were computed into a morality scale (α=.60, M=4.58, SD=.80).
Political knowledge
Political knowledge is best measured by using factual questions (Delli Carpini & Keeter, 1993) which is why five multiple choice questions were created. Five answer options were provided for each one including one correct answer and a ‘don’t know’ option. Responses were first coded manually (correct answer=1 - all other options=0) after which correct answers were summed up on all five questions (0=no correct answers - 4=all answers correct) to create a political knowledge scale for analysis, despite poor reliability (α=.353, M=2.77, SD=.89).
Mediator
Emotions
Discrete emotions were measured by asking participants to indicate on a seven-point Likert scale the extent to which they felt (1=not at all – 7=a lot) eight different emotions. Based on prior research (Lecheler, Schuck et al., 2013; Nabi, 1999), eight emotions were selected, four positive ones (hope, happiness, empathy, pride) and four negative ones (anger, disgust, sadness, fear). A principal component analysis with Oblique rotation (Direct Oblimin) showed that the items measure two latent constructs (Eigenvalue >1). For components, see Appendix C. Internal
2 The item “When the government makes laws, the number one principle should be ensuring everyone is treated
fairly” was removed as reliability increased from α=.59 to a better, yet not ideal, α= .60.
3 Reliability was not improved by deleting any items. Because reliability was poor, each individual item and a
subscale were used as proxy and all analyses were rerun with all of them. As findings didn’t differ from those results discovered when using the scale, it was kept. Still, caution is required when interpreting the findings.
18
consistency was good so two scales were built, one for positive emotions (α =.84, M=3.10, SD=1.35) and one for negative emotions (α=.86, M=3.33, SD=1.45).
Dependent variables
Opinion on whistleblowing
Participants’ opinion on whistleblowing was measured with four items on a seven-point Likert scale (0=strongly disagree – 7=strongly agree) which related to the whistleblowing actions portrayed directly in the stimuli such as “Mr. Petersen acted in the interest of the public”. The four items showed good reliability (α =.86) and were then averaged to form an opinion towards whistleblowing scale (M=4.43, SD=1.42).
Attitude on whistleblowing
Participant’s attitude on whistleblowing was also measured on a seven-point Likert scale (0=strongly disagree – 7=strongly agree) but with five statements related to more general
attitudes towards whistleblowing, such as “In some cases it is necessary to break the law in order to act in the best interest of the public”. The five items showed good reliability (α=.71) and were averaged to form an attitude towards whistleblowing scale (M=5.03, SD=1.05).
Behavioural intentions regarding whistleblowing
Boyle et al.’s (2016) willingness to take expressive action scale was used to determine behavioural intentions regarding whistleblowing. It included six items measured on a seven-point Likert scale (0=extremely unlikely – 7=extremely likely) which asked likeliness to engage in various behaviours relating to the article such as “talk to friends or family” about it. A reliability
19
analysis showed good internal consistency (α=.89) and the six items were averaged to form a behavioural intention scale (M=3.06, SD=1.31).
Actual behaviour regarding whistleblowing
Participants were also asked if they wished to get additional information related to the topic of whistleblowing in order to see if they not only intended to behave, but actually engaged in real behaviour. The intention of the question was not revealed until debriefing, so committing to the action in the moment went beyond intentions and was considered to be real behaviour. All in all, 61.7% of participants did not want additional information while 38.3% did opt for
receiving additional information.
Results
The direct effects of the Courage frame, Self-interest frame and control group were measured with one-way analyses of variance, independent samples t-tests and logistic regression analyses. Additionally, models 1 and 4 of Andrew Hayes’ PROCESS (2013) were used to test all indirect effect hypotheses using 5,000 bootstrap samples and 95% bias-corrected accelerated confidence intervals. In order to test different analyses, the three conditions were dummy-coded as follows: Courage frame (1) vs. Self-interest frame (0), Courage frame (1) vs. control group (0) and Self-interest frame (1) vs. control group (0).
Manipulation check
When checking if the primary manipulation was successful, a one-way ANOVA showed that participants perceived all three versions of the article differently when asked whether the
20
actions in the article focused on courage or self-interest, (F(2, 159)=53.25, p<.001). Independent t-tests confirmed that people exposed to the Courage frame perceived the article as focusing on courage the most (M=5.58, SD=1.18), the control group was placed in the middle (M=3.58, SD=1.35), and people exposed to the Self-interest frame perceived it as focusing on courage the
least (M=3.04, SD=1.47), as expected.
Another one-way ANOVA looking at perceptions on valence was also significant, F(2, 159)=58.14, p<.001). Independent T-tests confirmed that people exposed to the Courage frame (M=4.94, SD=1.06) perceived the article as more positive than both the control group (M=3.25, SD=1.01) and people exposed to the Self-interest frame (M=3.04, SD=1.20). However, the
difference between the Self-interest frame and control group was not significant indicating that manipulation between those two had not worked in terms of valence.
Overall, however, manipulation was successful and worked in the expected direction. For t-values and post-hoc results of manipulation check items see Appendix D.
Main effects
Opinion & Attitude
A one-way ANOVA discovered a significant, moderate direct effect of journalistic news frames on opinion, F(2,159)=17.15, p<.001, η²=.18. An LSD post-hoc test4 and independent sample t-tests revealed that the mean differences between people exposed to the Courage frame (M=5.28, SD=1.10) and the Self-interest frame (M=4.07, SD=1.30) (t(103)= -.52, p<.001) as well as between the Courage frame and the control group (M=3.97, SD=1.44) (t(108)= -5.34, p<.001)
4 LSD post-hoc test showed significant differences between Courage frame & Self-interest frame (Mdiff=1.22,
p<.001) and between Courage frame & control group (Mdiff=1.31, p<.001), but not between Self-interest frame &
21
were significant. Another one-way ANOVA found a significant, small direct effect of
journalistic news frames on attitude, F(2,159)=3.18, p=.044, η²=.04. An LSD post-hoc test5 and independent sample t-tests revealed that the mean differences between people exposed to the Courage frame (M=5.32, SD=.90) and the Selfinterest frame (M=4.86, SD=1.14) (t(103)=
-.2.29, p=.024) as well as between the Courage frame and the control group (M=4.91, SD=1.05) (t(108)=2.21, p=.029) were significant. Since people exposed to the Courage frame had more positive opinions and attitudes than those exposed to the Self-interest frame, H1a was supported.
While exposure to the Self-interest frame led to more negative opinions and attitudes than exposure to the Courage frame, when looking at the mean difference between the Self-interest frame and the control group, it was not significant for opinion (t(107)=.37, p=.712) or for
attitude (t(107)=.23, p=.822). Therefore, exposure to the Self-interest frame resulted in more
negative opinions or attitudes than the Courage frame but not more negative ones in general, thus only partially supporting H1b.
Behavioural Intentions & Actual Behaviour
When testing for H1c, a one-way ANOVA showed no significant difference between overall means of the three groups, F(2,159)= 2.20, p=.114). However, the LSD post-hoc test6 and an independent sample t-test showed that people exposed to the Courage frame (M=3.36,
SD=1.31) expressed higher intentions to behave than those exposed to the Self-interest frame
(M=2.84, SD=1.32) (t(103)=-2.03, p=.045). As no significant differences7 were found between
5 LSD post-hoc test showed significant differences between Courage frame & Self-interest frame (Mdiff=.46,
p=.024) and between Courage frame & control group (Mdiff=.41, p=.038), but not between Self-interest frame &
control group (Mdiff=-.05, p=.812).
6 LSD post-hoc test shows a significant difference between Courage frame & Self-interest frame (Mdiff=.52,
p=.043), but not between Courage frame & control group (Mdiff=.36, p=.148) or Self-interest frame and control
group (Mdiff=-.16, p=.530).
22
the Courage frame and control group (M=3.00, SD=1.29) or the Self-interest frame and control group, there’s only partial support that these behavioural intentions were caused by the
manipulation. Also, logistic regression analyses showed no significant differences between exposure to the Courage frame vs. Self-interest frame (B=.20, SE=.40, OR=1.23, p=.607, Negelkerke R²=.003), Courage frame vs. control group (B=.43, SE=.40, OR=1.53, p=.279, Negelkerke R²=.014) or Self-interest frame vs. control group (B=.22, SE=.40, OR=1.25, p=.577, Negelkerke R²=.004) on actual behaviour. Overall, journalistic news frames were not a
significant predictor of people’s actual behaviour, but as those exposed to the Courage frame did indeed show higher behavioural intentions than people exposed to the Self-interest frame, H1c was partially supported.
Moderation
Morality
A moderation analysis did not reveal significant effects of exposure to the Courage frame vs. Self-interest frame on opinion via one’s self-perceived morality (b=.12, SE=.27, p=.648, 95% BCa CI [-.4117, .6589]). Differences between Courage frame vs. control group (b=-.21, SE=.30, p=.482, BCa CI [-.8014, .3808]) and Self-interest frame vs. control group (b=-.33, SE=.33,
p=.315, BCa CI [-.9889, .3211]) were not significant either. Similarly, when testing the
moderating effect of morality on attitude, no significant results were found between the Courage frame vs. Self-interest frame (b=-.11, SE=.21, p=.619, BCa CI [-.5312, .3177]), Courage frame
vs. control group (b=.05, SE=.22, p=.812, BCa CI [-.3899, .4964])8 or Self-interest frame vs. control group (b=.16, SE=.25, p=.530, BCa CI [-.3434, .6635]). As morality does not moderate
8 Despite no significant linear trend, the moderation is significant for the respondents with average morality
23
the relationship between journalistic news frames and people’s opinions or attitudes, neither H2a nor H2b were supported.
When looking at whether morality moderates the relationship of journalistic news frames and people’s intention to act with regards to whistleblowing, no significant results were found between Courage frame vs. Self-interest frame (b=-.33, SE=.31, p=.287, 95% BCa CI [-.9356, .2797])9, Courage frame vs. control group (b=-.41, SE=.32, p=.202, BCa CI [-1.0420, .2227]) or Self-interest frame vs. control group (b=-.08, SE=.32, p=.800, BCa CI [-.7206, .5572]).
Similarly, with regards to actual behaviour, no significant results were found between Courage frame vs. Self-interest frame (b=-1.05, SE=.56, p=.060, BCa CI [-2.1491, .0456]), Courage
frame vs. control group (b=-.48, SE=.54, p=.380, BCa CI [-1.5363, .5862]) or Self-interest frame
vs. control group (b=.58, SE=.62, p=.355, BCa CI [-.6441, 1.7975]). Morality does not moderate the relationship between exposure to news frames and behavioural intentions or actual
behaviour, thus H2c was not supported.
Political knowledge
A moderation analysis did not reveal significant effects of exposure to the Courage frame vs. Self-interest frame on opinion via political knowledge (b=.06, SE=.22, p=.776, 95% BCa CI [-.3656, .4882]). Differences weren’t significant between Courage frame vs. control group (b=.08, SE=.22, p=.713, BCa CI [-.3614, .5268]) or Self-interest frame vs. control group either (b=.02, SE=.24, p=.928, BCa CI [-.4481, .4908]). When looking at the moderating effect of political knowledge on attitude, there were no significant results between the Courage frame vs.
9 Despite no significant linear trend, the moderation is significant for the respondents with lower morality (b=.77,
24
Self-interest frame (b=-.01, SE=.18, p=.955, BCa CI [-.3646, .3442])10, Courage frame vs.
control group (b=-.15, SE=.16, p=.340, BCa CI [-.4713, .1643])11 or Self-interest frame vs. control group (b=-.14, SE=.18, p=.425, BCa CI [-.4984, .2118]). Political knowledge does not moderate the relationship of journalistic news frames and people’s opinions or attitudes, thus neither H3a nor H3b were supported.
When looking at the moderating effect of political knowledge on behavioural intentions, no significant results were found between the Courage frame vs. Self-interest frame (b=-.00,
SE=.23, p=.990, 95% BCa CI [-.4604, .4548])12, Courage frame vs. control group (b=.05,
SE=.22, p=.828, BCa CI [-.3933, .4904]) or Self-interest frame vs. control group (b=.05, SE=.22,
p=.817, BCa CI [-.3870, .4897]). With regards to actual behaviour, no significant results were
discovered between the Courage frame vs. Self-interest frame (b=.07, SE=.36, p=.853, BCa CI [-.6429, .7771]), Courage frame vs. control group (b=-.34, SE=.37, p=.366, BCa CI [-1.0674, .3938]) or Self-interest frame vs. control group (b=-.40, SE=.37, p=.276, BCa CI [-1.1305, .3228]). As political knowledge does not moderate the relationship between journalistic news frames and behavioural intentions or actual behaviour, H3c was not supported13.
Mediation
Emotions
10 Despite no linear trend, the moderator is significant for the group with average political knowledge (b=.45,
SE=.20, p=.023, BCa CI [.0631, .8464]).
11 Despite no linear trend, the moderator is significant for the groups with lower political knowledge (b=.53,
SE=.25, p=.040, BCa CI [.0254, 1.0343]) and average political knowledge (b=.36, SE=.18, p=.048, BCa CI [.0041,
.7115]).
12 Despite no linear trend, the moderator is significant for the group with average political knowledge (b=.51,
SE=.25, p=.046, PCa CI [.0088, 1.0201]).
13 All of these analyses were rerun with each of the five knowledge items individually as well as a subscale created
25
A mediation analysis revealed a significant indirect effect of exposure to the Courage frame vs. Self-interest frame on opinions via positive emotions (PE=.632, BootSE=.149, BCa CI
[.3868, .9810]). Exposure to the Courage frame increased positive emotions (b=1.270, SE=.240, p<.001) which were positively related to opinions on whistleblowing (b=.498, SE=.082, p<.001).
Also, when looking at the Courage frame vs. control group, positive emotions significantly mediated this relationship (PE=.711, BootSE=.172, BCa CI [.4230, 1.1092]). Exposure to the Courage frame increased positive emotions (b=1.359, SE=.243, p<.001) which were positively
related to opinions on whistleblowing (b=.523, SE=.247, p=.044). When looking at the mediating effect of news frames on attitudes via positive emotions, Courage frame vs. Self-interest frame showed significant results (PE=.272, BootSE=.095, BCa CI [.1143, .4920]). As before, exposure to the Courage frame increased positive emotions, which in turn led to more positive attitudes towards whistleblowing (b=.214, SE=.077, p=.006). However, here the mediation for Courage
frame vs. control group was not significant (PE=.154, BootSE=.1088, BCa CI [-.0528, .3784])14.
Overall, exposure to the Courage frame increased positive emotions which were positively related to opinions and attitudes, however, since the effect of Courage frame vs. control group on attitude was not significant, H4a was partially supported.
No significant indirect effect for Courage frame vs. Self-interest frame on opinions via negative emotions was found (PE=-.067, BootSE=.065, BCa CI [-.2285, -.0389])15. The
relationship between Self-interest frame vs. control group via negative emotions, however, was significant (PE=.277, BootSE=.105, BCa CI [.1034, .5269]) although in an unexpected direction.
14 Again, exposure to the Courage frame increased positive emotions, but positive emotions did not have an effect
on attitude (b=.113, SE=.073, p=.123).
15 Exposure to the Self-interest frame did not have an effect on negative emotions (b=.298, SE=.276 , p=.283), but
26
In fact, exposure to the Self-interest frame decreased negative emotions (b=-.834, SE=.278, p=.003) which led to less negative opinions (b=-.333, SE=.075, p<.001). Similarly, no significant
indirect effect was found for Courage frame vs. the Self-interest on attitudes via negative emotions (PE=-.070, BootSE=.072, BCa CI [-.2527, .0410])16, yet the relationship between the Self-interest frame vs. control group was significantly mediated by negative emotions (PE=.193,
BootSE=.088, BCa CI [.0599, .4234]). Exposure to the Self-interest frame again unexpectedly decreased negative emotions which led to less negative attitudes (b=-.231, SE=.068, p=.001). Since exposure to the Self-interest frame didn’t increase negative emotions that led to more negative opinions or attitudes, H4b was not supported.
A significant indirect effect for Courage frame vs. Self-interest frame on behavioural intentions via positive emotions was found (PE=.596, BootSE=.160, BBa CI [.3193, .9642]).
Again, exposure to the Courage frame increased positive emotions which were positively related to behavioural intentions (b=.472, SE=.0944, p<.001). This mediation was also significant for Courage frame vs. control group (PE=.565, BootSE=.163, BCa CI [.2896, .9235]. Exposure to
the Courage frame again increased positive emotions, which were positively related to
behavioural intentions (b=.416, SE=.090, p<.001). Additionally, a significant indirect effect of
Courage frame vs. Self-interest frame on actual behaviour via positive emotions was discovered
(PE=.632, BootSE=.151, BCa CI [.3818, .9835]). Exposure to the Courage frame increased positive emotions which were positively related to engaging in actual behaviour (b=.489, SE=.082, p<.001). The mediation was also significant for Courage frame vs. control group
(PE=.716, BootSE=.297, BCa CI [.22671, 1.4095]. Exposure to the Courage frame increased
16 Again, exposure to the Self-interest frame did not have an effect on negative but negative emotions did have a
27
positive emotions which were positively related to engaging in actual behaviour (b=.527,
SE=.177, p=.003). Since exposure to the Courage frame increased positive emotions which were
positively related to both behavioural intentions and actual behaviour, H4c was supported.17
Discussion
Existing theory indicates that people are influenced by news frames as different portrayals of the same issue alter thinking in opposing directions (Slothuus, 2008). This thesis makes its contribution by looking into a previously unknown area of framing effects, that of whistleblowing, and provides empirical evidence that journalistic news frames do indeed have great impact on the way people think and behave. The inherently positive Courage frame and the more negative Self-interest frame showed mostly opposing effects since exposure to the former led to more positive opinions and attitudes towards whistleblowing, as well higher behavioural intentions, than exposure to the latter and vice versa. Thus, this study is the first to document that news coverage about whistleblowing does have the potential to shape public understanding and opinion towards the issue. Both direct and indirect effects were strongest for opinion on
whistleblowing which was to be expected as opinions refer directly to the issue at hand, whereas attitudes, for instance, are harder to change due to being more determinedly placed dispositions (Domfeh & Bawole, 2011). No direct effects of news frames on actual behaviour were
discovered, although this could also be partly due to more technical reasons in the analysis, such
17 It is also worth noting that mediation analyses were conducted with different emotions individually. Being
exposed to the Courage frame increased empathy, hope and pride. The first two were positively related to opinion, while pride was positively related to behavioural intentions and actual behaviour. See Appendix E for details.
28
as the variance being reduced due to the dichotomous nature of the measurement (Martin, Martin, Smith & Hewstone, 2007).
Another contribution of this thesis was answering, to some extent, the question as to why or how people exposed to the news frames were affected by them. The effect of the Courage frame on all four dependent variables was mediated by positive emotions. Here the current study
makes an important contribution since identifying positive emotions as an important mediator of framing effects is in line with recent research in different topical contexts (Berger & Milkman, 2012; Lecheler, Bos & Vliegenthart, 2015). While this study didn’t find any significant overall moderating effects when testing for the conditioning role of concepts such as morality or political knowledge, in multiple cases the group with average morality or average political knowledge did, interestingly, show to be significantly affected by the framing, whereas respondents scoring high or low on these concepts were not. While this was not anticipated or given great prominence, and findings suggest that these factors don’t moderate effects in a linear fashion, it indicates that their role might be more complex, and one that future research could further explore. Other framing research has documented that such middle groups can be most susceptible to attitude changes (Nelson, Oxley et al., 1997). People with high sophistication, for example, have been found to already possess strong attitudes which can obliterate media
influences whereas people scoring low on these variables might not understand framed messages (Lecheler & de Vreese, 2013). Therefore, this thesis supports prior findings which indicate that people positioned between extreme groups are capable of understanding a framed message yet not sophisticated enough to refute it (Nelson, Oxley et al., 1997).
This study found the inherently positive Courage frame to be more effective than the negative Self-interest frame. As negative news has often been found to be more powerful (Chen
29
& Berger, 2013), this can seem counterintuitive. However, as political news is mostly negative, and even so-called neutral news outlets tend to focus on negative aspects when doing political reporting (McIntyre & Gibson, 2016; Trussler & Soroka, 2014), these findings are in line with previous studies that discovered that due to positive news being rare and surprising, it can therefore have a stronger effect on emotional responses and attitudes (Berger & Milkman, 2012; Druckman & Nelson, 2003; Lecheler, Bos et al., 2015). It must also be noted that the
ineffectiveness of the Self-interest frame could in part be due to the unsuccessful manipulation between that and the control group in term of valence. Why people did not fully differentiate between the Self-interest frame and the control group could be linked to whistleblowing attitudes in general. As the public have expressed disdain towards whistleblowers in the past (Public Concern at Work, 2011), information might only be transformative when it’s explicitly positive whereas anything lacking positive attributes, such as the neutral frame, can be perceived
similarly to a negative frame.
The thesis has contributed to the worlds of framing and whistleblowing, however, as the topic has not been researched thoroughly before, there is room for improvement. First, in terms of the design, having a control group and using random assignment increased external validity, and using a between-subject design helped keep the manipulation concealed and reduce
participant fatigue (Field & Hole, 2003). However, using a one factorial design doesn’t allow for any comparisons between different types of whistleblowing cases, which should be taken into account when conducting future research. Researchers could compare whistleblowing that is financial in nature to a public safety issue, or look at whether attitudes depend on organizations being public or private, particularly since media coverage tends to focus on public cases even though more instances occur in the private domain (Heumann et al., 2016; Public Concern at
30
Work, 2011; Wahl-Jorgesen & Hunt, 2012). Second, research into news coverage regarding whistleblowing and opinion polls on the topic have shown that effects can differ between regions such as Europe and the United States (Tavani & Grodzinsky, 2014), hence in order to generalize results globally, more cross-country studies are required (Heumann et al., 2016). Third, as this study did not find significant moderating factors, other relevant factors explored in the future could include pre-existing topic-specific knowledge (Cacciatore, Scheufele & Iyengar, 2016) or political knowledge with better reliability than what was discovered in this thesis.
To conclude, the findings of this thesis offer unique insights into the effects of journalistic news frames on public views regarding whistleblowing. While prior research has discovered significant effects of news frames on various political issues (de Vreese &
Boomgaarden, 2003; Feinholdt et al., 2006; Nelson & Oxley, 1999; Schuck & de Vreese, 2006) this study was the first to investigate the effects of news frames specifically related to
whistleblowing. The current study built on recent content-analytic research and showed how two prominent issue-specific news frames, which play an important part in public debate about this topic, do carry the potential to affect public opinion. Implications of journalistic news frames can reach far since by affecting public attitudes through the use of frames, the media have the
potential to encourage or discourage whistleblowers to come forth. All in all, the frames that journalists decide to use, whether intentionally or inadvertently, have a great impact on public attitudes, also when it comes to whistleblowing.
31 References
Baron, R., & Kenny, D. (1986). The Moderator-Mediator Variable Distinction in Social
Psychological Research: Conceptual, Strategic and Statistical Considerations. Personality and Social Psychology, 51(6), 1173-1182.
Berger, J., & Milkman, K. (2012). What Makes Online Content Viral? Journal of Marketing Research, 49(2), 192-205.
Boyle, M.P., Schmierback, M., Armstrong, C.L., Cho, J., McCluskey, M., McLeod, D.M., & Shah, D.V. (2006). Expressive Responses to News Stories about Extremist Groups: A Framing Experiment. Journal of Communication, 6(2), 271-288.
Brown, A.J., Lewis, D., Moberly, R., & Vandekerckhove, W. (2014). International Handbook on Whistleblowing Research. Cheltenham, UK & Northampton, MA, USA: Edward Elgar. Cacciatore, M.A., Scheufele, D.A, & Iyengar, S. (2016). The End of Framing as We Know it …
and the Future of Media Effects. Mass Communication and Society, 19(1), 7-23.
Chadwick, A., & Collister, S. (2014). Boundary-Drawing Power and the Renewal of Professional News Organizations: The Case of The Guardian and the Edward Snowden NSA Leak. International Journal of Communication, 8(22), 2420-2441.
Chen, Z., & Berger, J. (2013). When, Why, and How Controversy Causes Conversation. Journal of Consumer Research, 40(3), 580-593.
Chong, D., & Druckman, J.N. (2007). A Theory of Framing and Opinion Formation in Competitive Elite Environments. Journal of Communication, 57(1), 99-118.
de Vreese, C. H., & Boomgaarden, H. (2003). Valenced News Frames and Public Support for the EU: Linking Content Analysis and Experimental Data. Communications: The European Journal of Communication, 3, 261–281.
32
Delli Carpini, M.X., & Keeter, S. (1993). Measuring Political Knowledge: Putting First Things First. American Journal of Political Science, 37(4), 1179-1206.
Di Salvo, P., & Negro, G. (2015) Framing Edward Snowden: A Comparative Analysis of Four Newspapers in China, United Kingdom and United States. Journalism, 17(7), 805-822. Domfeh, K.A., & Bawole, J.N. (2011). Muting the Whistleblower Through Retaliation in
Selected African Countries. Journal of Public Affairs, 11(4), 334-343. Druckman, J.N., & Nelson, K.R. (2003). Framing and Deliberation: How Citizens’
Conversations Limit Elite Influence. American Journal of Political Science, 47(4), 729-745.
Entman, R.R. (1993). Framing: Towards Clarification of a Fractured Paradigm. Journal of Communication, 43(4), 51-58.
Feinholdt, A., Schuck, A.R.T., Lecheler, S.K., & de Vreese, C.H. (2016). Shifting Frames: Conditional Indirect Effects of Contested Issues on Perceived Effectiveness Through Multiple Emotions. Journal of Media Psychology, 29(2), 1-11.
Field, S., & Hole, G. (2003). How to Design and Report Experiments. London: Sage Publications.
Giammarco, E.A. (2016). The Measurement of Individual Differences in Morality. Personality and Individual Differences, 88, 26-34.
Graham, J. Haidt, J., & Nosek, B.A. (2009). Liberals and Conservatives Rely on Different Sets of Moral Foundations. Journal of Personality and Social Psychology, 96(5), 1029-1046. Handley, R.L., & Rutigliano, L. (2012). Journalistic Field Wars: Defending and Attacking the
National Narrative in a Diversifying Journalistic Field. Media, Culture & Society, 24(6), 744-760.
33
Harwood, W.H. (2016). Secrecy, Transparency and Government Whistleblowing. Philosophy & Social Criticism, 43(2), 164-186.
Heumann, M., Friedes, A., Redlawsk, D., Cassak, L., & Kesari, A. (2016). Public Perceptions of Whistleblowing. Public Integrity, 18(1), 6-24.
Heumann, M., Friedes, A., Cassak, L., Wright, W., & Joshi, E. (2013–14). The World of Whistleblowing: From the Altruist to the Avenger. Public Integrity, 16(1), 25–51. Hindman, E. B., & Thomas, R. J. (2014). When Old and New Media Collide: The Case of
WikiLeaks. New Media & Society, 16(4), 541-558.
Landert, D., & Miscione, G. (2017). Narrating the Stories of Leaked Data: The Changing Role of Journalists after WikiLeaks and Snowden. Discourse, Context & Media, In Press.
Lecheler, S., Bos, L., & Vliegenthart, R. (2015). The Mediating Role of Emotions: News Framing Effects on Opinions about Immigration. Journalism & Mass Communication Quarterly, 92(4), 812-838.
Lecheler, S., & de Vreese, C. H. (2012). News Framing and Public Opinion: a Mediation Analysis of Framing Effects on Political Attitudes. Journalism & Mass Communication Quarterly, 89(2), 185–204.
Lecheler, S. K., Schuck, A. R. T., & de Vreese, C. H. (2013). Dealing with Feelings: Positive and Negative Discrete Emotions as Mediators of News Framing Effects.
Communications, 38(2), 189–209.
Lee, N.J., McLeod, D. M., & Shah, D. V. (2008). Framing Policy Debates Issue Dualism, Journalistic Frames, and Opinions on Controversial Policy Issues. Communication Research, 35(5), 695–718.
34
Luther, C. A., & Radovic, I. (2014). Newspapers Frame Julian Assange Differently. Newspaper Research Journal, 35(1), 64-81.
Lynch, L. (2013). WikiLeaks after Megaleaks: The Organization’s Impact on Journalism and Journalism studies. Digital Journalism, 1(3), 314-334.
Martin, R., Martin, P.Y., Smith, J.R., & Hewstone, M. (2007). Majority versus Minority Influences and Prediction of Behavioral Intentions and Behavior. ScienceDirect, 43(5), 763-771.
McIntyre, K.W., & Gibson, R. (2016). Positive News Makes Readers Feel Good: A “Silver-Lining Approach to Negative News Can Attract Audiences. Southern Communication Journal, 81(5), 304-315.
Nabi, R.L. (1999). A Cognitive-Functional Model for the Effects of Discrete Negative Emotions on Information Processing, Attitude Change, and Recall. Communication Theory, 9(3), 292-320.
Nelson, T.E., & Oxley, Z.M. (1999). Issue Framing Effects on Belief Importance and Opinion. The Journal of Politics, 61(4), 1040-1067.
Nelson, T.T, Oxley, Z.M., & Clawson, R.A. (1997). Toward a Psychology of Framing Effects. Political Behavior, 19(3), 221-246.
Ottosen, R. (2012). WikiLeaks: Ethical Minefield or a Democratic Revolution in Journalism? A Case Study of the Impact of Afghanistan Coverage in the Norwegian Daily, Aftenposten. Journalism Studies, 13(5-6), 836-846.
Pan, Z., & Kosicki, G.M. (1993). Framing Analysis: An Approach to News Discourse. Political Communication, 10(1), 55-75.
35
Park, H., Blenkisopp, J., & Park, M. (2014). The Influence of an Observer’s Value Orientation and Personality Type on Attributes Toward Whistleblowing. Journal of Business Ethics, 120(1), 121-129.
Pew Research Center (2010). Most Say WikiLeaks Release Harms Public Interest. Retrieved from:
http://www.people-press.org/2010/12/08/most-say-wikileaks-release-harms-public-interest/
Pew Research Center (2013). Snowden Leak Seen as Harming Public Interest. Retrieved from:
http://www.pewresearch.org/fact-tank/2013/12/17/snowden-leaks-seen-as-harming-the-public-interest/
Public Concern at Work (2016). Whistleblowing: Time for Change: A 5 Year Reviews by Public Concern at Work.
Public Concern at Work (2011). Whistleblowing: Beyond the Law. The Biennial Review. Qin, J. (2015). Hero on Twitter, Traitor on News How Social Media and Legacy News Frame
Snowden. The International Journal of Press/Politics, 20(2), 166-184.
Scheufele, D.A. (2000). Agenda-Setting, Priming, and Framing Revisited: Another Look at Cognitive Effects of Political Communication. Mass Communication & Society, 3(2-3), 297-316.
Schuck, A.R.T., & de Vreese, C.H. (2006). Between Risk and Opportunity: News Framing and its Effects on Public Support for EU Enlargement. European Journal of Communication, 21(1), 5-32.
Slothuus, R. (2008). More than Weighting Cognitive Importance: A Dual-Process Model of Issue Framing Effects. Political Psychology, 29(1), 1-28.
36
Tavani, H.T., & Grodzinsky, F.S. (2014). Trust, Betrayal and Whistle-Blowing: Reflections on the Edward Snowden Case. ACM SIGCAS Computers and Society – Special Issue on Whistleblowing, 44(3), 8-13.
Trussler, M., & Soroka, S. (2014). Consumer Demand for Cynical and Negative News Frames. The International Press/Politics, 19(3), 360-379.
Valkenburg, P.M, Semetko, H.A., & de Vreese, C.H. (1999). The Effects of News Frames on Readers’ Thought and Recall. Communication Research, 26(5), 550-569.
Vogele, C., & Baudermann, L. (2016). Whistleblowing zwischen Zivilcourage und Denunziantentu: Eine Frame-Analyse zur Berichterstattung uber Whistleblowing in deutschen und deutschschweizerischen Printmedien. Medien & Kommunikationswissenschaft, 64(6), 518-541.
Wahl-Jorgensen, K., & Hunt, J. (2012). Journalism, Accountability and the Possibilities for Structural Critique: A Case Study of Coverage of Whistleblowing. Journalism, 13(4), 399-416.
YouGov (2016). British Public Reject UN Ruling on Julian Assange. Retrieved from:
37 Appendix A: Questionnaire on Qualtrics
TEXT1
Dear participant,
I would like to invite you to participate in a research study to be conducted under the auspices of the Graduate School of Communication, a part of the University of Amsterdam. This study looks into people’s opinions about current topics in the news media. In this online survey, a news article will be displayed followed by a series of questions. As this research is being carried out under the responsibility of the ASCoR, University of Amsterdam, we can guarantee that:
1) Your anonymity will be safeguarded, and that your personal information will not be passed on to third parties under any conditions, unless you first give your express permission for this.
2) You can refuse to participate in the research or cut short your participation without having to give a reason for doing so. You also have up to 24 hours after participating to withdraw your permission to allow your answers or data to be used in the research.
3) Participating in the research will not entail your being subjected to any appreciable risk or discomfort, the researchers will not deliberately mislead you, and you will not be exposed to any explicitly offensive material.
4) No later than five months after the conclusion of the research, we will be able to provide you with a research report that explains the general results of the research.
For more information about the research and the invitation to participate, you are welcome to contact the project leader Vilma Nurmela (vilmanurmela@gmail.com) at any time. Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this research, you can contact the designated member of the Ethics Committee representing the ASCoR, at the following address: ASCoR secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐ 525 3680; ascor‐secr‐fmg@uva.nl. Any complaints or comments will be treated in the strictest confidence.
We hope that we have provided you with sufficient information. We would like to take this opportunity to thank you in advance for your assistance with this research, which we greatly appreciate.
Kind regards, Vilma Nurmela
TEXT2
I hereby declare that I have been informed in a clear manner about the nature and method of the research, as described in the invitation previously.
I agree, fully and voluntarily, to participate in this research study. With this, I retain the right to withdraw my consent, without having to give a reason for doing so. I am aware that I may halt my participation in the experiment at any time. If my research results are used in scientific publications or are made public in another way, this will be done such a way that my anonymity is completely safeguarded. My personal data will not be passed on to third parties without my express permission.
If I wish to receive more information about the research, either now or in the future, I can contact the project leader Vilma Nurmela at any time. Should I have any complaints about this research, I can contact the designated member of the Ethics Committee representing the ASCoR, at the following address: ASCoR secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐ 525 3680; ascor‐secr‐fmg@uva.nl.
