Transparency in Journalism, Trust in the Media:
Evaluating the impact of user-generated content on journalistic practice and
audience perceptions
Name: Ciarán O’Connor
Student Number: 11368268
Master’s Thesis; Graduate School of Communication
Master’s Programme; Communication Science
Supervisor: Michael Hameleers
Word Count: 8,500
opening up journalism, it is proposed that news organisations can strengthen relationships with their readers, viewers, or followers. One way journalism has opened up in recent years has been the increasing reliance on amateur footage, known as user-generated content (UGC), shot by civilians at the scene of an incident. Their footage is regularly discovered on social media and used by news organisations. The contribution of non-professionals in the journalistic process is recognised as a indicator that journalism is opening up and becoming more transparent. Research to date has explored the relationship between transparency and trust, but little is understood about the effects of how UGC, as a marker of transparency, affects audience trust. This research aims to fill that gap.
Using a mixed methods approach, this thesis built on previous research and examined how often UGC is used and credited by news organisations. It then tested the relationship between transparency and trust by investigating whether various forms of transparent uses of UGC in news reports impact trust. Results indicated that while news organisations regularly use UGC, including source information on UGC isn’t an industry-recognised norm. Further analysis showed individuals prefer known, established sources of information over reports of
eyewitnesses on social media as a form of transparency conducive to forming trust in news reports with UGC.
The overall conclusion indicates transparency in journalism is still an ideal, the link between transparency and trust has yet to be definitively proven and requires future research.
study since I first set out to come to Amsterdam. Many thanks are due to my lecturers and teachers who have helped me get to this point, but particularly to my supervisor, Michael Hameleers, whose encouraging support and guidance helped me shape this thesis and bring my ideas to reality.
I am particularly grateful to Louise Tierney, a colleague and friend whose constant curiosity and interest in my research was invaluable at times, and later, her involvement as a coder in the research itself was incredibly helpful. I am also thankful to Claire Wardle for
assisting my research by providing me with a sample codebook. My colleagues in Storyful News agency also deserve a thank you for supporting me in my decision to study in UvA and allowing me the time and space to undertake a Master’s while also remaining close to the newsroom. Without my work in Storyful, I wouldn’t have acquired the knowledge or interest in this area to pursue this Master’s degree.
A big thank you also to Aishling Lennon, Gearoid Lennon, and Val O’Connor for proofreading this document when it was near the finish line and required an outsider’s perspective to see what I could no longer see and suggest final checks and edits.
Lastly, and most importantly, to my family. Margaret, Val, Dec, and Dave have all helped me arrive at this point in their own ways. Their unconditional support is a source of great strength for me, and I will always be grateful for their help.
Theoretical Framework ……….. 2
What is UGC ……….. 2
UGC in the News ……….……….. 3
Transparency ……….. 4
The ‘Opening up of Journalism’ ……… 5
UGC and Transparency ……….. 7
Trust in the Media ……….. 8
Trust vs. Credibility ………... 9
Trust Over Time ………... 10
UGC and Trust ……….… 11
Using UGC to Examine Transparency and Trust ……….…………...….... 11
The Aim of this Research ……… 12
Hypotheses ...………....….…….…. 13
Methodology ……….……….. 14
Study 1: Content Analysis ……….. 14
Sample ……….…………... 14
Coding Protocol ………..… 15
Intercoder Reliability ……….. 15
Analysis Strategy ……… 15
Sample ……….………….... 18 Independent Variable ………... 19 Dependent Variable ………. 19 Treatment ……….………… 20 Procedure ……….………… 21 Manipulation Check ………. 22 Measuring Trust ……….….. 22 Results ……….………. 23 Discussion ……….………... 25 Conclusion ……….……….. 30 Bibliography ……….………... 33 Appendices ……….……….. 39
News organisations are opening up. A shift towards more transparency can be seen on social media where journalists regularly speak about their work with the public, or in comment sections of news websites where people share their opinion or post questions about the reporting at hand. This change has been driven significantly by advancements in digital communication technology, namely the rise of the internet, that have facilitated greater information access and interaction than previous decades (Fung, Graham, & Weil, 2007; Karlsson, 2011). The public now expect to know more about news organisations, and other civic institutions. News organisations recognise this and advocate transparency as a way of restoring and building audience trust.
The BBC recently published guidelines outlining how it is “working to strengthen trust and transparency in online news” (2017) while a Reuters Institute report concluded that
newsrooms are gradually revealing more ways that “demonstrate the transparency” (Newman, 2017) of their journalism to instill public trust. Does transparency have an impact on journalistic practice, or audience perceptions? The increased use of amateur footage in news reports, sourced from non-professionals and used by journalists in their reporting, suggests it impacts both by opening up journalism (Wardle, Dubberley & Brown, 2014). This impact is little understood, and thesis aims to answer these questions by conducting a content analysis and an experiment. But first, at an academic level, how is the inclusion of amateur footage, known as user-generated content (UGC), in modern news understood in journalism research, and secondly, at a professional level, how do news organisations use it?
Theoretical Framework What is UGC?
In the view of Hellmueller, Vos, & Poepsel (2013) user-generated content refers not only to amateur pictures and videos, but also audience comments in sections underneath articles or in emails to the newsroom, and is understood as a form of participatory journalism (Hellmueller & Li, 2015). This thesis takes a more focused view of the term, looking purely at visual forms of UGC like amateur videos posted on social media, and how they are used by news organisations. UGC will be exclusively explored in this paper, because this falls under a burgeoning area of study where little research to date has examined the interplay between visual UGC, its effect on newsroom practice, and impact on audience perceptions. Using the definition of Wardle et al., (2014), UGC in news coverage is understood as “photographs and videos captured by people who are not professional journalists and who are unrelated to news organisations” (2014, p. 5).
UGC is a common feature in modern journalism, helping journalists tell stories in innovative, engaging ways, and is not a new phenomenon. Abraham Zapruder’s 1963 footage of the shooting of President John F. Kennedy shows UGC was considered newsworthy long before the term existed. In the modern context, the 2004 tsunami in Asia, the 2005 London bombings, and significantly, the 2011 Arab Spring and ongoing Syrian conflict are considered watershed moments for the utilisation of UGC in journalism (Wardle et al., 2014).
UGC is primarily shot by non-professional individuals who happened to record footage of an incident and were able to share it afterwards, nowadays, usually on social media. Such footage like this now regularly generates news coverage. Because of advancements in digital and
social media, people outside news organisations can now interact with, contribute to, and challenge journalism more than in the past, and crucially, can also become sources with UGC. News organisations use social media to source UGC from inaccessible places like Syria because it offers “proximity and immediacy” (Mast & Hanegreefs, 2015, p. 595). This interactivity and speed are key for journalists who now report on emerging stories from Syria or elsewhere in real-time from a newsroom in Dublin, using social media to find sources and UGC on the ground. But how is UGC used?
UGC in the News
The civil war in Syria provides a good case study for analysing how UGC is used, as coverage of the conflict has relied on UGC since its outbreak in 2011. Research has consistently shown that crediting the source of UGC in news coverage is not commonplace. Mast &
Hanegreefs (2015) examined the crediting of 294 “non-professional” UGC items from Syria used by Flemish news organisations between March 2011 and December 2012, and observed that half (50.7%) contained no source credit. In a content analysis of BBC Arabic and Al Jazeera Arabic’s 2011 news coverage of Syria which used UGC, Harkin, Anderson, Morgan & Smith (2012) found almost 75% of the 64 UGC items analysed contained no credit. This thesis will examine recent coverage of Syria to see if similar practice continues.
Wardle et al., (2014) demonstrated that the problem is not limited to coverage of Syria. In their analysis of 2,115 UGC items about general news events used by eight news organisations over three weeks, only 16% of UGC used on TV was credited. Despite UGC’s democratising role in bringing people external to newsrooms into the journalistic process, why is crediting not regarded as necessary?
Throughout interviews with 64 individuals in 38 news organisations worldwide, Wardle et al., highlighted some of the reasons such as “the pressure of breaking news situations,
concerns about screen clutter caused by crediting, and unease about crediting certain
organisations,” (2014, p. 84) and added that confusion and ignorance are common in newsrooms. They conclude that one of the main problems is that there is a lack of systemised practice in how UGC is used and credited (2014). While Reuters (2017), the Associated Press (2013) or the BBC (2006) all have guidelines for using UGC, they all differ and an industry standard is sorely lacking.
The prevalence of UGC and social media today is part of a bigger societal shift which now favours sharing more information with others. This is symptomatic of a greater shift towards increased transparency by ourselves, brands, or political parties, and can be seen at personal and professional levels, in both individuals and institutions.
Transparency
Transparency is held in high esteem across a number of disciplines and sectors of society. Increased openness is seen as one way to make institutions more accountable and more
trustworthy (Kovach & Rosenstiel, 2007) to their relevant publics. While the concept of
transparency is not new, its increased relevance for institutions in recent decades has been aided by advancements in technology (Fung et al., 2007). Digital and online technology means storing and accessing information has never been easier, (Allen, 2008; Karlsson, 2011) allowing people outside institutions greater abilities to scrutinise decisions, pose questions, and critique practices. For professions that rely on close relationships with the public, including politicians and
century is disappearing. This evolution has forced institutions to rethink routines and behaviours that were once regarded as normal.
The maturation of communication technologies like social media allows the public to view once-guarded practices that are now exposed to the public. In October 2017, outgoing Republican Senator Bob Corker used a public tweet to label President Donald Trump’s White House an “adult day care center” (Corker, 2017). Similarly, journalists now interact with people publicly on social media, probing them for eyewitness information and UGC to corroborate reporting around breaking news events, like political demonstrations or terrorist attacks, further altering the traditional one-way mode of journalism.
The ‘Opening Up’ of Journalism
Transparency and journalism have been intertwined since the early professionalisation of the profession, especially, as Koliska suggests, if “sourcing and attribution are conceived as the most basic forms of transparency in journalism” (2015, p. 6). Koliska highlights that, as early as 1920, Walter Lippmann “proposed sourcing as a crucial element to give journalism the clout of a respected profession” (2015, p. 7). In the modern age of journalism, digital technology enables journalists to share more source and attribution (crediting) information than ever before, along with other new tools Karlsson labels ‘rituals of transparency’ (2010). Digital media affords journalists new methods to show their work, yet to what extent such transparency affects audiences will be examined in this thesis. Social media has accelerated this opening up of the journalistic process, and journalists now interact directly with the audience. Karlsson
characterises this new openness as “specific techniques that can be used routinely by journalists and identified and understood by users” (2010, p. 537).
Traditional journalism facilitated minimal interaction between reader and writer. Now, new online features encourage this interaction. Karlsson remains one of the few researchers to explain what this openness looks like on news websites. Karlsson (2010) compared the numerous online tools of transparency utilised on the websites of the New York Times, the Guardian, and Sweden’s Dagens Nyheter, identifying features like timestamps, corrections, links to external documents outside the website, yet did not include UGC. From a sample of 335 online articles, Karlsson found that all sites employed at least one feature of transparency in their articles, concluding that these tools help audiences learn more about the journalistic process and to interact with journalists.
Transparency in journalism has often been equated with the traditional norm of objectivity, and even as a new form of objectivity (Hellmueller et al., 2013; Karlsson, 2010; Weinberger, 2009). Tuchman defined objectivity as a norm central to a style of journalism that prioritised information over opinion, a “strategic ritual” (1972, p. 660) that protected journalists from the risks of the trade, and she identified ways journalists practice objectivity, like ensuring reporting is balanced. For media critic Jay Rosen, objectivity is a myth that contributes to a “view from nowhere” (2010). Journalists today still regard objectivity as a pillar of the profession, but do they perceive a distinction between objectivity and transparency?
Hellmueller et al. label transparency as a “new occupational norm and strategy for truth-telling” (2013, p. 287). They tested US journalists’ attitudes towards objectivity, neutrality, and Karlsson’s two aspects of transparency, disclosure and participatory (2010), which will be explained below. On average, objectivity as a “strategy of factualness” (Hellmueller et al., 2013, p. 295) was the strongest preference among journalists. Disclosure transparency, such as links to
external sources of information, was valued as the second strongest truth-telling strategy, followed by neutrality. Journalists ranked participatory transparency, such as the inclusion of user-generated information, as the least important strategy for truth-telling (2013). Their findings also point to a division between journalists who have always worked online prioritising
transparency over objectivity, versus an opposite view from those who began in traditional journalism.
Agarwal and Barthel (2013) recorded similar results when they spoke to journalists from online and traditional news outlets. Younger, online journalists prioritise transparency, with one journalist saying they “need to be honest about their information-gathering practices” (2013, p. 382), while strongly rejecting objectivity and neutrality. Digital communication is regarded by scholars as significant in shifting attitudes that once favoured objectivity towards favouring transparency (Deuze, 2005; McNair, 2017; Plaisance, 2007). Plaisance even states that digital technologies have “reaffirmed claims that openness is the best way to build public trust” (2007, p. 192).
UGC and Transparency
Deuze defines transparency in journalism as “the increasing ways in which people both inside and external to journalism are given a chance to monitor, check, criticise and even
intervene in the journalistic process’’ (2005, p. 455). ‘Inside’ the process are journalists who now have greater powers to show their work to audiences with online technologies. ‘External’ to the process are non-newsroom individuals, often social media users whose content is now regularly used in reporting, and who are representative of how non-professionals participate in the
transparency, allowing people external to the newsroom to contribute with amateur images and videos. As newsrooms incorporate UGC sourced from non-professionals, they open up the journalistic process to the audience. Yet, how their UGC is incorporated presents new challenges for newsrooms.
Karlsson defines two forms of transparency in journalism today. Disclosure transparency refers to how “news producers can explain and be open about the way news is selected and produced” (2010, p. 537) and participatory transparency refers to how the audience and non-professionals are “invited to participate in different stages in the news production
process” (2010, p. 537). Disclosure transparency is seen in the increasing openness of
journalism, and evident through how UGC is verified, how a source is credited, or whether any stage of this process is communicated to the audience by journalists ‘inside’ the organisation (Deuze, 2005). Participatory transparency relates to how a non-professional actor contributes to news coverage by providing UGC, or eyewitness testimony. Though ‘external’ to the
organisation (Deuze, 2005), their contribution significantly shapes newsroom practice and generates coverage of salient news events. Karlsson’s two strands of transparency will be used to form an independent variable. But, as UGC furthers transparency in opening up the journalistic process, what are its potential effects on audience trust?
Trust in the Media
In today’s world, trust is essential for individuals to navigate interactions with people and institutions that cannot be avoided in our complex society where labour and services are
separated (Simmel, 1950). This division allows experts to rise in all professional fields, and people rely on their knowledge to compensate for their own insufficient knowledge in an area
(Giddens, 1990). Trust is a “lubricant of a social system” (Fukuyama, 1995, p. 151) that helps people narrow the uncertainty between knowing and not knowing (Simmel, 1950). The distance between knowing and not knowing is central to understanding trust’s importance for journalism. The distance between knowledgeable experts, like journalists, and an uninformed audience is referred to as a “competence gap” (Parsons, 1978, p. 46) or a "cognitive leap” (Lewis & Weigert, 1985, p. 971). Parsons states this gap is “bridged by something like what we call trust” (Parsons, 1978, p. 46).
Trust vs. Credibility
Research has only recently started to focus on media trust as a key concept in the exchange between news organisations and audiences. Media credibility traditionally received more attention from scholars (Kohring & Matthes, 2007; Roberts, 2007). Koliska states that in this digital media environment, research needs to examine the “interplay between news
producers and news users” (2015, p. 115), like exchanges on social media. As early as 1997, Singer detected the impending impact of this interplay on journalism when a journalist told her “the interactive capabilities of online media can open the gate between journalist and reader, ’humanising and personalising the reporter’” (1997, p. 85).
In the past, trust has been understood as just one aspect of credibility (Karlsson, Clerwall, & Nord, 2014; Kohring, 2004). Karlsson et al. (2014) tested the effects of transparency on readers’ perceptions of journalistic and news content credibility, but found no effect. Roberts (2007) measured the effects of computer-mediated transparency on journalistic credibility, such as external links to additional information about a topic or about the editorial process, but also found no effects. Kohring & Matthes (2007) questioned which way the direction goes - does
increased trust lead to increased credibility, or vice versa. For them, the difference lies in risk, a component of trust, which is essential to the exchange between knowing and not knowing and is built upon ‘journalistic selectivity’ (2007) to highlight some information, and exclude other information. But how is trust created?
Trust Over Time
Repeated exposure to news media affects trust. Tsfati & Cappella (2003) observed that the level of scepticism, defined as the opposite of trust, among media consumers was negatively related to exposure to mainstream news, demonstrating that repeated encounters with the media did not lower trust. So, it seems trust in mainstream media is created over time, yet over time, they also observed that exposure to non-mainstream media positively related to scepticism, sowing mistrust in mainstream media (2003). Might journalism’s pursuit of audience trust be self-defeating? Arguably, if journalism serves the public interest and exposes wrongs in society, this may erode public trust in institutions that are central to society. Koliska calls this
journalism’s “irrational loop,” (2015, p. 112) and states it’s necessary in complex societies, even though “a lack of trust in journalism could ultimately result in a dysfunctional society that is stricken by the absence of trust” (2015, p. 110). Evidently, the relationship between journalism, trust, and society is multifaceted.
Tsfati demonstrated the effect of time again, this time with online mainstream media, showing that repeated exposure had a negative effect on media scepticism (2010) which he said fostered trust as “the more people trust mainstream media, the more they consumed news from online mainstream news outlets” (2010, p. 37).
UGC and Trust
What of UGC’s potential effect on trust, in the context of other transparency features? The use of UGC in news reporting has increased dramatically in the last 10 years due to the mass, worldwide proliferation of camera phones and the development of global communication technologies like social media that help disseminate such content. Thus, research that analyses the medium to long-term effects of UGC is limited, with no study examining its role in the link between transparency and trust. Including UGC in news reports presents challenges of its own as a journalist must verify its authenticity, explain its inclusion, and justify its use instead of
professional footage. If transparency is understood as providing more information, which can limit uncertainty and achieve what Koliska calls a point of “knowledge saturation,” (2015, p. 96) then by reaching this point, “trust may occur” (2015, p. 96).
UGC has been found to have a negative effect on information credibility, (Bivens, 2008; Singer, 2010) with one reason cited as the difficulty in verifying such content. Grosser (2016) states UGC might be bad for credibility, but trust in online media consists of many parts, two of which she defines as characteristics of trustworthiness that contribute positively towards
credibility - “diversity of information” and “currentness of information” (2016, p. 1040). Grosser adds that credibility could be strengthened by further disclosure transparency in naming and explaining sources and their interests (2016).
Using UGC to Examine Transparency and Trust
Despite being regarded as a key tool to increase audience trust, little evidence exists that transparency has any positive effects on credibility (Karlsson et al., 2014; Roberts, 2007) or trust (Koliska, 2015). Just one experiment (Meier & Reimer, 2011) indicated, using Kohring &
Matthes’ (2007) model to measure journalistic selectivity, that transparency can lead to a small increase in trust. There have been no experiments to test if, when transparency is applied to UGC, it leads to more trust. This thesis will adopt Meier and Reimer’s design, and use disclosure transparency, participatory transparency, and a combination of both (full transparency) to focus on UGC and test the link between transparency and trust. This thesis represents one of the first attempts to measure if varying forms of transparency, related to UGC, impact audience trust in online news reports.
Kohring (2004) places journalists within this scope of selectivity, members of an expert system who the public rely on to reduce risk. The audience trust journalists to select and contextualise information, providing knowledge about local, national, and international
incidents. As a public service providing citizens with information about institutions in their lives, journalism’s impact on trust in society is crucial, and the selectivity they employ in choosing which topics to cover and what information to highlight is key to understanding media trust (Kohring, 2004).
Kohring & Matthes (2007) focused on journalistic selectivity to propose standardised dimensions to measure media trust and this defines how trust is understood and measured as the dependent variable in this thesis.
The Aim of this Research
This thesis will contribute to the understanding of UGC’s role in transparency and its effect on trust in two ways. The first involves building on previous research to measure the extent to which coverage of the Syrian conflict uses UGC, how its sources are credited, and how transparent news organisations are with their use of UGC. This leads to the first research
question:
To what extent is User Generated Content (UGC) used in online news reports?
The second way constitutes the main investigation in this thesis. Here, the causal link between transparency and trust will be tested, through the use of UGC and its understanding as indicators of disclosure and participatory transparency (Karlsson, 2010). Participants will be asked to evaluate their levels of trust after exposure to a news report, depending on various forms of transparency, as a way of testing trust through journalistic selectivity (Kohring & Matthes, 2007). This leads to the second research question:
To what extent can transparency of amateur footage in news reporting lead to more trust among respondents?
Using Kohring’s & Matthes’ (2007) multifactor model of trust, I propose the following hypotheses:
H1: A disclosure transparency article will be more trusted than a non-transparent article H2: A participatory transparency article will be trusted more than a non-transparent article. H3: A full transparency article will be trusted more than a non-transparent article.
H4: A full transparency article will be trusted more than a disclosure transparency article. H5: A full transparency article will be trusted more than a participatory transparency article.
Methodology
A mixed methods approach was used. Firstly, a small-scale exploratory content analysis was conducted on news coverage of the Syrian civil war to compare to previous research about use of UGC in such coverage (Harkin et al., 2012; Mast & Hanegreefs, 2015; Wardle et al., 2014). Secondly, an experiment, which constitutes the main focus of this thesis, was conducted to test if, when news reports include UGC, more transparency leads to more audience trust. The methodology used and results obtained for each method are discussed separately.
Study 1: Content Analysis
A content analysis was conducted on coverage from the BBC, CNN, and Al Jazeera English from December 2016 about the conflict in Aleppo, Syria, which provides a good case study for the use of UGC in news reports because, as one news editor told Wardle et al., due to the danger of reporting in the country “there’s no way for us to cover Syria other than
UGC” (2014, p. 13). December 2016 represented the “climax” of the battle for Aleppo (Hassan, 2016), with a ceasefire called on December 14 when government forces took control of the city. The ceasefire allowed news organisations more access so the pre/post-ceasefire periods could potentially reveal a contrast in reliance on UGC by news organisations.
Sample
The BBC, CNN, and Al Jazeera English were chosen as they represent three international news outlets that have covered the Syrian conflict in its entirety. Reports which mentioned “Aleppo” were sourced from searches on their websites, and 118 news articles were gathered from a random start date of December 2, every third day afterwards, concluding December 29, 2016. Articles were examined for UGC, understood as amateur footage shot by non-professional
actors. One major restriction was applied to certain news articles. For news reports with multiple pieces of UGC, only the first piece was coded, as this was considered the most newsworthy. This limitation is discussed below.
Coding Protocol
A sample codebook was obtained from Wardle et al., from their UGC study (2014). This guided my analysis in constructing variables. Key questions were developed to investigate whether UGC was included in a news report, and if the source of the UGC was credited. The category of UGC was also recorded (full codebook in Appendix A). Both key questions
contained a simple yes/no answer. These were central to understanding the transparent element of coverage.
Intercoder Reliability
A second coder was trained to ensure there was clarity about the variables and coding manual. We examined a random selection of 15 articles (12.7% of full sample). After our first attempt, intercoder reliability wasn’t achieved so another training session was conducted and instructions were clarified. 15 new articles were selected and intercoder reliability was achieved. For the question ‘Was UGC used in this news report?’, Krippendorff’s Alpha = 0.82, percentage agreement = 93.3%. Similarly, for ‘was the UGC credited?’, Krippendorff’s Alpha = 0.82, percentage agreement = 91.6%. Full data collection then commenced.
Analysis Strategy
The aim of the small-scale content analysis is to examine the use and crediting of UGC in news coverage of Syria and build on previous research about UGC use in news coverage of the conflict. For this, simple descriptive analysis was employed in examining the dataset to provide
insights into UGC use among the three news organisations, such as the frequency of use through the observed period. Also, descriptive analyses allowed for an examination of source crediting by these organisations, aiding the overall aim of my research to investigate transparency with
regards to UGC.
Results: How Often is UGC Used and Credited by News Organisations
From the search of “Aleppo” related articles from the BBC, CNN, and Al Jazeera English in December 2016, 118 articles were found. As seen in table 1 below, CNN accounted for 50% of the sample with 59 articles about Aleppo, Al Jazeera English with 26.4%, and the BBC with 24.6% For the key question of ‘Was UGC used in this news report’ (N = 118), 64 articles (54.2%) did include UGC in their coverage. Al Jazeera English used the highest proportion of UGC, 20 of their 31 articles included UGC (64.5%), followed by the BBC with 57.1%, and CNN with 47.5%.
Table 1 - Question: Was UGC used in this news report
For the key question of ‘Was the UGC credited in the news report’ (N = 64), the source of UGC in 50 articles (78.1%) was credited, as seen in table 2 below. Of their 28 articles that used UGC, CNN credited everything (100%), followed by the BBC (75%) and Al Jazeera English (50%).
Was UGC used? Yes No N = 118
BBC 16 12 28 (24.6%)
CNN 28 31 59 (50%)
Al Jazeera English 20 11 31 (26.4%)
Table 2 - Question: Was UGC credited in this news report
The category of UGC was also recorded. UGC depicting conflict in Aleppo was the largest section, accounting for 23 (35.9%) news articles, testimonial videos from residents in Aleppo accounted for 16 (25%) news articles, and UGC about humanitarian efforts after an attack accounted for 10 (15.6%) news articles.
During the five days monitored between December 2 - 14, 46 news articles used UGC, accounting for 71.9% (n = 64). In the five days monitored between December 17 - 29, just 18 articles used UGC, accounting for 28.1%. On December 14 alone, when a ceasefire was declared, the three news organisations had 22 articles with UGC.
In summary, the content analysis has demonstrated that UGC was a central component of news organisations’ coverage of Aleppo and featured in 54.2% of news reports examined. When news organisations had greater access to Aleppo after December 14, they used less UGC, compared to when the conflict was still ongoing. The high percentage for crediting also highlights the importance placed in source citation too. Regarding the research question, it’s clear UGC is a salient feature of online news reports, and crediting of UGC sources, recognised as a form of transparency, is much practiced by these news organisations.
Was UGC credited? Yes No N = 64
BBC 12 4 16 (25%)
CNN 28 0 28 (43.75%)
Al Jazeera English 10 10 20 (31.25%
Study 2: Experiment
The first study demonstrated that UGC is used and credited frequently in news coverage. The next question regards how crediting of UGC, and other methods of transparently using UGC, affect perceptions of trust among the audience. To test this, an experiment was developed.
Design
An online experiment examined the link between transparent uses of UGC in news reports and audience trust in journalistic selectivity by using an empirically-tested
conceptualisation of trust. To test this relationship, hypotheses were formulated and a 4-factor between-subjects experimental design was developed. Respondents were randomly assigned to one of four news articles, each which represented a form of transparency: non-transparent
(control), disclosure transparency, participatory transparency, and full transparency (combination of disclosure and participatory). Trust was measured in a posttest survey.
Sample
Respondents were sourced through social media. I shared the survey link and stated the experiment was to test attitudes about online journalism, not disclosing the true aim (see
Appendix C). The survey was shared across Facebook and Twitter and steps were taken to ensure this convenient sample achieved variation regarding age, nationality, and background. The limitations of this sample are discussed below. No incentives were offered for participation. At least 50 respondents were desired for each treatment. 441 respondents opened the survey link, with a completion rate of 53.5%, meaning 236 respondents completed the survey and remained in my sample for analysis.
(M = 2.45, SD = 1.24) clustered around the 25-34 grouping which included 112 respondents (51.7%), the 18 - 24 group contained 40 respondents (16.9%) and the 35-44 group contained 33 respondents (14%). Most respondents were Irish (163), accounting for 69.1% of the sample, followed by the United States with 20 respondents (8.5%), and the UK with 11 respondents (4.7%). Regarding education, 126 respondents (53.4%) completed a bachelor’s degree, and 81 respondents (34.3%) completed a master’s degree.
Independent Variable
The use of transparency centred on Karlsson’s definition by focusing on disclosure and participatory transparency (2010), as this understanding relates directly to UGC, outlined in table 4 below. Across four fictitious news articles transparency was operationalised in different forms; non-transparent (control), disclosure transparency, participatory transparency, and full
transparency (combination of disclosure and participatory).
Dependent Variable
Kohring & Matthes’ (2007) standardised method for measuring trust in journalistic selectivity contains four first order factors. Trust in the selectivity of topics refers to the trust in a journalist for choosing topics for coverage, but was excluded here, like Koliska (2015), as this thesis doesn’t analyse trust over multiple news items, just one. Trust in the accuracy of
depictions evaluates journalistic work in verifying information. Trust in the selectivity of facts evaluates how facts relating to a topic are contextualised. Trust in journalistic assessment measures the “usefulness and appropriateness” (Kohring & Matthes, 2007) of journalistic commentary. Each factor is composed of variables that, together, produce a composite view of trust. Similar to Koliska (2015), only eleven of Kohring & Matthes’ (2007) sixteen questions
were employed in this experiment, seen in table 3 below. Respondents evaluated, on a scale of 1 to 5 (strongly disagree/agree), to what extent they agreed with the statements below. Together, they formed the measure for trust, my dependent variable.
Table 3 - Kohring & Matthes’ (2007) multifactor model of journalistic selectivity
Treatment
Transparency was operationalised in four different versions of the treatment (seen in table 4 below). All treatments contained identical news articles and UGC video from a fictitious news organisation about an airstrike on the Syrian town of Khan Shaykhun in November 2017 (see Appendix E). My content analysis findings informed the topic selection, confirming news coverage of Syria relies on UGC, therefore replicating a realistic scenario in how respondents encounter news about Syria. The control article suggested UGC video showed the aftermath of a “suspected government airstrike”. More transparent treatments provided additional information about the ‘local resident’ who recorded the video, links to external news sources, Google Satellite View of the location, and the ‘resident’s’ fictitious Twitter account.
Factors of Trust Items per Factor
Trust in accuracy of
depictions ● Information in the article would be verifiable if examined. ● The reported information is true. ● I think the article recounts the facts truthfully.
● Facts that I received regarding topic are correct. Trust in fact
selectivity ● The essential points are included. ● Focus of article is on important facts.
● All important information regarding the topic is provided. ● Reporting includes different points of view.
Trust in journalistic
assessment ● Criticism is expressed in an adequate and well-founded manner. ● Journalist’s assessments regarding the topic are useful. ● Journalist’s evaluations of the topic are well founded.
As a journalist, I produced the article, based on a New York Times investigation about an April 2017 airstrike on Khan Shaykhun (Browne, 2017), and reporting from my colleagues in Storyful News agency who spoke with the resident who filmed the video. Editors examined the article for accuracy and balance before it was amended to its current form.
Table 4 - Transparency features operationalised in treatments
Procedure
Respondents were invited to participate in a study about the quality of online journalism (see Appendix C). The survey first asked respondents demographic questions and then general
Transparency
Type Relationship to UGC Features of Transparency in Treatments N = 236
1. Non-transparent (control)
● No transparency features
● Text is non-specific about source, location, accuracy of video
N = 56 2. Disclosure
transparency ● Including hyperlink to original source of UGC, on social media like Twitter ● Including corroborating
news coverage that report on incident shown in the UGC ● Including hyperlinks to
independent sources of information that help verify video authenticity, like Google Satellite View
Includes all of 1, also:
● Hyperlink to “social media” where local resident posted video
● Hyperlink to corroborating Reuters report ● Hyperlink to Google Satellite, confirming
location of video
N = 57
3. Participatory
transparency ● Finding/contacting sources on social media, using their testimony to corroborate information about news event, asking permission for UGC.
● Embedding original versions of UGC in article, like videos shared on Twitter
Includes all of 1, also:
● Embedded tweets from the local resident, who is named, showing their original video, and their initial comments about incident, incorporated into news report ● Details about local resident posting video
on Twitter, including time and date ● Further quotes from local resident, speaking ‘directly’ with BTV, after contacting him on social media
N = 65
4. Full
questions about their level of media trust and preferred sources of news. Respondents were informed they would next read a news article about an incident in Syria. The survey platform randomly assigned respondents proportionally to the four treatments. Respondents then clicked through to a posttest survey (see appendix D), and were unable to return to the article.
Respondents first answered eleven questions relating to Kohring & Matthes’ (2007) trust scale before answering how much attention they gave the article, how transparent they considered the article, and how trustworthy they perceive articles with UGC. Trust and transparency questions were based on five-point Likert scales (strongly disagree/agree).
Manipulation Check
In the posttest survey, respondents were asked two questions as a manipulation check. A cross-tabulation evaluated the manipulation and investigated the relationship between the type of transparent article respondents encountered and whether they thought the journalist specified what social media platform the UGC originated on. The social media platform was only specified in two conditions (participatory and full transparency), and wasn’t specified in the other two (non-transparent and disclosure). A Pearson’s Chi-Square test showed this relationship is
statistically significant, χ2 (3) = 49.84, p = .000, indicating the manipulation was successful and respondents noted the mention, or lack of mention, of the social media platform accordingly, based on the transparency condition they were exposed to.
Measuring Trust
I conducted a principal components analysis to measure eleven trust items and create the dependent variable for overall trust. An eleven-item principal component analysis resulted in one component that explained 41.4% of variance in the items. For reliability of the eleven items,
Cronbach’s Alpha score = .85, indicating a strong, reliable scale. A new continuous dependent variable measured on a five-point Likert scale was computed and named Overall Trust in Journalism (M = 3.15, SD = 0.54).
Results: Testing the Relationship Between Transparency and Trust
Five hypotheses were proposed to test the effect of different forms of transparency on levels of trust, in relation to UGC in a news report. It was expected that both forms of
transparency specified in this research, disclosure (H1) or participatory (H2), would result in higher levels of trust among respondents versus the control, non-transparent article. It was then expected that a combined form of transparency (disclosure and participatory) would result in higher levels of trust, first versus the non-transparent article (H3), and also versus the separate forms of transparency of disclosure (H4) or participatory (H5). An analysis of covariance
(ANCOVA) test was conducted to investigate the relationship when controlling for two different covariates, the level of attention given to the article by respondents, and how important
respondents regarded transparency in journalism. A post-hoc Bonferroni test was used then. A Levene’s test showed a non-statistically significant value, p = .510, meaning equality of variances was assumed. ANCOVA analysis also reduces the probability of a Type II error.
Results indicate low variability in levels of trust between the four transparency conditions (Non-transparent article, M = 3.07, SD = 0.48; disclosure transparent article, M = 3.27, SD = 0.57; participatory transparent article, M = 3.05, SD= 0.51; fully transparent article, M = 3.21, SD = 0.57). Of all four forms of transparency, disclosure transparency resulted in the highest levels of trust among respondents.
transparent article received did not have a statistically significant effect on the overall levels of trust among respondents, F (3, 230) = 2.358, p = .053, the model as a whole was close to significance, and with a one-tailed hypothesis could be regarded as such.
Specifically regarding the covariates, analysis showed that when controlling for the level of attention given to the article by respondents, there is a significant effect of the type of
transparency on overall levels of trust, F (1, 320) = 4.557, p = .034. When controlling for how important respondents regarded transparency in journalism, there is no significant effect of transparency on the overall level of trust, F (1, 320) = 0.480, p = .489.
However, after a Bonferroni post-hoc test, no significant results were found, meaning the overall levels of trust reported by the respondents in the non-transparent group (Adj M = 3.07, SE = .071), disclosure transparency group (Adj M = 3.28, SE = .071), participatory transparency group (Adj M = 3.04, SE = .066), and full transparency group (Adj M = 3.21, SE = .070) did not differ significantly. The relationship that came closest to differing significantly was between those exposed to disclosure and participatory transparency conditions, p = .090. See appendix B for full results and descriptive statistics of this analysis.
The partial Eta squared value for the model shows a small effect size (.033), signifying that 3.3% of the variance in the dependent variable of trust is explained by the independent variable of transparency. A small effect size like this is not uncommon in media research.
After collecting the data, analysing the results, and capturing non-significant differences between transparency groups, even after controlling for covariates, no effects on audience trust were found. Therefore, the ANCOVA results mean the null hypothesis was retained and all
hypotheses (H1 - H5) were rejected.
Discussion
It has been argued that the incorporation of UGC into news reporting is an important development in digital journalism because it offers “proximity and immediacy” (Mast &
Hanegreefs, 2015, p. 595). Previous research has provided insights into how UGC is understood and employed as a tool of transparency (Hellmueller et al., 2013; Mast & Hanegreefs, 2015; Wardle et al., 2014). However, as a tool of transparency, research has yet to examine how UGC affects audience perceptions. This study sought to fill that gap and relied on a content analysis and experiment to assess how UGC is used and credited (study 1), and how such practices
impact audience trust in news reporting (study 2). The research and societal implications of these findings will be discussed here.
So, how is UGC used and credited by news organisations, and how does this compare with previous research? The content analysis demonstrated that the BBC, CNN, and Al Jazeera English regularly relied on UGC in their coverage of Aleppo, best reflected in the pre and post-ceasefire periods of December 14. When access was limited due to the conflict, UGC was
essential for news organisations. Afterwards, professional footage was preferred. This echoes the finding of Wardle et al. that “user-generated content is used when other images are not available” (2014, p. 39) and sheds light on my research question too. UGC is lauded for its authenticity, immediacy, and eyewitness qualities, but as demonstrated, UGC still represents just a stopgap, temporary method for coverage, rather than a new bastion for transparent journalism.
organisations almost all the time. This is surprising considering previous research (Harkin et al., 2012; Mast & Hanegreefs, 2015; Wardle et al., 2014) recorded exceptionally lower levels of UGC source attribution (detailed in theoretical framework), displaying a distinct lack of
transparency. For Mast & Hanegreefs, it is irresponsible of news organisations to “abstain from alerting audiences to the complexity of the sectarian Syrian conflict” (2015, p. 611) and assume their audience “will simply “trust” the news (agency)” (2015, p. 611) when UGC is used. For the BBC, CNN, and Al Jazeera English, has the lesson been learned?
In my sample, transparency, at least in the form of simple UGC source attribution, seems to be a regular routine among those three organisations now. However, it’s worth remembering this research comes over two years since the most recent study into UGC use in Syrian coverage. Since 2015, newsrooms have likely become more accustomed to sourcing and using UGC when reporting about Syria, so perhaps the level of crediting is particular to this story, but with a different conflict or news event, it might not be so high.
The level of crediting observed may be symptomatic of a limitation of this study. One restriction was applied to news articles with multiple UGC pieces, with only the first piece of UGC coded as this was regarded as the most newsworthy UGC. Still, it can be expected that the leading UGC item in news reports is considered the most important by news organisations and therefore, the best indicator of the degree of transparency employed in newsrooms. More time would have allowed for recording extra uses of UGC and provided more detailed findings about crediting practices, possibly identifying larger discrepancies between newsrooms.
Every newsroom is different. CNN policy states all footage not shot by CNN must be credited, regardless if it’s professional footage, or amateur UGC (Wardle et al., 2014). This
explains why CNN credited every piece of UGC, and demonstrates how newsroom policy directly influences reporting practices, further highlighting the variance between news organisations’ attitudes towards transparency.
These findings clearly show an industry standard towards crediting sources of UGC is lacking. UGC’s ability to heighten transparent news coverage should be recognised in newsroom policies as it encourages inventive reporting and engaging with people external to the newsroom. Yet, considerations over verifying UGC and standardising attribution should not be sacrificed in the name of speed and technological innovation. Although, would such policies affect audience perceptions?
In the next step, the effects of transparent uses of UGC, like crediting, on audience trust were tested in an experimental study. Trust among respondents was not significantly affected by any of the transparent treatments of UGC across four conditions, meaning the theoretical
assumptions tested in the hypotheses were not realised. Different trust levels were still observed, and though minimal, results indicated those exposed to disclosure transparency registered the highest levels of trust, while those exposed to participatory transparency registered the lowest, based on Kohring & Matthes’ (2007) model.
Comparing levels of trust placed in these two forms of transparency, people placed more trust in established, verified sources of information like Reuters, than the testimony of a
reportedly credible eyewitness who shared his experience on Twitter. Marginal differences apply, but perhaps the conclusion here is that to uncover the truth about an incident, people do place trust in a journalist’s selection of information, providing the journalist selects credible, reliable sources of information. UGC used without sufficient transparent verification and corroboration
might still be too much for some. Hellmueller et al. (2012) similarly found US journalists regarded disclosure transparency to be a stronger “truth-telling” strategy than participatory.
Surprisingly, respondents exposed to the control, non-transparent article registered even higher trust levels than the participatory group. Participatory respondents were given the same information as those in the control group, along with additional details about the video from the alleged source’s Twitter account. This further suggests that sources and information discovered on social media sow scepticism in the audience, casting doubts over the veracity of the facts presented. It’s possible the audience pays more attention to information sourced from official organisations than information sourced on social media. There was a significant relationship between transparency and trust, when accounting for the level of attention given to the news article by respondents.
It’s unknown if repeated exposure to news reports with social media features would diminish trust, but perhaps, repeated exposure to such a style of reporting could result in findings similar to Tsfati & Cappella (2003). That is, over time, the audience might become accustomed to reporting that includes social media features and scepticism may decrease, even if trust doesn’t ultimately increase.
Factors such as the level of attention given to a news article or how people value
transparency in journalism were shown to alter the relationship between transparency and trust. A single factor like transparency by itself may not be enough to induce trust. This was reflected in the small effect size observed, transparency explained just 3.3% of the variance in trust. Media consumption does not happen in a vacuum and small effects are common in media research when accounting for the impact of personal factors on media behaviour. Researchers state trust in
journalism is multifaceted (Grosser, 2016; Koliska, 2015) and affected by people’s
characteristics and preferences regarding news sources. As outlined above, modern societies are complex. For individuals to bridge the gap between knowing and not knowing (Simmel, 1950), they rely on trust, in society or journalism, yet this does not occur overnight.
Researchers have called for longitudinal studies (Karlsson et al., 2014; Koliska, 2015) yet none have looked at the relationship over time, indicating an obvious path for future research. As for news organisations who regularly use UGC, the path isn’t so clear. Consideration needs to be given to how they integrate UGC. Bringing about professional changes by making it central to reporting and cultural changes by recognising and explaining the value of UGC transparency to audiences seem like sensible ways of better embracing UGC, whilst fostering trust. To that end, to answer the research question, alone, transparent use of UGC has little impact in building trust in audiences. However, how does this compare to previous research?
The findings of my analysis mirror recent studies. The link between transparency and trust has still to be definitively proven. Meier & Reimer (2011) recorded small, but positive effects of transparency on trust in German audiences, though they focused on a different
conceptualisation of transparency, product and process transparency, which paid more attention to the selection of multiple news items regarding a topic, which is perhaps better suited to capture the nuances of trust. Koliska (2015) replicated their study, though he found no effects on US audiences and suggested cultural factors made the difference. Reflecting on the research design, an experiment was the most appropriate method to measure the effect of transparency as it allowed for a controlled setting, strong external validity, and the effects of different treatments could be measured on a randomised sample. To go a step further, longitudinal experimental
studies would best capture the potential cumulative effect of transparency on trust. Like this study, Meier & Reimer (2011) and Koliska (2015) used Kohring &
Matthes’ (2007) model to measure trust, but differently. That difference may shed further light on the relevant success of the German study versus the US study, and towards a limitation in my research. Meier & Reimer (2011) measured trust on a seven point scale, whereas Koliska (2015) used a five point scale, like this study. With a homogenous sample who were young, experienced with online media, educated, and already quite trustworthy, it’s possible a ceiling effect occurred and a five point scale wasn’t enough to reflect the distinctions in their levels of trust. However, general levels of trust were constructed from an 11 item scale from Kohring & Matthes’ (2007) model, so respondents’ scores represented reliable measures of trust, based on the treatment they were exposed to.
The sample also presented a limitation. A random sample was not achieved as the vast majority of respondents were recruited via social media, meaning those who don’t use social media couldn’t participate. A more inclusive sample would include people of varying
nationalities, socio-economic backgrounds, and ideally, levels of media trust or scepticism.
Conclusion
This research has illustrated one thing very clearly. Transparency does have an impact, be it on journalistic practice, as seen in the content analysis, or on audience perceptions, as seen in the experiment. Whether or not that effect is understood by audiences remains an open question. As detailed above, increased transparency is a consequence of advancements in digital
(Karlsson, 2011; Kovach & Rosenstiel, 2007), and also for individuals on all sides of the media. Research to date has focused on social and digital media’s impact on journalistic transparency and how their practices have changed. I agree with Wardle et al., (2014) that future research should undertake a greater examination of how “people formerly known as audience” (Rosen, 2006) have now become as much a part of the media themselves, and how people who capture and share UGC feel about this.
Transparency’s power to induce trust is regarded as a journalistic ideal by professionals and scholars alike. Former New York Times Public Editor Margaret Sullivan said as much when describing how journalists should bolster trust with readers by letting them “get to know their backgrounds, their personalities and how they do their jobs” (2013). However, the link has still to be scientifically proven, and remains just that, an ideal. Tools for transparency exist in every newsroom, arguably though, perhaps newsroom culture is yet to catch up. Rather than an ideal, recognising transparency as a professional norm, akin to balance, objectivity, or accuracy, and making it central to the journalistic process could help solidify its role. UGC represents one way to encourage that change by incorporating UGC transparency into reporting, sharing insights on sources, discovery, and verification to foster stronger relationships with the audience.
There is no silver bullet for building trust in the media. However, as UGC becomes a central newsroom resource, improper source attribution or verification serves to undermine its potential to connect journalists with their sources and audience. Transparency is just one tool in a journalist’s arsenal, but it can help news organisations to work with an audience that are no longer only consumers, but also contributors. In the current climate where people demand more
from journalists, as Mast & Hanegreefs conclude, increased transparency has the potential to “restore, rather than undermine the authority and trustworthiness of journalism” (2015, p. 597).
References
Agarwal, S., & Barthel, M. (2015). The Friendly Barbarians: Professional Norms and Work Routines of Online Journalists in the United States. Journalism, 16(3), 376-391. http://dx.doi.org.proxy.uba.uva.nl:2048/10.1177/1464884913511565 Allen, D. (2008). The Trouble with Transparency. Journalism Studies, 9(3), 323-340. Associated Press. (May, 2013). AP Social Media Guidelines Update, Including
Newsgathering in Sensitive Situations. Retrieved, October 16, 2017, from https://
blog.ap.org/announcements/ap-social-media-guidelines-update-including-newsgathering-in-sensitive-situations
BBC. (2006). User-Generated Content (Video, Audio and Stills Contributions From Members of the Public in BBC News Output). Retrieved, October 16, 2017, from http://www.bbc.co.uk/guidelines/editorialguidelines/assets/advice/
user_generated_content.pdf
BBC. (2017). Learn how the BBC is Working to Strengthen Trust and Transparency in Online News. Retrieved, January 20, 2018, from http://www.bbc.com/news/ help-41670342
Bivens, R. (2008). The Internet, Mobile Phones And Blogging. Journalism Practice, 2(1), 113-129.
Browne, M. (New York Times) (2017, May 1). The Times Uses Forensic Mapping to Verify a Syrian Chemical Attack. Retrieved November 17, 2017, from https://
www.nytimes.com/2017/05/01/insider/the-times-uses-forensic-mapping-to-verify-a-syrian-chemical-attack.html
Corker, B. [SenBobCorker]. (2017, October 8). It's a shame the White House has become an adult day care center. Someone obviously missed their shift this morning
[Tweet]. Retrieved from https://twitter.com/senbobcorker/status/ 917045348820049920
Deuze, M. (2005) What is Journalism? Professional Identity and Ideology of Journalists Reconsidered. Journalism, 6(4): 442–64.
Fukuyama, F. (1995). Trust: The Social Virtues and the Creation of Prosperity. New York, New York: The Free Press.
Fung, A., Graham, M., & Weil, D. (2007). Full Disclosure: The Perils and Promise of Transparency. New York, New York: Cambridge University Press.
Giddens, A. (1990). The Consequences of Modernity. Stanford, California: Stanford University Press.
Grosser, K. (2016) Trust in Online Journalism, Digital Journalism, 4(8): 1036-1057, https://doi.org/10.1080/21670811.2015.1127174
Harkin, J., Anderson, K., Morgan, L., and Smith, B. (2012). Deciphering User-Generated Content in Transitional Societies: A Syria Coverage Case Study. Center for Global Communication Studies, Annenberg School for Communication, University of Pennsylvania
Hassan, H. (The Guardian) (2016, December 18). Aleppo: Elegy for a Doomed City Whose History Spans Centuries. Retrieved, November 14, 2017, from https://
www.theguardian.com/world/2016/dec/18/aleppo-elegy-for-doomed-city-syria-assad
Hellmueller, L., & Li, Y. (2015) Contest Over Content: A Longitudinal Study of the CNN iReport Effect of the Journalistic Field. Journalism Practice. 9(5): 617-633 https:// doi.org/10.1080/17512786.2014.987553
Hellmueller, L., Vos, T., & Poepsel, M. (2013). Shifting Journalistic Capital? Journalism Studies, 14 (3): 287–304 http://dx.doi.org/10.1080/1461670X.2012.697686
Karlsson, M. (2010). Rituals of Transparency. Journalism Studies, 11(4), 535- 545. Karlsson, M. (2011). The Immediacy of Online News, the Visibility of Journalistic
Processes and a Restructuring of Journalistic Authority. Journalism, 12(3), 279-295.
Karlsson, M., Clerwall, C., & Nord, L. (2014). You Ain't Seen Nothing Yet. Journalism Studies, 15(5), 1-11. https://doi.org/10.1080/1461670X.2014.886837
Kohring, M. (2004). Vertrauen in Journalismus: Theorie und Empirie. Konstanz, Germany: UVK Verlagsgesellschaft.
Kohring, M., & Matthes, J. (2007). Trust in News Media: Development and Validation of a Multidimensional Scale. Communication Research, 34(2), 231-252.
Koliska, M. (2015). Transparency and Trust in Journalism: An Examination of Values, Practices, and Effects (Doctoral dissertation). Retrieved from https://
Kovach, B. & Rosenstiel, T. (2007). The Elements of Journalism: What News People Should Know and the Public Should Expect (2nd ed.). New York, New York: Crown.
Lewis, J., & Weigert, A. (1985). Trust as a Social Reality. Social Forces, 63(4), 967-985. Mast, J. & Hanegreefs, S. (2015). When News Media Turn To Citizen-Generated Images
of War. Digital Journalism, 3(4), 594-614
Meier, K. & Reimer, J. (2011). Transparenz im Journalismus: Instrumente,
Konfliktpotentiale, Wirkung. Publizistik, 56(2), 133-155. https://doi.org/10.1007/ s11616-011-0116-7
Newman, M. (Reuters Institute) (2017). Executive Summary and Key Findings. Retrieved, January 20, 2018, from https://reutersinstitute.politics.ox.ac.uk/sites/ default/files/Digital%20News%20Report%202017%20web_0.pdf
Parsons, T. (1978). Action Theory and the Human Condition. Berkeley, California: Free Press
Plaisance, P. (2007). Transparency: An Assessment of the Kantian Roots of a Key
Element in Media Ethics Practice. Journal of Mass Media Ethics, 22(2-3), 187-207 Reuters. (2017). Handbook of Journalism: A Brief Guide to the Standards and Values of
Reuters Video News. Retrieved, October 16, 2017, from http:// handbook.reuters.com/index.php?
title=A_Brief_Guide_to_the_Standards_and_Values_of_Reuters_Video_News Roberts, M. (2007). Measuring the Relationship Between Journalistic Transparency and
Rosen, J. (2006, June 27). The People Formerly Known as the Audience. Retrieved, January 12, 2018, from http://archive.pressthink.org/2006/06/27/ppl_frmr.html Rosen, J. (2010, November 10). The View from Nowhere: Questions and Answers.
Retrieved, November 8, 2017, from http://pressthink.org/2010/11/the-view-from-nowhere-questions-and-answers/
Singer, J. (1997). Still Guarding the Gate? The Newspaper Journalist's Role in an Online World. Convergence: The International Journal of Research into New Media Technologies, 13(1), 72-89. http://dx.doi.org/10.1177/135485659700300106 Simmel, G., in Wolff, K. H. (1950). The Sociology of Georg Simmel. Glencoe, Illinois:
Free Press.
Sullivan, M. (New York Times) (2013, January 5). When Reporters Get Personal. Retrieved, November 14, 2017, from http://www.nytimes.com/2013/01/06/public-editor/when-reporters-get-personal.html
Tsfati, Y. (2010). Online News Exposure and Trust in the Mainstream Media: Exploring Possible Associations. American Behavioral Scientist, 54(1), 22- 42.
Tsfati, Y., & Cappella, J. (2003). Do People Watch What They Do Not Trust? Exploring the Association Between News Media Scepticism and Exposure. Communication Research, 30(5), 504-529.
Tuchman, G. (1972). Objectivity as Strategic Ritual: An Examination of Newsmen's Notions of Objectivity. The American Journal of Sociology, 77(4), 660-679.
Wardle, C., Dubberley, S., & Brown, P. (Tow Center for Digital Journalism) (2014, December 3). Amateur Footage: A Global Study of User-Generated Content, Columbia University Academic Commons, https://doi.org/10.7916/D88S526V Weinberger, D. (2009). Transparency is the New Objectivity. Retrieved, November 10,
2017, from http://www.hyperorg.com/blogger/2009/07/19/transparency-is-the-newobjectivity
Appendices
Appendix A - Codebook for content analysis
Appendix B - Descriptive statistics for analysis of experimental results
Appendix C - Social media recruitment and information for prospective respondents Appendix D - Experiment pretest and posttest
Appendix E - Transparency treatments for experiment Appendix A - Codebook
Appendix B - Descriptive Statistics for Analysis of Experimental Results
Table 1: ANCOVA Results and Descriptive Statistics for Overall Levels of Trust by Transparency Type and Covariates, the Level of Attention Given to the Article, and How Important Respondents View Transparency in Journalism
Note. R2 = .05, Adj. R2 = .03. *p < .05
Table 2: Multiple Comparisons and Mean Differences in Overall Levels of Trust by Transparency type, Controlling for Covariates, the Level of Attention Given to the Article, and How Important Respondents View Transparency in Journalism
Note. DT = Disclosure transparency, PT = Participatory transparency, FT = Full transparency. * p < .05, where p-values are adjusted using the Bonferroni method.
DV: Type of Transparency IV: Overall Trust
Observed Mean Adjusted Mean SD n
Non-transparent 3.07 3.07 0.48 56
Disclosure 3.27 3.28 0.57 57
Participatory 3.05 3.04 0.52 65
Full transparency 3.21 3.21 0.57 58
Source SS df MS F
Level of attention given 1.29 1 1.29 4.58*
Importance of transparency 0.14 1 0.14 0.48
Type of transparency 2.20 3 0.73 2.6
Error 65.00 230 0.28
Comparison Mean Difference SE Bonferroni Adjusted 95% CI Non-transparent vs. FT -.13 .10 -.4, .13 Non-transparent vs. DT - .21 .10 -.47, .06 Non-transparent vs. PT .03 .10 -.23, .29 DT vs. PT .24 .10 -.02, .49 DT vs. FT .07 .10 -.19, .34 PT vs. FT -.16 .10 -.42, .09
Appendix C - Social Media Recruitment and Information for Prospective Respondents
Appendix D - Experiment Pretest and Posttest
1) Pretest
Appendix E - Transparency Treatments for Experiment
1) Control, non-transparent article 2) Disclosure transparent article
3) Participatory transparent article 4) Fully transparent article
