Anti-western partisan media websites in Lithuania : targeting issue publics and creating an echo-chamber of biased information

Master‘s Thesis

University of Amsterdam Graduate School of Communications

Anti-western Partisan Media Websites in Lithuania: Targeting Issue Publics and

Creating an Echo-chamber of Biased Information.

Name: Povilas Kriksciunas Student ID: 10915788

Master‘s programme: Communication Science: Political Communication Supervisor: dhr. dr. J.M. (Jonas) Lefevere

Date: 23 June 2016

Abstract.

This study has developed a framework to understand the strategies employed by online anti-western partisan media websites in order to gather a target audience by applying the theory of issue publics and how these websites attempt to persuade the audience into believing information with an anti-western agenda by connecting the reader to a network of other online anti-western partisan media websites. To do so, this study conducted a quantitative content analysis of articles published on online anti-western partisan media websites and Facebook pages, with posts referring to these articles, in order to answer the following research question: to what extent do online partisan media websites in Lithuania contain information with an western bias, to what extent do different online anti-western partisan media websites and Facebook pages provide web-links to other online anti-western partisan media websites, and to what extent do they address different issue publics? The results revealed three important findings. First, the Facebook pages aim to serve certain issue publics. Second, the Facebook pages do not have a clear anti-western bias, but the website articles referred to in these posts have a clear anti-western bias. Third, the external web-links provided in the posts of the Facebook pages and website articles form a network between these websites. The added value of this study consists of three factors. First, this study attempted to bridge a gap between the two academic theories of partisan media and issue publics. Second, this study has displayed how social media can be used to lure a reader in by appearing to present unbiased information, when in fact the websites the reader is directed to disseminate biased information. Third, this study provided some insights into how online partisan media websites can create an echo-chamber of biased information by providing web-links to one another in their articles.

Introduction.

In the past few years there has been a growing amount of information with an anti-western bias being spread in countries that previously belonged to the Cold War era‘s Eastern bloc (Miazhevich, 2014; Simons, 2015). In 2015, the Russian government increased the funding for its global English-speaking news channel RT to 300 million USD (Shuster, 2015), and there have been many smaller online websites with an anti-western bias appearing in vernacular languages in Eastern bloc states (Pancerovas, 2015). After the Russian military actions in Georgia and Ukraine, the governments of Estonia, Latvia, and Lithuania have called upon NATO and the European Union to help ensure security in the Baltic region (Pridham, 2014). Furthermore, they recognized an informational war as a direct threat to national security (Krutaine, 2015). Research on anti-western partisan media websites is lacking, yet it is critical for our understanding of their persuasion strategies.

This thesis will investigate the persuasion strategies of online partisan media websites with an anti-western bias in Lithuania and will contribute to the understanding of these socially relevant issues. Over the past few years anti-western websites have become a growing concern as they aim to influence public opinion on political issues and events (Bugorkova, 2015). Also, this thesis will address the growing influence of news content found online. Furthermore, it will look into the effects of online partisan media websites which can be very persuasive, yet misleading (Levendusky, 2013b).

Moreover, this study has academic relevance, as it looks into how the academic theory of “partisan media“ (Groeling, 2013; Van Boven et al, 2012; Westfall et al, 2015) can be used together with the application of the academic theory of “issue publics“ (Chen, 2012; Claassen & Nicholson. 2013; Henderson, 2013; Tremayne, 2006) in order to gather a target audience. To the author’s knowledge, these two academic theories have not been

thoroughly studied together before. I argue, that these partisan media websites appeal to the interests of different issue publics in order to gather their audience (Kim, 2009). They combine this issue-specific approach with a general bias in how they then cover these issues. By providing web-links to other anti-western partisan media websites, they create an “echo-chamber“ of biased information (Bakshy et al, 2015).

Goals of the research.

This research will explore and describe the persuasion strategies of online anti-western partisan media websites in Lithuania. It will look into how these websites gather a target audience by applying the theory of issue publics (Tremayne et al, 2006) and how they attempt to persuade the audience into believing information with an anti-western agenda by connecting the reader to other online anti-western partisan media websites. It does so through a quantitative content analysis of articles published on these websites, and Facebook pages with posts referring to these articles. For the purpose of this research, the following research question has been formed: To what extent do online partisan media websites in Lithuania contain information with an anti-western bias, to what extent do different online anti-western partisan media websites and Facebook pages provide web-links to other online anti-western partisan media websites, and to what extent do they address different issue publics?

Theoretical framework.

This theoretical framework section will first describe the growing influence of online media on public opinion and how online media allow people to self-select the information they are exposed to depending on their interests. Second, it will describe how partisan media websites can appeal to the interest of specific target groups, called issue publics. Third, it will look into how partisan media can target such an issue public by catering to its

needs and by offering desired arguments and conclusions. Finally, it will describe how partisan media can form a network of biased information, called an “echo-chamber“, and talk about how an echo-chamber with an anti-western bias may possibly be in existence in Lithuania. Finally, it will present hypotheses of this research based on the theoretical framework provided in this section.

News media have been described to serve as the link connecting the public opinion of citizens and the world of politics (Strömbäck, 2005). The media have a lot of influence in forming public opinion on political issues (Scheufele and Tewksbury, 2007), as news media have been found to be the main source of political information for the general public (Strömbäck, 2008). The most influential mediums have generally been referred to as “traditional media” – such as the written press, TV, and radio. However, in recent years, the Internet has become more and more frequently used by citizens in order to consume news (Lee and Chyi, 2015). Also, when compared to traditional media, online news is regarded as more credible than television or radio, but less credible than newspapers (Mitchelstein and Boczkowski, 2010). However, even though overall online news media have been perceived as a credible source of information, it also differs from traditional media in a crucial respect: the Internet provides people with a much greater ability to self-select the content which they wish to be exposed to, as content found online tends to be more personalized and caters to the interests of the consumer (Sunstein, 2001). Such personalization may even be generated automatically through computer algorithms which tracks the interests and preferences of a user and then selects content which is most likely to be engaging (Bakshy et al, 2015; Flaxman et al, 2013; Zuiderveen Borgesius et al, 2016). Therefore, many websites focus on content tailored to the interests of a specific audience. Such an audience, gathered based on its members’ interests on a certain topic, is called an issue public.

Issue publics.

Krosnick (1990) said that the citizenry can be divided into many small issue publics, according to certain issues which interest certain citizens. Kim (2009) described issue publics as groups of people who are intensely concerned about particular issues. Claasen and Nicholson (2013) identified members of issue publics as people who pay special attention to a particular issue. This is in line with Chen (2012) who stated that members of an issue public are people who find a certain specific issue important to them. People may become members of issue publics when they have material self-interests on certain issues (Converse, 1964; Krosnick, 1990; Price et al, 2006), when these issues are relevant to their personal values (Krosnick, 1990; Goidel and Nisbet. 2006; Kim, 2009, Wojcieszak, 2014), or when they adopt a stance on a particular issue from a group they feel a strong identification with (Converse, 1964; Krosnick, 1990; Price et al, 2006). Henderson (2013) stated that members of an issue public are people who are well-informed about a particular political issue, even though their overall knowledge of broad political issues may be low. Even though Krosnick (1990) said that a citizen can only be a member of a few issue publics, according to Chen (2012) a citizen may be a member of several issue publics. Chen (2012) identified that members of several issue publics tend to have higher involvement in online activities which allow them to find information and express their opinion on these issues, i.e. online blogs. Price et al (2006) pointed out that some citizens may be part of issue publics due to high media exposure and adoption of the main issues on the public agenda. However, members of issue publics are more prone to have selective-exposure bias and become more politically polarized (Chen, 2012; Iyengar et al, 2008). This can be used by online partisan media websites in order to appeal to a certain issue public and keep the audience exposed to one-sided information.

This research will look into how certain issue publics are targeted in order to gather an audience for a network of anti-western partisan media websites. It is hypothesized that these issue publics are first targeted on the social media network site Facebook through specific Facebook pages which are aimed to address a certain issue. Then, this audience is connected to a larger network of online anti-western partisan media websites through web-links provided in the posts of these Facebook pages. This leads to the first hypothesis:

H1: Facebook pages aim to serve certain issue publics. There is a significant difference in the main issues addressed in different Facebook pages.

Partisan media.

Even though by appealing to issue publics anti-western partisan media websites may attempt to draw in an audience, their key goal is to disseminate information with an anti-western bias. When it comes to news consumption, Levendusky (2013b) found that people have two broad classes of goals: to reach the correct conclusion, and to reach the desired conclusion. Even when consuming unbiased media, a person with a biased prior opinion will discard the arguments of the other side and only approve arguments supporting his own position. Therefore, such a person will feel that both goals have been reached. Furthermore, selective-exposure bias may be reinforced through self-selection of news content online and consumption of biased media. Partisan media, which provide biased information, help the audience reach the desired conclusion by catering to their bias.

To understand how this bias is reflected in the actual content of these online anti-western websites, I draw on literature on partisan media. According to Levendusky (2013a), news media outlets in the 21st century have taken a trend to not only report news, but to also take a certain position with which they want to present their coverage. This is in line with Conry and Moehler (2015) who found that partisan media tend to favor one side

of the argument and reinforce political bias through news consumption. Partisan media outlets are not necessarily concerned about presenting facts, but about offering a certain point of view and helping to make sense of the world around us (Levendusky, 2013b). According to Groeling (2013), this trend of media outlets not only reporting the news, but also taking a certain political stance in their reports, is similar to the way news media outlets used to work in the 19th century, when most news outlets used to favor a certain political organization. Levendusky (2013b) argues that viewership of partisan media may be driven by directional goals - the desire to reach the preferred conclusion, which fits with the viewer’s original opinion. Taber and Lodge (2006) found that people tend to seek out information which confirms their existing views rather than conflicting with them, this is called a confirmation bias. Partisan media appeal to their audience by providing information which their readers are seeking in order to confirm their views. Taber, Cann, and Kucsova (2008) found that when processing new information and evidence, people were unable to ignore their previous beliefs. This is in line with Taber and Lodge (2006) who found that people tend to accept information which supports their original opinion without being critical, but try to find counterarguments to information which conflict with their original opinion. Therefore, partisan media websites appeal to their audience by providing counterarguments to information which conflicts with their position. Also, viewership of partisan media tends to be highly driven by self-exposure (Taber and Lodge, 2006; Taber, Cann, Kucsova, 2008; Levendusky, 2013b). Westfall et al (2015) and Van Boven et al (2012) found that this creates a divide in society and makes the public perceive political polarization in a country even higher than it may actually be. Therefore, partisan media appeal to the audience and attempt to polarize them by providing one-sided information and counterarguments to the information of the other side. So far, the role of partisan media has been studied mostly in the context of the USA or western Europe. This

research will try to use the theories of partisan media by applying them in the context of one of the Baltic states and looking at online websites and Facebook pages which are expected to have an anti-western bias. Based on the aforementioned information, it can be hypothesized that online anti-western partisan media websites in Lithuania appeal to their audience’s confirmation bias by providing one-sided information and arguments to counter the position of the other side. On the other hand, Facebook pages, which are aimed at issue publics and not a polarized audience, may provide more neutral information, due to the fact that the issue publics which they target may be members of the general public and therefore not have a confirmation bias. This leads to the second hypothesis:

H2: The Facebook pages’ posts will not be found to have a clear anti-western bias, but the website articles provided in the posts will be found to have an anti-western bias.

Echo-chamber.

Finally, I posit that anti-western partisan media websites form a so-called “echo chamber”: when several channels of partisan media link and refer their audience to each other, this creates a network of biased media. Audiences who are tuned into such a network may feel like they are getting arguments from different sources and will therefore perceive the content found on this network as reliable. Such networks are known as “echo-chambers” of biased information.

According to Bakshy et al (2015), an echo-chamber is a network in which people are exposed to information from like-minded individuals. This is in line with Tremayne et al (2006), who found that user-generated political content online (i.e. blogs) tend to form networks between their portals (“blogrolls“) and that these networks are based on certain issue publics. Barbera et al (2015) described echo-chambers as networks where people are connected based on their political preference on social media platforms, i.e. Twitter. Some

portals then become centers of traffic for such networks and people need only a small amount of web-links to travel inside such a network (Tremayne et al, 2006). Levendusky (2013b) stated that partisan media help create a “self-protective enclave“ of biased information, which reconfirms the viewer‘s original opinion. Also, Barbera et al (2015) found that people with conservative political views were more likely to be a part of echo-chambers which are in line with their political views.

For the purpose of this research, this study will look into a possible online echo-chamber with an anti-western bias in Lithuania. This echo-echo-chamber is expected to be created through several mediums, such as certain pages found on the social media site Facebook and online anti-western partisan media websites. Facebook pages will be used to attract members of certain issue publics by providing seemingly neutral content on issues which the audience will find interesting, but will then direct the audience towards online partisan media websites with an anti-western bias. By following the web-links provided in the posts of a Facebook page, the reader will end up traveling through an “echo-chamber” of online partisan media websites with an anti-western bias. This leads to the third hypothesis:

H3: The external web-links provided in the posts of the Facebook pages and website articles form an echo-chamber network between these websites.

Method.

To analyze how online anti-western partisan media websites apply the theory of issue publics in order to gather a target audience, this study analyzed posts found on Facebook pages aimed at specific issue publics and website articles found in the web-links of these Facebook posts. To analyze how online partisan media website use web-links in order to create an echo-chamber of biased information, web-links found in the posts of the

Facebook pages and web-links found in the website articles were used to analyze how these websites refer to one another and connect the reader to a network of websites with an anti-western bias. The codebook used for this analysis is found at Appendix A. To analyze the content of the Facebook pages and website articles, this study used a quantitative content analysis method. This method of research is objective in its nature and provides high reliability and validity (Riffe et al, 2005, p. 35).

Data collection method.

The quantitative content analysis was carried out in two parts. The first part consisted of analyzing content posted in Facebook pages with an anti-western agenda in the Lithuanian language, the tone of the post (anti-western, neutral, or pro-western), the length of the post, the amount of attention the post received, web-links to online website articles provided in the posts, and the issues covered in the posts. The second part consisted of analyzing the content of website articles found in the web-links from the posts of Facebook pages’ posts (found in the first part of the research), the tone of the article (anti-western, neutral, or pro-western), the length of the article, the amount of attention the article received, and the web-links provided in the article.

An inter-coder reliability check was conducted. All research units were tested and coded independently by two different coders. In order to see whether the variables were reliable and could be analyzed, Krippendorff’s alpha (α) was calculated for all variables (see Appendix B). All of the variables had an α value above 0,7, many of the variables had an α value of 1,00.

Sampling method.

The sample of this study included Facebook pages in the Lithuanian language, which were expected to have a hidden anti-western agenda, and website articles found in the

web-links in the posts of these Facebook pages. The sampling and research was carried out in two parts. The first part included Facebook pages in the Lithuanian language, which were expected to have a hidden anti-western agenda in their posts. The amount of followers a Facebook page has served as the sampling criterion for choosing which pages were included in the sample - the pages with the highest number of people following their newsfeeds were included in the sample, as it was assumed that these pages also reach the largest audience. Due to the fact that there were no official statistics which Facebook pages have an anti-western bias in their newsfeed, the pages were selected by the author and may therefore be subject to a certain level of bias. However, some external user-generated lists of Facebook pages with an anti-western bias were used as an additional means of choosing which pages should be included in the sample, i.e. “EU Mythbusters“, “www.Unlike.lt“, and “www.Socialbakers.com”. The project “EU Mythbusters” is a weekly review which exposes anti-western propaganda online by debunking false information provided on various websites, blogs, and media portals. The project “www.Unlike.lt” lists different Facebook pages and warns people who follow them that these pages have a hidden anti-western agenda. The project “www.Socialbakers.com” offers a list of the most popular Facebook pages in Lithuania (and other countries) – this list was used to search for Facebook pages which the author had never heard of, even though most of them were not relevant to the political theme of this research. Overall, the selection of Facebook pages is considered to be representative of the population of Lithuanian Facebook pages with an anti-western agenda.

In the first stage of the research, the posts of six Facebook pages were coded in the time period of one year, from April 24th, 2015, to April 24th, 2016. Around 50 to 70 posts were coded from each Facebook page. Some Facebook pages had more posts than others, therefore the total number of all posts on a certain Facebook page in the selected time

period were counted before the start of the coding and then divided by 70 to calculate the sampling criterion for every Facebook page: therefore, if a Facebook page had a total of 70 posts in the selected time period - all posts were coded, if a Facebook page had a total of 130 posts in the selected time period – every other post was coded, resulting in a total of 65 posts coded for that Facebook page, etc. In the second stage of the sampling, online website articles which were found in web-links provided in the posts of the Facebook pages from the first stage were included in the sample. The variables which were used in the coding are described in the Codebook (see Appendix A).

Dependent variables.

Issue. The sampled Facebook posts and website articles were coded as addressing one or several issues in their texts (including images, video and audio material), i.e. internal politics, international politics, economy, military and defense, etc. These issues were identified depending on the topics that the posts or articles were based on. For the issue (topic) list see Appendix C. This was used to test hypothesis H1: Facebook pages aim to serve certain issue publics. There is a significant difference in the main issues addressed in different Facebook pages.

Tone. The sampled Facebook posts and website articles were coded as having an anti-western (= -1), neutral (= 0), or pro-anti-western tone (= 1). A Facebook post or website article was understood as having an anti-western tone, if the issues and problems it addressed were portrayed to have been caused by the policies of the European Union, NATO (or its countries), the “Western way of life” (capitalism), or the pro-western direction of the Lithuanian government; also, if a post or article demonized the pro-western government of Ukraine (i.e. calling it a fascist regime with no substantial evidence, etc.) or portrayed Russia as the better alternative to “the West”. However, if a Facebook post or website

article was critical of internal political issues in Lithuania, but with no regards to the European Union or NATO bearing responsibility for these issues, it was coded as neutral. This was used to test hypothesis H2: The Facebook posts will not be found to have a clear anti-western bias, but the website articles provided in the posts will be found to have an anti-western bias.

Independent variables.

For Facebook posts and website articles the following independent variables were coded:

Facebook page. This variable tracked which Facebook page the post appeared on. It took a nominal value of 1 to 6, with 1 being “! Peticija PRIEŠ priverstinę imigraciją !”, 2 being “Mes už litą Lietuvoje”, 3 being “Jungtinės Karalystės Lietuvių Bendruomenė”, 4 being “Lietuvos ŽEMEI reikia savanorių”, 5 being “Lietuvos žiniasklaida skleidžia propaganda”, and 6 being “Europos sąjungos minusai ir pliusai”.

Website. The website domain where the website article was hosted, i.e. domain.com; for public platforms the base extension was included, i.e. youtube.com/channel.

Web-link. The web-link, found in a Facebook post or website article, which directs the reader to a website article, i.e. domain.com/article.

For Facebook posts and website articles the following control variables were coded:

Length. The amount of words in a Facebook post or website article. Pictures, videos, and audio were not counted as words. Also, if a Facebook page had only shared a post of another Facebook user or page, and not written added any text on their own behalf, the length of the post was coded as “0”.

Attention. The amount of “reactions” (previously known as “likes”), “shares”, and “comments” under a Facebook post or website article were coded as a means of tracking how engaging the Facebook post or website article was with its audience. The sum of all “reactions”, “shares”, and “comments” was used as a single variable.

Main issue of Facebook page. Also, a specific issue was identified with each Facebook page, according to its name, in order to serve the page‘s issue public. These issues were identified as the main issues for each Facebook page:

Table 1. Main issues identified per Facebook page

Facebook page English translation of

Facebook page name Main issue “! Peticija PRIEŠ priverstinę

imigraciją !”

“Petition against forced immigration (refugee

quotas)“

Social relations

“Mes už litą Lietuvoje“

“We‘re for the litas (national currency) in

Lithuania“

Economy “Jungtinės Karalystės Lietuvių

Bendruomenė“

“Lithuanian community

in the United Kingdom“ International politics “Lietuvos ŽEMEI reikia

savanorių“

“The land of Lithuania is

in need of volunteers“ Internal politics “Lietuvos žiniasklaida skleidžia propagandą“ “Lithuanian media is disseminating propaganda“ Communication “Europos sąjungos minusai ir

pliusai“

“The pros and cons of

the European Union“ International politics

Results.

H1 expected that: “Facebook pages aim to serve certain issue publics. There is a significant difference in the main issues addressed in different Facebook pages“. In order to test this hypothesis, the percentage of how often the main identified issue was mentioned in a post on its designated Facebook page was compared to the mean percentage of the same issue being mentioned in a post, on average, on any of the other Facebook

pages. This was done by recoding each Facebook page into a dummy variable in which it was compared to the other Facebook pages. The results for this are displayed in Figure 1.

Figure 1. Frequency of main issue being mentioned.

Overall, these results seem to support H1, except for one page: “Jungtinės Karalystės Lietuvių Bendruomenė”. This is because this page had a lot of posts on the issue of local events (0,28 versus 0,012 average) and “human interest” (0,4 versus 0,015 average) with posts on topics such as dating, food, travelling, funny pictures, etc. Yet, this should have been anticipated, as the page is aimed at the local Lithuanian community living in the United Kingdom, and therefore talked not only about political topics, but had posts to engage with the local community as well. The results for the page “Europos sąjungos minusai ir pliusai“ were not significant, but it is worth noting that the main issue of “international politics” was still mentioned at an above-average frequency. Also, this page had a high frequency of the issue “societal relations” being mentioned in a post (0,286

! Peticija PRIEŠ priverstinę imigraciją !

Mes už litą Lietuvoje Jungtinės Karalystės Lietuvių Bendruomenė Lietuvos ŽEMEI reikia savanorių Lietuvos žiniasklaida skleidžia propagandą Europos sąjungos minusai ir pliusai Average 0,12 0,04 0,34 0,29 0,03 0,32 Issue 0,75 0,72 0,20 0,48 0,78 0,36 p 0,000 0,000 0,026 0,011 0,000 0,528 t -11,354 -10,388 2,281 -2,610 -15,381 -0,634 0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00 Pr o b ab ili ty

versus 0,12 average), mostly because the page approached posts concerning refugees from a societal relations standpoint, i.e. integration of migrants or refugees.

H2 expected that: “The Facebook pages’ posts will not be found to have a clear western bias, but the website articles provided in the posts will be found to have an anti-western bias“. In order to test this hypothesis, the overall tone of all Facebook posts and all website articles provided in the web-links of the Facebook posts were compared in a paired-sample T-test. The results were found to be statistically significant (T(316) = 7,129, p = 0,000, 95%CI [0,176 ; 0,310]), and the overall tone of all Facebook posts was found to have a mean value of -0,21, while the overall tone of the website articles provided in the web-links of the Facebook posts was found to have a mean value of -0,46. Overall, these results support H2, because the tone of the website articles found in the web-links of the posts was found to be significantly more anti-western than the tone of the Facebook posts. However, it is interesting to note, that while the tone of the website articles was found to be more anti-western than the tone of the Facebook posts, the overall tone of Facebook posts was still found to be anti-western. This is due to the fact that all of the Facebook posts were either neutral (73%) or had an anti-western tone (27%), while most of the website articles were either neutral (46,87%) or anti-western (50,16%). Only a small portion of the website articles had a pro-western tone (2,97%).

The same procedure was applied to calculate the differences in the overall tone of the Facebook posts and the website articles provided in the web-links of these posts per each Facebook page. These results are displayed in Figure 2.

Figure 2. Overall tone of Facebook posts compared to tone of website articles found

We can see that all Facebook pages had significantly more website articles with an anti-western tone provided in their web-links when compared to their posts, even though the results of the Facebook pages “Jungtinės Karalystės Lietuvių Bendruomenė“ and “Lietuvos žiniasklaida skleidžia propagandą“ were not significant. Overall, these results support the hypothesis that Facebook pages lure their audience in by appealing to issue publics and only then expose them to an anti-western bias in websites provided in their web-links.

H3 expected that: ”The external web-links provided in the posts of the Facebook pages and website articles form an echo-chamber network between these websites.“ In order to test this hypothesis, it was split between web-links provided in the Facebook posts and web-links provided in the website articles themselves. Therefore, H3a expected that: “The amount of web-links provided in the Facebook posts pointing to websites within the echo-chamber is higher than the amount of web-links pointing to other websites“. First, all web-links provided in Facebook posts were recoded into two groups: “outside” (=0) or

! Peticija PRIEŠ priverstinę imigraciją !

Mes už litą Lietuvoje Jungtinės Karalystės Lietuvių Bendruomen ė Lietuvos ŽEMEI reikia savanorių Lietuvos žiniasklaida skleidžia propagandą Europos sąjungos minusai ir pliusai Facebook -0,37 -0,41 -0,11 -0,12 -0,17 -0,11 Web-links -0,67 -0,69 -0,14 -0,45 -0,30 -0,46 p 0,002 0,002 0,768 0,000 0,086 0,000 t 3,298 3,266 0,298 5,000 1,738 4,660 -1,00 -0,90 -0,80 -0,70 -0,60 -0,50 -0,40 -0,30 -0,20 -0,10 0,00 Ov e ral l to n e , an ti -we ste rn

“within“ (=1). The “outside” group contained all links outside of the echo-chamber – links to domestic and foreign mainstream media websites, Facebook profile pages of people whose posts were shared on the Facebook pages but who themselves had no ties to any of the pages within the echo-chamber, and other random websites. The “within” group contained links that were deemed to be part of the echo-chamber - other Facebook pages and websites with an anti-western bias, Facebook profile pages, YouTube channels of people who have close ties with the anti-western websites (owners or contributors), and foreign English or Russian media websites that were found to have an anti-western bias1.

In order to test H3a, the amount of web-links which point to websites outside or within the echo-chamber were compared in a one-sample T-test. After running the analysis, it was found that 70,19% of the web-links pointed to websites in the “within” group, the results were significant, the confidence interval does not overlap with 50% (T(519) = 34,959, p = 0,000, 95%CI [,6625; ,7414]).

The distinction between the “within” and “outside” group of websites was made on the assumption that there would be a difference in the tone of the article. In order to test this assumption, an independent sample T-test was run. After running the analysis, it was found that website articles in the “within” group had an anti-western tone of -0,6 (with -1 being anti-western), compared to -0,1 of website articles in the “outside” group. These results were significant (T(174) = 8,106, p = 0,000, 95%CI [,384; ,632]).

H3b expected that: “The amount of web-links, which were found in website articles from the Facebook posts, pointing to websites within the echo-chamber is higher than the amount of web-links pointing to other websites“. First, all websites, which were found in the Facebook posts, were recoded into two groups: “within” or “outside” of the

1 _{Even if foreign English and Russian media websites were excluded from the analysis, the results remained}

chamber. Second, the web-links found in these website articles were recoded into two groups: “link-within” or “link-outside”. The criteria, for which web-links belong within or outside of the echo-chamber was the same as for H3a – these websites were known to have an anti-western bias or were closely affiliated to these websites (i.e. Facebook page or YouTube channel which belongs to the same website or content contributor of the website), therefore, it is possible that some additional websites with an anti-western bias were not included in the “echo-chamber”, because these web-links were not investigated any further. After running a cross-tabulation analysis, it was found that website articles which belong to the echo-chamber provided a similar amount of web-links in both groups: 251 web-links outside and 299 web-links within the echo chamber2.

In order to test H3b, the web-links from websites within the echo-chamber were compared to see how many pointed to websites within or outside of the echo-chamber. This was done with a single-sample T-test. After running the analysis, it was found that 54% of the web-links, which were found in website articles from within the echo-chamber, pointed to other websites within the echo-chamber, and 46% pointed to websites outside of the echo-chamber. These results were significant (T(549) = 25,573, p = 0,000, 95%CI [,50; ,59]), however it can be seen that the amount of web-links pointing within and outside of the echo-chamber were very similar.

Even though the amount of web-links pointing outside of the echo-chamber was relatively high, overall, these results support H3 that external web-links provided in the posts of the Facebook pages and website articles form an echo-chamber network between these websites.

2 _{In comparison, website articles from outside of the echo-chamber provided 234 web-links outside and only}

Overall, the conclusion regarding H3 seems to indicate that websites provide more web-links to other websites within the echo-chamber. However, it could be that this tendency is greater for some websites than others. To test this, a linear regression analysis was run, in order to determine how the proportionate amount of web-links within and outside of the echo-chamber provided in the website articles correlated with a particular website, the anti-western tone of the article (recoded values: anti-western = 1, neutral = 0, pro-western = -1), the length of the article (amount of words), and the amount of attention (sum of Facebook “reactions”, Facebook ”shares”, and comments) the article received. The dependent variable of the proportionate amount of web-links within and outside of the echo-chamber was coded by calculating the mean value of the web-links provided in an article, in regards to whether these links were pointing outside (=0) or within the echo-chamber (=1). To get estimates for all websites, I ran the model for each website, always including a dummy variable indicating whether the article appeared on that website (1) or on any of the other websites (0). The results for this are found in Table 2.

Table 2. Linear regression analysis results for websites. Coefficients are standardized

regression coefficients.

Sarmatai.lt Faktai.lt Kulgrinda.lt Interesas.lt Ekspertai.eu Website -,353** -,255* ,405*** ,249* -,422*** Attention -,322** -,265* -,202 -,247* -,108 Length ,143 ,117 ,008 ,196 ,06 Tone -,045 -,013 -,064 -,065 -,078 Adj. R² ,206 ,102 ,217 ,141 ,220 Note: * = p < .05, ** = p < .01, *** = p <.001.

It can be seen that the proportionate amount of web-links pointing within or outside of the echo-chamber in a website article depended on the website itself. Some websites (kulgrinda.lt, interesas.lt) had a higher proportionate amount of web-links pointing to websites within the echo-chamber, while others (sarmatai.lt, faktai.lt, ekspertai.eu) were more prone to use web-links pointing outside of the echo-chamber. It can also be seen that the amount of attention a website article received correlated with how low the proportionate amount of web-links pointing to websites within the echo-chamber was. This means that website articles, which were more prone to use web-links pointing outside of the echo-chamber, received more attention from the readers. This is an interesting finding, because, usually, web-links to websites pointing outside of the echo-chamber were used to mainstream media website articles, in order to either criticize the mainstream media article itself, or to use a mainstream media article to support own arguments presented in the website article (i.e. arguing against accepting refugees and providing mainstream media articles about crimes which were committed by refugees).

I accepted H1, because, overall, Facebook pages were found to address their main issue in their posts at a substantially higher rate than other Facebook pages. However, to test whether these results hold, if we control for the length of the Facebook posts and their tone (anti-western = 1, neutral = 0, pro-western = -1), a logistic regression was run. The results for this are found in Table 3. The Facebook pages were added as a categorical variable, with “Europos sąjungos minusai ir pliusai” set as the reference category. This choice was made based on the results displayed in Figure 1: for this Facebook page, the attention to its main and other issues was relatively similar, so it makes for a good reference to compare the other Facebook pages to.

Table 3. Logistic regression analysis results for Facebook pages.

Independent variable B S.E. Wald

Constant -,832** ,264 9,939

Facebook page (reference: Europos sąjung…) 53,890

- Peticija PRIEŠ 1,385*** ,385 12,965

- Mes už litą 1,297** ,420 9,540

- Jungtinės Karalystės -,944* ,449 4,410 - Lietuvos ŽEMEI ,580 ,388 2,237 - Lietuvos žiniasklaida 1,861*** ,378 24,223 Tone of post 1,232*** ,310 15,759 Length of post ,001 ,002 ,241 Pseudo R² (Nagelkerke) Note: * = p < .05, ** = p < .01, *** = p <.001.

The analysis gives mostly consistent results with the bivariate tests: compared to the reference Facebook page (“Europos sąjung…”), “Peticija PRIEŠ…”, “Mes už litą…”, and “Lietuvos žiniasklaida...“ have significantly more attention to its main issue. However, “Jungtinės Karalystės...“ provided significantly negative results - as it also did in Figure 1. Lastly, the results for “Lietuvos ŽEMEI…“ were not significant. Overall, these results support H1. In terms of control variables, the length of the post does not seem to have a significant effect on mentioning the main issue or not, while the tone of the post does suggest that the main issue of a Facebook page is significantly more likely to be mentioned in a Facebook post, if the tone of the Facebook post is anti-western.

Conclusion and discussion.

This study, on a basis of a quantitative content analysis, has developed a framework to understand the strategies employed by online anti-western partisan media websites in order to gather a target audience by applying the theory of issue publics (Tremayne et al, 2006) and how these websites attempt to persuade the audience into believing information with an western agenda by connecting the reader to a network of other online anti-western partisan media websites. This study has societal relevance, as it addressed online anti-western websites, which have become a growing concern in the Baltic region of NATO and the European Union, as they aim to influence public opinion on political issues and events (Bugorkova, 2015; Jackūnaitė, 2014; Krutaine, 2015; Pancerovas, 2015; Shuster, 2015; Simons, 2015). The results of this study contribute to the academic theory of “partisan media“ (Groeling, 2013; Van Boven et al, 2012; Westfall et al, 2015) by looking into how online partisan media websites may apply the theory of “issue publics“ (Chen, 2012; Claassen & Nicholson. 2013; Henderson, 2013; Tremayne, 2006) in order to gather a target audience.

The framework of this study was developed through a quantitative content analysis of articles published on online anti-western partisan media websites and Facebook pages, with posts referring to these articles, in order to find out to what extent do online partisan media websites in Lithuania contain information with an anti-western bias, to what extent different anti-western partisan media websites and Facebook pages provide web-links to other online anti-western partisan media websites, and to what extent they address different issue publics. The results revealed three important things. First, the Facebook pages investigated in this study aim to serve certain issue publics: there are substantial differences in the main issues addressed in different Facebook pages. Second, even though the overall tone of the posts on these Facebook pages was anti-western, the Facebook

pages’ posts do not have a clear anti-western bias, but the website articles referred to in these posts have a clear anti-western bias. Third, the external web-links provided in the posts of the Facebook pages and website articles form an echo-chamber network between these websites: the amount of web-links provided in both the Facebook posts and website articles pointing to websites within the network is higher than the amount of web-links pointing to websites outside of the network.

Indeed, in relation to the first hypothesis, it was found that these Facebook pages tend to appear to be more neutral in the information they provide, when compared to the online anti-western websites to which they direct the reader. The website articles had a substantially higher anti-western tone when compared to the Facebook posts themselves. With that being said, the results of two of the six Facebook pages were not significant, so while the overall pattern seems to confirm the expectations, there are differences. This can be explained by the fact that one of the Facebook pages targeted a local Lithuanian community living in the United Kingdom, and did not solely address politically themed issues, while appealing more to the themes of human interest and local events. The other Facebook page, which did not provide significant results to the first hypothesis, focused more on the issue of internal politics and, therefore, was not found to be as anti-western as it did not refer to events concerning international politics and the West as much as the other Facebook pages. The other Facebook pages, which provided significant results, were still found to have, on average, an anti-western tone, however, the online website articles to which they provided web-links to in their posts were found to be significantly more anti-western than the posts in the Facebook pages. This can be explained by the fact that Facebook pages are aimed at a much wider audience of issue publics than the website articles and therefore have a more neutral tone. Then, once this audience is lured in by the Facebook pages and click on a website article in a post, they are taken to the online

anti-western partisan media websites, which present one-sided information with an anti-anti-western bias.

Furthermore, the third hypothesis that the Facebook pages and online anti-western partisan media websites form an echo-chamber (Bakshy et al, 2015; Levendusky, 2013b; Tremayne et al, 2006) was supported. However, the Facebook pages themselves only referred to the echo-chamber in a single direction and were not referred to by the online anti-western partisan media websites. This can be explained by the fact that the Facebook pages are only used as a means to gather an initial target audience through the theory of issue publics, which at first may be more neutral, but then becomes polarized by being exposed to one-sided information in the online anti-western partisan media websites, and become more prone to seek out additional one-sided information in order to confirm their newly formed anti-western bias. In other words, the Facebook pages served as an entry point to the echo-chamber, which would then provide web-links to other online anti-western partisan media websites inside this echo-chamber.

However, while Facebook pages only referred to the echo-chamber in a single direction, they were much more prone to provide links to websites within the echo-chamber than outside. In comparison, and somewhat contrary to the expectation, the online anti-western partisan media websites were much more prone to provide web-links to websites outside of the echo-chamber, especially mainstream media articles. Still, such web-links usually appeared in case a negative story was published in the mainstream media: the story would then be used in the online anti-western partisan media websites to depict the West and its policies in a negative light (i.e. to argue against accepting refugees and providing mainstream media articles about a crime which was committed by a refugee, or to argue against using the euro currency while pointing to problems in the eurozone, etc.). It can be argued that in this manner the online anti-western partisan media websites

made their arguments appear more credible, as they provided sources to information which is confirmed in the mainstream media and the general public. This can be confirmed by the fact that online anti-western partisan media websites (i.e. ekspertai.eu) which provided more web-links to websites outside of the echo-chamber, also received more attention for their articles.

Though the above results do further our understanding of the strategies employed by anti-western online media portals, there are some limitations to this study. First, these findings represent only a small scale of online anti-western partisan media websites and Facebook pages in Lithuania. There could be other echo-chambers in existence, but are unknown to the author, as web-links to these websites may not be shared on social media sites like Facebook and they may not have been displayed in the lists of online anti-western websites (“www.Unlike.lt“, “EU MythBusters“, and “SocialBakers.com“). Therefore, the results from the sample of this research may not be generalizable to the whole population, as other online anti-western partisan media websites may employ different tactics in order to attract and persuade their audience.

Also, due to the fact that it was unknown which online anti-western partisan media websites were part of an echo-chamber, the Facebook pages and the posts in these pages served as the selection criterion for these websites. Therefore, for future research, the sampling could be more precise by coding only articles in online anti-western partisan media websites which are known to be part of an echo-chamber. This way, it could be ensured that all relevant articles are coded, because due to the sampling criteria of this study (through posts in Facebook pages) some relevant articles may have been missed, if web-links to these articles were not provided in one of the posts of the Facebook pages. Furthermore, the hypothesis that Facebook pages target issue publics in order to lure them into the websites may be tested more accurately by looking at additional variables, such as

the date of when the article was published and how soon it was shared in a post on one of the Facebook pages, or whether older articles are shared in the Facebook posts due to some event taking place in the world which may cause the public to seek out more information on a certain issue. Finally, other social media sites other than Facebook may be used by online partisan media websites in order to gather an audience by applying the theory of issue publics, i.e. Twitter. For this study, the social media site Facebook was chosen due to the fact that it is the most popular social network among Internet users in Lithuania, with 76% using this social network, while Twitter is not used widely by the general public – only 2% of Internet users have an account on this social network (Social Network Usage Statistics, 2015). Therefore, in other countries where social media preference among Internet users is different, different tactics may be used by online partisan media websites in order to attract an audience, depending on the peculiarities of the social network used and on the political context of the country.

In conclusion, this study has developed a framework in order to understand how online anti-western partisan media websites gather a target audience by using the theory of issue publics and how they attempt to persuade the audience into believing information with an western agenda by connecting the reader to a network of other online anti-western partisan media websites. Accordingly, the added value of this study consists of three factors. First, to the author’s knowledge, the theory of “issue publics“ has not been applied to the theory of “partisan media“ before. This is the first study which combined these two academic theories in such a manner. Therefore, this study attempted to bridge a gap between these two academic theories in the context of online partisan media websites and social media platforms. Second, this study has displayed how social media networks can be used to direct a reader who is a member of an issue public by appearing to present unbiased information, when in fact the websites the reader is being directed to disseminate

biased information. Third, this study provided some insights on how online partisan media websites can create an echo-chamber of biased information (in this case, with an anti-western bias) by gathering a target audience through the theory of issue publics. Also, while there has been a lot of researched carried out on partisan media in the context of the USA or western Europe, not much research has been carried out on how partisan media works in the context of Lithuania. Further research could be conducted to test the findings of this study, such as whether social media pages can be used as a persuasion tool in order to gather an originally neutral target audience by catering to issues of their concern and appearing to provide neutral information, but then re-directing them to multiple partisan media websites which provide biased information in order to polarize the originally neutral target audience.

Bibliography.

Bakshy, E., Messing, S., & Adamic, L. A. (2015). Exposure to ideologically diverse news and opinion on Facebook. Science, 348(6239), 1130-1132.

Barberá, P., Jost, J. T., Nagler, J., Tucker, J. A., & Bonneau, R. (2015). Tweeting From Left to Right Is Online Political Communication More Than an Echo Chamber?. Psychological science.

Bugorkova, O. (2015, March 19). Ukraine conflict: Inside Russia's 'Kremlin troll army'. BBC News. Retreived from:

http://www.bbc.com/news/world-europe-31962644 (accessed March 24, 2016).

Chen, H. T. (2012). Multiple issue publics in the high-choice media environment: Media use, online activity, and political knowledge. Asian Journal of Communication, 22(6), 621-641.

Claassen, R. L., & Nicholson, S. P. (2013). Extreme Voices Interest Groups and the Misrepresentation of Issue Publics. Public Opinion Quarterly, 77(4), 861-887.

Conroy-Krutz, J., & Moehler, D. C. (2015). Moderation from bias: A field experiment on partisan media in a new democracy. The Journal of Politics, 77(2), 575-587.

Converse, P. E. (2006). The nature of belief systems in mass publics (1964). Critical Review, 18(1-3), 1-74.

Pridham, G. (2014). EU/Ukraine Relations and the Crisis with Russia, 2013-14: A Turning Point. The International Spectator, 49(4), 53-61.

Flaxman, S., Goel, S., & Rao, J. M. (2013). Ideological segregation and the effects of social media on news consumption. Available at SSRN 2363701.

Goidel, K., & Nisbet, M. (2006). Exploring the roots of public participation in the controversy over embryonic stem cell research and cloning. Political Behavior, 28(2), 175-192.

Groeling, T. (2013). Media bias by the numbers: Challenges and opportunities in the empirical study of partisan news. Political Science, 16(1), 129.

Henderson, M. (2014). Issue publics, campaigns, and political knowledge. Political Behavior, 36(3), 631-657.

Iyengar, S., Hahn, K. S., Krosnick, J. A., & Walker, J. (2008). Selective exposure to campaign communication: The role of anticipated agreement and issue public membership. The Journal of Politics, 70(01), 186-200.

Jackūnaitė. (2014, May 07). Karai dėl protų: kaip propaganda stengiamasi kelti isteriją. LRT. Retreived from:

http://www.lrt.lt/naujienos/lietuvoje/2/42834#wowzaplaystart=5039000&wowzaplaydu ration=170000

Kim, Y. M. (2009). Issue publics in the new information environment selectivity, domain specificity, and extremity. Communication Research, 36(2), 254-284.

Krosnick, J. A. (1990). Government policy and citizen passion: A study of issue publics in contemporary America. Political behavior, 12(1), 59-92.

Krutaine, A. (2015, March 12). Latvia minister cites threat from 'information war' over Russian minority. Reuters. Retreived from:

Lee, A. M., & Chyi, H. I. (2015). The Rise of Online News Aggregators: Consumption and Competition. International Journal on Media Management, 17(1), 3-24.

Levendusky, M. (2013a). Partisan media exposure and attitudes toward the opposition. Political Communication, 30(4), 565-581.

Levendusky, M. S. (2013b). Why do partisan media polarize viewers?. American Journal of Political Science, 57(3), 611-623.

Miazhevich, G. (2014). Russia Today's coverage of Euromaidan. Russian Journal of Communication, 6(2), 186-191.

Mitchelstein, E., Boczkowski, P. J. (2010). Online news consumption research: An assessment of past work and an agenda for the future. New Media & Society.

Oxford Dictionaries. Oxford University Press.

Retreived from: http://www.oxforddictionaries.com/definition/english/anti-western (accessed January 17, 2016)

Pancerovas, D. (2015, August 03). Lietuvos radikalai naudoja rusiškos ir sovietinės propagandos technologijas. 15min.lt. Retreived from:

http://www.15min.lt/naujiena/aktualu/karo-zona/a-784-518608 (accessed January 15, 2016).

Price, V., David, C., Goldthorpe, B., Roth, M. M., & Cappella, J. N. (2006). Locating the issue public: The multi-dimensional nature of engagement with health care reform. Political Behavior, 28(1), 33-63.

Riffe, D., Lacy, S., & Fico, F. G. (2005). Analyzing Media Messages: Using

Quantitative Content Analysis in Research, Lawrence Erlbaum Associates. Mahwah-New Jersey.

Scheufele, D. A., Tewksbury, D. (2007). Framing, agenda setting, and priming: The evolution of three media effects models. Journal of Communication, 57, 9-20.

Shuster, S. (2015, March 05). The global news network RT is the Russian

government's main weapon in an intensifying information war with the West—and its top editor has a direct phone line to the Kremlin. Time.com. Retreived from:

http://time.com/rt-putin/

Simons, G. (2015). Perception of Russia's soft power and influence in the Baltic States. Public Relations Review, 41(1), 1-13.

Social Network Usage Statistics Using Desktop in Lithuania. (2015). Retreived from: https://www.statsmonkey.com/sunburst/21630-lithuania-desktop-social-network-usage-statistics-2015.php

Strömbäck, J. (2005). In Search of a Standard: four models of democracy and their normative implications for journalism. Journalism Studies, 6(3), 331-345.

Strömbäck, J. (2008). Four phases of mediatization: An analysis of the mediatization of politics. The International Journal of Press/Politics, 13(3), 228-246.

Sunstein, C. (2001). Fragmentation and cybercascades. In Republic.com (pp. 51-88). Princeton University Press.

Taber, C. S., Cann, D., & Kucsova, S. (2009). The motivated processing of political arguments. Political Behavior, 31(2), 137-155.

Taber, C. S., & Lodge, M. (2006). Motivated skepticism in the evaluation of political beliefs. American Journal of Political Science, 50(3), 755-769.

Tremayne, M., Zheng, N., Lee, J. K., & Jeong, J. (2006). Issue publics on the web: Applying network theory to the war blogosphere. Journal of Computer‐Mediated Communication, 12(1), 290-310.

Van Boven, L., Judd, C. M., & Sherman, D. K. (2012). Political polarization projection: social projection of partisan attitude extremity and attitudinal processes. Journal of personality and social psychology, 103(1), 84.

Westfall, J., Van Boven, L., Chambers, J. R., & Judd, C. M. (2015). Perceiving Political Polarization in the United States Party Identity Strength and Attitude Extremity Exacerbate the Perceived Partisan Divide. Perspectives on Psychological Science, 10(2), 145-158.

Wojcieszak, M. (2014). Preferences for political decision-making processes and issue publics. Public Opinion Quarterly, 78(4), 917-939.

Zuiderveen Borgesius, F. J., Trilling, D., Moeller, J., Bodó, B., De Vreese, C. H., & Helberger, N. (2016). Should We Worry About Filter Bubbles?. Internet Policy Review. Journal on Internet Regulation, 5(1).

Appendix A

Codebook

Facebook pages Website articles

1. Facebook page (where the post was found)

1. Website (where the article was found. for websites: only domain, no extension, i.e. domain.com; for public platforms include base extension, i.e. youtube.com/channel) 2. Post name (first 5 words of the post, or

brief description, if not applicable, i.e. only picture)

2. Web-link of the website article (full web-link, i.e. reuters.com/article/xxx)

3. Date of post 3. Date of article

4. Length of post (words) 4. Length of article (words) 5. Number of Facebook “likes“ the post has

received

5. Number of Facebook “likes“ the article has received (if applicable)

6. Number of Facebook “shares“ the post has received

6. Number of Facebook “shares“ the article has received (if applicable)

7. Number of comments under the post

7. Number of comments under the article (if applicable)

8. Issue (topic) addressed in the post (see Appendix C)

8. Issue (topic) addressed in the article (see Appendix C)

9. Is the post anti-western?

o

o criticizing or blaming the EU o criticizing or blaming NATO

9. Is the article anti-western?

o

o criticizing or blaming the EU o criticizing or blaming NATO

o criticizing or blaming the Lithuanian government for its pro-western (pro-EU, or pro-NATO) direction, or demonizing it without substantial arguments

o demonizing the “western way of life” (capitalism, consumerism) o demonizing the pro-western Ukranian government

o pro-Russian

o

o

o criticizing or blaming the Lithuanian government for its pro-western (pro-EU, or pro-NATO) direction, or demonizing it without substantial arguments

o demonizing the “western way of life” (capitalism, consumerism) o demonizing the pro-western Ukranian government

o pro-Russian

o

o

10. Web-link provided in the post

10. Web-links provided in the

webpage (exclude internal web-links of different articles pointing towards the same website domain, BUT include web-links to different media platforms even if they may belong to the same owner, i.e. Facebook page, YouTube channel; exclude ads (best to ignore by installing an ad-blocking web-browser extension), but include friendly endorsements of other blogs/websites, i.e. "other blogs: xxx", "our friends: xxx",

"sources: xxx") 11. Is the information provided in the

web-link anti-western?

o

o criticizing or blaming the EU o criticizing or blaming NATO

o criticizing or blaming the Lithuanian government for its pro-western (pro-EU, or pro-NATO) direction, or demonizing it without substantial arguments

o demonizing the “western way of life” (capitalism, consumerism) o demonizing the pro-western Ukranian government

o pro-Russian

o

Appendix B

Table 4. Krippendorff’s alpha (α) for the inter-coder reliability test

Variable Krippendorff’s alpha (α)

Date 1,00

Length 1,00

Likes 1,00

Shares 1,00

Comments 0,9023

Issue 1 (Internal politics) 0,7008

Issue 2 (International politics) 0,8312

Issue 3 (Military and defense) 0,7771

Issue 4 (Internal order (crime)) 1,00

Issue 5 (Economy) 1,00

Issue 6 (Labor and industrial relations) 1,00 Issue 7 (Business, commerce, industry) 1,00

Issue 8 (Transportation) 1,00

Issue 9 (Health, welfare, social services) 1,00

Issue 10 (Population) 1,00 Issue 11 (Education) 0,7679 Issue 12 (Communication) 1,00 Issue 13 (Housing) 1,00 Issue 14 (Environment) 1,00 Issue 15 (Energy) 1,00

Issue 16 (Science and technology) 1,00

Issue 17 (Social relations) 1,00

Issue 18 (Sports) 1,00

Issue 20 (Fashion) 1,00

Issue 21 (Ceremonies, events) 1,00

Issue 22 (Human interest) 1,00

Issue 23 (Weather) 1,00

Issue 24 (Religion) 0,74

Tone 0,8385

Tone of the article in the web-link 0,8825

Web-link 0,8825

Appendix C

Issue (topic) list (based on the Belgian version of the comparative agenda‘s codebook, found at http://www.comparativeagendas.info/)

 Internal politics

 International politics

 Military and defense

 Internal order (crime)

 Economy

 Labor and industrial relations (unions, etc.)

 Business, commerce, industry

 Transportation

 Health, welfare, social services

 Population  Education  Communication  Housing  Environment  Energy

 Science and technology

 Social relations

 Sports

 Culture

 Fashion

 Human interest (celebrities, food, holidays, etc.)

 Weather