• No results found

Do news media produce similar products? : a content analysis of newspaper coverage in the United States, China, and Taiwan of the South China sea dispute

N/A
N/A
Protected

Academic year: 2021

Share "Do news media produce similar products? : a content analysis of newspaper coverage in the United States, China, and Taiwan of the South China sea dispute"

Copied!
50
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Master Thesis

Graduate School of Communication

Master’s Programme: Erasmus Mundus Journalism

Do News Media Produce Similar Products?

A Content Analysis of Newspaper Coverage in the United States, China, and Taiwan of the South China Sea Dispute

Qing Li

University of Amsterdam

Student number: 11081880 Supervisor: Dr. Jonas Lefevere 27th May, 2016

(2)

Abstract

News media play a major role in shaping citizens’ opinion of politics, including foreign policy. This research takes a first step in empirically applying the new institutionalism theory, which suggests a homogeneity hypothesis of news content. To test this hypothesis, a quantitative content analysis was conducted on nine US, Chinese, and Taiwanese national newspapers’ coverage on the issue of the South China Sea dispute between 2000 and 2016. Findings to a great extent confirmed the homogeneity hypothesis of news content regarding the reliance on authoritative sources and presented viewpoints. Within each of the three examined countries, national newspapers reported similar fractions of authoritative sources, and supported significantly their national government’s political stances. Across the three countries, variances in the reliance on authoritative sources and presented opinion orientation were revealed: the US newspapers reported the highest dependence on authoritative sources. Additionally, the US and Chinese newspapers held strongly opposing viewpoints on the policies of the other country. By and large, the homogeneity hypothesis of news content held true in both Western and non-Western contexts.

(3)

Introduction

International laws cannot settle all international disputes, which leaves room for national interpretation and argumentation. A good example of an international dispute is found in Asia, where over the past few decades there has been a territorial dispute between several powers over the South China Sea. The South China Sea is one of the world’s busiest shipping routes, a productive fishing waterway, and contains massive quantities of oil and natural gas resources (Hemmings, 2011). Seven Asian neighboring countries have made conflicting claims on the islands in the region: the People’s Republic of China, the Republic of China (Taiwan), Malaysia, Vietnam, Indonesia, the Philippines, and Brunei. Amongst them, China is the center of controversy because of the ambiguity of its claim and its

assertiveness in building artificial islands in the region since 2014 (Tiezzi, 2015; Ma, 2016). As rival claimants, China and Taiwan have the most aggressive and identical claims to the U-shaped line encircling most of the South China Sea (Tiezzi, 2015). Apart from the seven Asian countries, the United States (US) is also deeply involved in the dispute. The US has no territorial claims over the South China Sea, but has strong political and economic interests in the region (Ma, 2016). The dynamics in the development of the dispute have substantial impact on the US foreign policy in the region (Hemmings, 2011).

With different nations holding contrasting claims, it appears that news media within each nation are following suit. China tends to be associated with the “China threat” (Hong, 2013, p. 27) and “the most dangerous source of instability in Asia” (Dillion, 2011, p. 51) by several non-Chinese media. Meanwhile, China is portrayed as a constructive force in the region by the Chinese national media. Although in general public attention to international affairs is very low, mass media produce the “default reality option” (Bennett et al., 2007, p. 56) to a large audience and influence the relationship between public opinion and

(4)

the media, and journalism in particular: As international affairs are complex matters, do national mass media take a position in these issues? Are media truly a free “marketplace of ideas” (Ingber, 1984, p. 2)? Or do they produce rather identical news content?

To understand how news media cover the South China Sea dispute, this research adopts the new institutionalism approach, which argues that news outlets within a given system are pushed towards similar practices. The theory predicts a certain level of homogeneity on the final product of the news making process, namely the news content (Cook, 1998; Ryfe, 2006; Esser & Umbricht, 2013). Whereas extent work in the new institutionalism literature mainly focused on Western democracies (e.g. Cook, 1998;

Sparrow, 1999; Entman, 2006), this research takes a first step in empirically applying the new institutionalism theory in a comparative research design, including both Western and Asian countries. The South China Sea dispute involves countries with different political systems and media cultures. The dispute offers a notable opportunity to analyze different journalistic practices and to contrast perspectives from a typical Western democracy (the US), a non-Western authoritarian country (China), and a non-non-Western democracy (Taiwan). The

following research question is addressed in this research: To what extent does homogeneity of news content exist in the US, Chinese, and Taiwanese newspapers’ coverage of the South China Sea dispute both within and across countries?

This thesis is structured as follows. In the next section, it briefly presents the

literature on new institutionalism, and develops expectations on news content both within and across countries. Then, it presents the content analysis measurement, before turning to the empirical analysis. Lastly, it discusses the findings, derives implications, and suggests directions for future research.

(5)

Theoretical Framework New Institutionalism

A number of different explanations have been offered to account for news content. Sociologists and communication scholars, who have been deeply influenced by the

sociological tradition, try to understand how journalists’ efforts are constrained by the

organizational and occupational demands. The basic orientation is that political news making is a reality-constructing process that follows the lead of government officials (Cook, 1998; Schudson, 2002; Sparrow, 2006). In the field of political communication, the new

institutionalism approach situates its analysis at a macro-social level and focuses on societal or cultural influences on news organizations (DiMaggio & Powell, 1983; Kaplan, 2006; Baum & Potter, 2008). It argues that news media are not simply distinct organizations, but make up a collective political institution (Cook, 1998; Sparrow, 1999; Ryfe, 2006). Cook (1998), who applied these neo-institutional insights to political communication, considers the core of institutionalism as a denial that all social phenomena can be reduced to individual psychology. Cook (1998) further defines institutions as “social patterns of behavior

identifiable across the organizations that are generally seen within a society to preside over a particular social sphere” (p. 70).

New institutionalism argues that there are numerous reasons why news media are recognized as a political institution. Within a similar political and regulatory environment, news media share identical outside politics, uncertainty, and professional values (DiMaggio & Powell, 1983). Sparrow (2006) identifies three kinds of uncertainties in the environment: whether or how to make a profit, how to establish legitimacy, and how to timely find

information. Since news organizations within a country operate in the same uncertain political and economic environment, they have developed various standard routines

(6)

of authoritative sources, including politicians, government officials, corporate spokespersons, and academic experts (Sparrow, 1999). Additionally, news media are a political institution due to the politics of communication. They converge on similar sources to benefit from the governmental information subsidies, which give them all a similar reliance on political power (Cook, 1998; Schudson, 2002; Bennett et al., 2007).

Scholars (e.g. Cook, 1998; Bennett et al., 2007) even go as far as to argue that news media have become part of the government, because journalists have embraced the part of disseminating governmental information to the public. As the fourth branch of the

government, news media are also partially dependent on other institutions to accomplish their own tasks, and thus share a similar fate with the three other constitutional branches (Wirt & Cater, 1960; Cook, 1998). Given these perspectives, new institutionalists claim that the news making process is not essentially influenced by individual journalists, but rather by routines, procedures, and rules (Cook, 1998). As a result, individual news practitioners are pushed towards similar news products.

Building on the above described arguments, the new institutionalists bring forward a homogeneity hypothesis, which expects identical similarities in the news production

processes within a given society and similar news content across various news outlets (Cook, 1998; Esser & Umbricht, 2013; Ryfe, 2006; Entman, 2006). However, several scholars (e.g. Shoemaker & Reese, 1996; Schudson, 2002) state that the influence at the individual level on the news content cannot be fully neglected. These authors argue that the construction of news also depends on the individual characteristics of news practitioners, as well as on the external political, cultural, and economic factors. Therefore, news can be homogenous and

remarkably diverse (Ryfe, 2006).

Much of the subsequent work in the academic field assumes the existence of homogeneity without empirically validating this assumption. The few empirical studies

(7)

utilizing the new institutionalism approach have only been conducted in a Western context (Ryfe, 2006; Cook, 2006; Benson, 2006). In an analysis of the US newspapers’ coverage of prisoner abuse in Iraq, Entman (2006) relates the new institutionalism theory to the framing theory, and identifies a contrast between homogeneous neglect of civilian death as opposed to the more varied treatment of prisoner abuse.

This thesis contributes to the academic literature by empirically testing the

homogeneity hypothesis on the national news coverage in both Western and non-Western contexts. This research is inspired by Cook’s (2006) approach and theory, which expects identical similarities over the range of sources, issues, and points of view in the news content. The homogeneity of national news is the baseline prediction of the new institutionalism claim (Entman, 2006). To test this hypothesis, this research focuses on the issue of the South China Sea dispute. Based on the new institutionalism theory, two sub-questions are raised:

(1) To what extent does homogeneity of sources exist in the US, Chinese, and Taiwanese national newspapers’ coverage of the South China Sea dispute?

(2) To what extent do national newspapers in the US, China, and Taiwan present homogeneous points of view on the South China Sea dispute?

Homogeneity Hypothesis of News Content

A central assertion of the new institutionalists’ approaches to news is that the news outlets, even across modalities (e.g. newspapers, televisions, radios), tend to focus on certain political actors for particular reasons with particular stories in mind (Cook, 2006; Entman, 2006). As the societal influences are constant in a single society, a certain level of

homogeneity of news content is expected within a country (Zhu et al., 1997; Cook, 1998; Sparrow, 1999; Ryfe, 2006). This research further investigates the homogeneity hypothesis on the use of sources and on the presented points of view in the news coverage of the South China Sea dispute.

(8)

Source.

Journalists need sources to get access to information, to provide different viewpoints, to offer contexts and interpretations, and to validate news accounts (Sigal, 1973; Tuchman, 1978). These sources can come from different social spheres, and the role of particular source categories can vary. Therefore, “the journalistic need for news sources is a constant, but the usage of news sources is a variable” (Dimitrova & Strömbäck, 2009, p. 75). One hypothesis of the new institutionalism theory is that journalists rely heavily on authoritative sources, which are the “societal indicators of achievement” (Sparrow, 1999, p. 126). The definition of authoritative sources varies in the new institutionalism literature. Cook (1998, p. 111) suggests a “gravitation toward officialdom” among news media, which refers to government officials having greater access to obtain coverage that is closer to their viewpoints, especially regarding international affairs. Inspired by Sparrow (1999), this research takes a broader approach to authoritative sources, which include politicians, political candidates, public officials, as well as academics, and other powerful actors.

Indexing theory offers additional support to the homogeneity hypothesis regarding the use of authoritative sources. Journalism is telling the story of power, as journalists calibrate the news based on the dynamic power sphere. Bennett et al. (2007) describes the process as the “mechanics of opening and closing the news gates” (p. 49). The news media create a weighing system to determine what gets into the news, who gets a voice, and what

prominence a source receives. The implicit weighing process is based on the positioning of powers that are in play, which means that the news gates open wider for the most powerful players. This regularized process has been called indexing by Bennett (1990), and the indexing theory offers an empirical framework of how news content is constructed

(Lawrence, 2006). The main assertion in the indexing theory is that mass media tend to index the range of voices and viewpoints according to the range of opinions expressed in

(9)

mainstream relations. Bennett (1990) tests the assertion with the media coverage of the US policymaking on Nicaragua in the mid-1980s, and measures the frequency, direction, and source of all opinions voiced in the coverage. Similar to the homogeneity expectations put forth by new institutionalism, the results show that opinions voiced in news stories originate overwhelmingly from government officials. Studies have largely demonstrated that US media are most likely dependent upon official sources, for instance, in covering foreign policy issues (e.g. Hallin, 1986; Brown et al., 1987; Zaller & Chiu, 1996).

Given the literature in new institutionalism and the indexing theory, news media rely to a greater extent on authoritative sources than on non-authoritative sources (e.g. ordinary citizens, social society representatives). In this case, all national newspapers from the US, China, and Taiwan are expected to show a substantial reliance on authoritative sources. Based on the homogeneity hypothesis, the news outlets within a country share similar routines, procedures, and rules. Consequently, the newspapers are expected to have the similar ratio of authoritative sources to non-authoritative sources within the US, China, and Taiwan, as they should rely on authoritative sources to the same degree. The following hypotheses are formulated regarding the use of authoritative sources in national newspapers in a given country:

H1a: The ratio of authoritative sources to non-authoritative sources is the same across national newspapers within the US.

H1b: The ratio of authoritative sources to non-authoritative sources is the same across national newspapers within China.

H1c: The ratio of authoritative sources to non-authoritative sources is the same across national newspapers within Taiwan.

From a cross-national perspective, the US, China, and Taiwan are fundamentally different in their political systems, societal cultures, and so forth. Given the fact that the

(10)

societal influences on news media vary across the three countries, the US, Chinese, and Taiwanese media are expected to have different routines, procedures, and rules, and thus varied ratios of authoritative sources to non-authoritative sources in the news content.

Another expectation derives from the cross-national comparison between the US and Chinese media. It is widely known that journalists in China are “dancing with chains on” (Zhao, 1998, p. 121). The Chinese national newspapers are state-controlled, and act as

“self-conscious organs of propaganda” for the government (Schudson, 2002, p. 250), in contrast to the US national newspapers, which are formally independent from the state. The Taiwanese newspapers had been heavily regulated by the government until the martial law was lifted in 1988 (Luo, 2008), and thus situate somewhere in the middle between the US and Chinese media regarding the freedom of press. Therefore, it is rational to expect that Chinese news media are more heavily dependent on authoritative sources, compared to news media in the US. Two cross-national expectations are made:

H1d: The US, Chinese, and Taiwanese national newspapers have different fractions of authoritative sources.

H1e: There is a greater fraction of authoritative sources in the Chinese national newspapers compared to the US national newspapers.

Points of view.

The homogeneity hypothesis also investigates the opinion orientation of news media both within and across countries. The new institutionalism literature argues that the national news media do not reflect a wide range of diverse political realities because they benefit from the governmental information, which gives media all a similar reliance on political power (Cook, 1998; Sparrow, 1999). As the authoritative actors talk with the loudest and

sometimes even the only voice in international affairs stories, the institutionalized press also speaks with a singular voice, and in effect, “the voice of government itself” (Bennett et al.,

(11)

2007). The news media’s reliance on authoritative sources leads them to essentially parrot the information obtained from authorities (Cohen, 1963). Subsequently, the new

institutionalism theory does not expect much diversity in the presented viewpoints in news content. Zaller and Chiu (2000) confirm that US media repeatedly rally along with the public when conflicts arise as a result of the US media’s reliance on authoritative sources for

information. Other scholars argue that dependence on authoritative sources, for instance government officials, does not guarantee pro-government news (e.g. Waisbord, 2000; Schudson, 2002). However, it is widely believed that most news discourses originate from sources, and the selection of these sources greatly influences the presented opinion

orientation of the news content (Hagen, 1993; Bennett et al., 2007). For instance, Hagen (1993) identifies that journalist’s opinion orientation is in accordance with the general viewpoints of most of the news sources.

In sum, authoritative sources, being part of the elite, would tend to favor their national government’s interests. In the South China Sea dispute, the US, Chinese, and Taiwanese governments stand for their own interests and benefits in their political claims. The national media within the three countries are thus expected to rely heavily on national authoritative sources, and consequently, to present more positive viewpoints of their national government than negative opinion orientations. The fraction of positive points of view is then expected to be identical across newspapers within a country. The following hypotheses on the opinion orientation of the news content within a country are formulated:

H2a: The fraction of news articles with a positive point of view on the US is similar across US national newspapers.

H2b: The fraction of news articles with a positive point on view of China is similar across Chinese national newspapers.

(12)

H2c: The fraction of news articles with a positive point on view of Taiwan is similar across Taiwanese national newspapers.

From a cross-country comparative perspective, the US, Chinese, and Taiwanese authorities hold opposing political claims in the South China Sea dispute, and possess different opinion orientations of each other. Regarding the relationship between the US and China in this issue, the two countries are potential competitors for leadership and dominance in the South China Sea. Since the early years of the Second World War, the US has exercised its leading influence in this region (Mearsheimer, 2010). However, the period of US

leadership and dominance is seen to come to a close, as China has emerged as a serious challenger to the status quo, which has brought instability and a bipolar balance of power in the region since the 1990s (Cha, 2010; Burgess, 2016). Regarding China and Taiwan, these two countries have competing claims of the islands in the South China Sea. Even though China has approached Taiwan for potential collaborations on this matter, Taiwan has

declined to cooperate with China on any territorial issue (Tiezzi, 2015). Meanwhile, Taiwan has remained an unofficial but close relationship with the US over the past four decades (Liao & Lin, 2015).

The new institutionalists would argue that news media tend to follow their national government’s positions in the dispute, leading the US newspapers to be critical of China, and the Chinese newspapers critical of the US. As in the US, Chinese, and Taiwanese

governments possess different claims in the South China Sea dispute, the national media would tend to parrot their own national authorities’ political stances. Consequently, the national media within the three countries will present different opinion orientations towards the US, China, and Taiwan. Thus, this thesis expects that

H2d: The US national newspapers have a similar fraction of news articles with negative viewpoints towards China.

(13)

H2e: The Chinese national newspapers have a similar fraction of news articles with negative viewpoints towards the US.

H2f: The US, Chinese and Taiwanese newspapers present different fractions of points of view towards the US.

H2g: The US, Chinese and Taiwanese newspapers present different fractions of points of view towards China.

H2h: The US, Chinese and Taiwanese newspapers present different fractions of points of view towards Taiwan.

Methods

A quantitative content analysis was employed to investigate the homogeneity hypothesis of news content within a country and across countries. This thesis examined the national broadsheet newspapers’ coverage of the South China Sea dispute, by analyzing the use of authoritative sources and the presented viewpoints in the news content. A quantitative content analysis was chosen because it enabled a systematical evaluation of media content with less dependence on subjective perceptions, compared to several qualitative approaches (Kolbe & Burnett, 1991; Krippendorff, 2004). Three countries were chosen, namely the US, China, and Taiwan. The three selected countries consist of a notable mix of different political systems and media landscapes, including a typical Western democracy (US), a non-Western authoritarian country (China) and a non-Western democracy (Taiwan). Moreover, they are to a different degree involved in the dispute and possess distinct political positions, which is expected to result in multiple and different perspectives on the issue. The content analysis was carried out on a sample of articles from the US, Chinese, and Taiwanese national newspapers.

(14)

Selection of Newspapers

For each country, three national daily broadsheet newspapers were selected. The three US newspapers include the New York Times, USA Today, and Washington Post, which were chosen based on the circulation and distribution (Cision, 2014). The three Chinese newspapers included the People’s Daily, Guangming Daily, and Xinhua Daily, which are state-owned, nationally distributed, and highly representative of the Chinese press landscape (Hassid, 2012). The three Taiwanese newspapers included the China Times, United Daily News, and Liberty Times, which are the three most influential broadsheet newspapers in Taiwan (Lee, 2007; Lin, 2008; Zhang & Chen, 2015). In sum, these nine national dailies are opinion-leading quality newspapers within the national press landscape in all three countries. Period of Study

The content analysis was conducted on articles published between January 1st, 2000 and April 10th, 2016. The time period was decided based on the availabilities of all the chosen newspapers archiving platforms across the three countries. Apart from Liberty Times, only available since 2005, the other eight newspapers are available since 2000. Since the South China Sea dispute has been going on for decades, this research tried to take the

maximum available period into account. New institutionalism theory expects the existence of homogeneity in the news content to endure over time (Cook, 1998), thus this research took a longitudinal approach and sampled over 16 years. The year of publication factor was taken into account in the assessment of the varied fractions of authoritative sources, which will be further explained in the results section.

Data Collection

All the US newspaper articles were collected through LexisNexis. All Chinese newspaper articles were collected through CNKI, a Chinese information platform with the highest reputation for archiving newspapers. The articles from the three Taiwanese

(15)

newspapers were derived from three different platforms: China Times articles were selected through KMW (知識贏家), which provides all newspaper articles from the China Times News Group; United Daily News articles were derived through United Daily Media Group Dataset (聯合知識庫); Liberty Times articles were derived through its own online dataset. All the listed datasets offered a complete collection of news articles from all nine chosen

newspapers. Sampling

A stratified sampling method was applied to ensure a representative sample of articles published from 2000 to 2016, and to ensure an equal chance of any article being selected. In the first step, for each newspaper a pool of articles was collected based on a specific

keywords scheme (see Appendix A), which contained the most relevant terms and phrases in connection to the South China Sea dispute. For the newspapers that published over 500 articles on the South China Sea dispute within the given time period (e.g. the New York Times and United Daily News), every 10th article was selected. For the others newspapers, all the relevant articles were collected in the first step. Afterwards, within each pool of articles, a random sampling was applied to choose 40 articles for each newspaper to code, which resulted in a total sample of 360 articles. A random string of numbers was generated,

following which articles were selected and coded. In addition, an article was only coded if it satisfied the following conditions: (1) the South China Sea dispute was the main issue in the article; (2) it referred explicitly to at least one of the three nations, namely the US, China, and Taiwan; (3) at least one source was mentioned.

Coding

The unit of analysis and coding unit were at the source level in the news articles, which means that each source in the articles was identified and coded. The coding of the sampled articles was conducted according to a 11-item codebook (see Appendix B for the

(16)

codebook), and all the variables were coded with numeric values. To assess the reliability of the coding process, an intercoder reliability test was carried out. Using the Krippendorff’s alpha, the intercoder reliability between two trained coders was assessed on a randomly selected 30-article sample (approximately 10% of the main sample) in a pilot test (see a list of variables in Appendix C). The alpha scores were in an acceptable range (from 0.70 to 1.00), except for the variable “sourced message’s points of view towards Taiwan” with an alpha score of 0.66. As this variable had little variance, one disagreement between two coders dropped the score from 1.00 to 0.66. The codebook was improved after the reliability test. Afterwards, the actual coding was conducted by one trained coder based on the modified codebook. In general, the intercoder reliability assessment demonstrated a satisfying level of consistency in coding decisions. This study is thus solidified by the verification of the intercoder reliability test (Krippendorff, 2004).

Measures of Variables.

The new institutionalism theory provided a conceptual definition of news media as a political institution. However, to identify the homogeneity of news content, an operational definition is needed as well. Regarding the operationalization of homogeneity, there is no agreed practice (Entman, 2006; Ryfe, 2006). This research was inspired by the theoretical approach of Cook (2006), which suggests examining the range of sources, issues, and points of view, in order to test the homogeneity of the news content. Because the scope of this study was limited to the issue of the South China Sea dispute, issue as a variable was not examined and the study confined its focus to the range of sources and the points of view.

Authoritative sources category tracked whether a source belongs to the authoritative category (1) or non-authoritative category (0). In this study, news sources were considered as actors, including both individuals and groups (e.g. government officials, national newspapers, organizations), whose voices or opinions were mentioned in the news content. In order for

(17)

an actor to be considered as a source, one statement, fact or quote, must be attributed to them (Dimitrova & Strömbäck, 2009). The operationalisation of authoritative sources was inspired by Sparrow (1999, p. 126) and included politicians, political candidates, public officials, academics, state organizations, etc. In this context, military forces were also regarded as authoritative sources. In addition, the Chinese news organizations were seen as authoritative sources, because they operate in an authoritarian media landscape and their coverage

explicitly or implicitly reflects the opinion orientations of the Chinese authorities. In sum, authoritative sources included government officials, military members, experts, state organizations, and the Chinese news organizations. Non-authoritative sources included ordinary citizens, non-state organization, and others1.

Points of view referred to the opinion orientation of an article or a sourced message, which were measured using a three-point scale (supportive = 1; neutral or mixed = 0; against or skeptical = -1). The articles’ points of view towards the US, China, and Taiwan were coded. An article was coded as “supportive” towards a country when the argument was in favour of the nation’s position or interest. An article was coded as “neutral or mixed” towards a country when the argument did not declare a clear opinion orientation towards the country, or held a mixed point of view. An article was coded as “against or skeptical” towards a country when the argument showed disagreement or skepticism towards a country’s position or interest.

1Government officials refer to officials elected or appointed to administrate a government

(e.g. politicians, public officers). Experts refer to people who have special skills or

knowledge (e.g. academics, lawyers, analysts). Military members include people who work for the armed force of a country. News organizations include newspapers, magazines, radio, television broadcasters, etc. Citizens include civil society representatives, business persons, ordinary citizens, etc. State organizations refer to the organizations or institutions that belong to the government, or are owned by the government. Non-state organizations include NGOs, multinational corporations, religious groups, etc. Others include unclassifiable

(18)

The main topic of the articles was also coded. Six major topics were identified on the issue of the South China Sea dispute: (1) territorial claims; (2) claims to the natural resources (including the oil and natural gas reserves, fish, etc.); (3) military presence (including the maintenance of freedom of navigation, military practices, etc.); (4) piracy and sea robbery; (5) regional management (including the regional disaster management, coast guards’ work, etc.); and (6) other economic development (including the pan-regional economic

development plans, etc.) (Rosenberg, 2002).

Additionally, other variables, such as the date of publication, type of article (1= news article; 2= feature article; 3= opinion piece), nationality of source were also coded with numeric values.

Results

To give an overview of the media coverage on the South China Sea dispute of the nine selected newspapers, table 1 presents the descriptive statistics for the type of sources, and the topics that were covered. Although all articles dealt with the South China Sea dispute, various subtopics got varying degrees of attention in the articles: the most dominant topic within all the articles from all nine newspapers was countries’ general “territorial claims” (51.1%, N=184). 27.8% (N=100) of the sample articles dealt with “military presence” and the topic “natural resources” encompassed 11.4% (N=41) of the whole sample.

(19)

Table 1

Descriptive Results of Sources and Topics Across Countries

National Newspapers US China Taiwan Type of source Government officials* 49.6% 39.9% 39.7% Military members* 13.6% 8.0% 9.9% Experts* 16.9% 13.6% 12.0% Citizens 4.9% 13.9% 7.9% State organizations* 7.8% 4.7% 12.2% Non-state organizations 2.4% 2.1% 2.5% News organizations** 5.0% 14.8% 15.5% Others 0.0% 3.1% 0.3% Total 100.0% 100.0% 100.0% N 619 426 393 Location of source Domestic sources 54.0% 44.6% 23.5% Foreign sources 46.0% 55.4% 76.5% Total 100.0% 100.0% 100.0% N 619 426 393 Topics Territorial claims 39.1% 57.3% 52.9% Natural resources 4.9% 18.6% 9.4% Military presence 53.6% 8.7% 27.7%

Piracy and sea robbery 1.8% 2.3% 1.0%

Region management 0.3% 13.2% 7.4%

Economic development 0.3% 0.0% 1.5%

Total 100.0% 100.0% 100.0%

N 619 426 393

Note. N=1,438. The unit of analysis is the source. An equal number of articles (n=40) were

coded for all nine newspapers, and multiple sources generally feature in an article. Source type marked with * are categorized as authoritative sources.

** Regarding the news organizations, only the Chinese news organizations are regarded as authoritative sources.

(20)

As shown in table 1, 1,438 sources were coded in all examined articles. Government officials were the most widely cited sources. Experts and military members were also commonly cited. Regarding the number of sources cited, the US newspapers had the highest number of sources (N=619) compared to the Chinese (N=426) and Taiwanese newspapers (N=393). The nationalities of sources were largely American (34.1%, N=470), Chinese (31.0%, N=427), Filipino (12.6%, N=174), and Taiwanese (6.8%, N=93) in the sample. In terms of the geographic locations of the sources, there was roughly an equal distribution between domestic sources and foreign sources in the US and Chinese newspapers. Taiwanese newspapers relied to a greater extent on foreign sources (N=290) than on domestic sources (N=89).

The first research question of this study concerned the ratio of authoritative to non-authoritative sources used in the newspapers within a country and across countries. The expectation was that all newspapers in a given country would use more authoritative than non-authoritative sources to the same extent, but that the ratios differ significantly between newspapers of different countries. As summarized in table 2, independent t-tests were conducted to test if the ratios are significantly (p<0.01) different across groups.

(21)

Table 2

Independent T-tests of Authoritative Sources Across Newspapers and Countries

H1 Authoritative Sources*

(N=1,424)

M SD

Within the US (H1a)

New York Times 0.92a 0.27

USA Today 0.91a 0.29 Washington Post 0.92a 0.27 Within China (H1b) People's Daily 0.66a 0.48 Guangming Daily 0.73a 0.44 Xinhua Daily 0.67a 0.48 Within Taiwan (H1c) China Times 0.82a 0.39

United Daily News 0.82a 0.39

Liberty Times 0.75a 0.43

Across countries (H1d & H1e)

US 0.92a 0.28

China 0.69b 0.46

Taiwan 0.80c 0.40

Note. M = Mean. SD = Standard Deviation. n/a= not available due to a limited number of

observations.

The unit of analysis is the source. An equal number of articles (n=40) was coded for all nine newspapers, and multiple sources generally feature in an article.

An alpha level of .01 is used for all the t-tests. Within each country, if the means of the two samples have the same subscripts (a, b, c), it indicates that these two groups do not differ significantly in their means. If the means of the two samples have different subscripts (a, b, c), these two groups report significant difference between their means.

(22)

In this research, three hypotheses were formulated at the newspaper level within the three countries. H1a predicted that the ratio of authoritative to non-authoritative sources should be the same across the three US newspapers. As shown in table 2, the findings regarding the first three lines show that the three US newspapers placed an overwhelming reliance on authoritative sources (M=.92, SD=.28). There was no significant difference among the three newspapers. H1a is thus strongly supported. H1b predicted that the ratio of authoritative to non-authoritative sources should be the same across the three Chinese newspapers. The findings suggest that the Chinese newspapers relied more heavily on authoritative sources than non-authoritative sources (M=.69, SD=.46). There were no significant differences between the ratios across the three Chinese newspapers. H1b is also supported. Finally, H1c predicted that the ratio of authoritative to non-authoritative sources was the same across the three Taiwanese newspapers. H1c is supported because Taiwanese newspapers reported a greater reliance on authoritative sources (M=.80, SD=.40), and no significant difference was found across the three Taiwanese newspapers. To conclude, the three hypotheses on the fraction of authoritative sources at the newspaper level are supported.

Secondly, from a cross-country perspective, H1d predicted that the US, Chinese, and Taiwanese national newspapers would report different fractions of authoritative sources. As shown in table 1, the US newspapers reported the highest reliance on authoritative sources (M=.92, SD=.28), whereas the Taiwanese (M=.80, SD=.40) and Chinese newspapers (M=.69, SD=.46) showed slightly less reliance on authoritative sources. Indeed, significant

differences among the ratios of these three countries were found. Therefore, H1d is

supported because all newspapers from the three countries used more authoritative sources, and the newspapers further demonstrated significant differences across countries. Although the Chinese news organizations were considered to be authoritative sources, it is important to mention that even when the Chinese news organizations were not considered as authoritative

(23)

sources, there were still significant differences (p<.01) in the fractions of authoritative sources between the Chinese and US newspapers, and the Taiwanese and US newspapers. However, there was no significant difference (p<.01) between the Chinese and Taiwanese newspapers. That said, H1d receives overall support. Finally, H1e predicted that there would be a greater fraction of authoritative sources in the Chinese newspapers compared to the US newspapers. The findings show that the fraction of authoritative sources in the US

newspapers was higher compared to in the Chinese newspapers. Therefore, H1e is rejected. One explanation is that the US newspapers have less access to non-authoritative sources in foreign policy issues, which leaves considerable room to officials to define the matter (Bennett et al., 2007). This will be further explained in the conclusion.

The second research question concerned the presented opinion orientation in all the newspapers’ coverage within a country and across countries. Five hypotheses at the

newspaper level and three hypotheses at the country level were formulated. The columns of table 3 present the mean opinion orientation towards each of the three countries, split per newspaper (rows).

(24)

Table 3

Independent T-tests of Viewpoints Across Newspapers and Countries

H2 Viewpoints* towards

US (N=195) China (N=342) Taiwan(N=117)

M SD M SD M SD

Within the US (H2a & H2d)

New York Times 0.67a 0.48 -0.65a 0.48 0.00a 0.58

USA Today 0.41a 0.55 -0.48a 0.51 0.22a 0.43

Washington Post 0.68a 0.47 -0.68a 0.53 0.36a 0.50

Within China (H2b & H2e)

People's Daily -0.79a 0.43 0.82a 0.46 n/a n/a

Guangming Daily -0.78a 0.44 0.93a 0.35 n/a n/a

Xinhua Daily -0.57a 0.79 0.98a 0.16 n/a n/a

Within Taiwan (H2c)

China Times -0.04a 0.60 -0.13a 0.66 0.72a 0.46

United Daily News 0.06a 0.54 -0.23a 0.54 0.79a 0.41

Liberty Times 0.44a 0.72 -0.67b 0.86 0.82a 0.51

Across countries (H2f, H2g, H2h)

US 0.58a 0.51 -0.60a 0.51 0.20a 0.50

China -0.73b 0.52 0.91b 0.35 n/a* n/a*

Taiwan 0.12c 0.64 -0.31c 0.61 0.79b 0.48

Note. M=Mean. SD=Standard Deviation. n/a=not available due to a limited number of observations. n/a* only two observations were made from Chinese newspapers.

The unit of analysis is the source. An equal number of articles (n=40) were coded for all nine newspapers, and multiple sources generally feature in an article.

Viewpoints*: A positive mean value indicates a positive viewpoint of a certain country, and a negative mean value indicates to a negative viewpoint of a certain country.

An alpha level of .01 is used for all the t-tests. Within each country, if the means of the two samples have the same subscripts (a, b, c), it indicates that these two groups do not differ significantly in their means. If the means of the two samples have different subscripts (a, b, c), these two groups report significant difference between their means.

(25)

Firstly, at the newspaper level, H2a predicted that the fraction of news articles with a positive viewpoint on the US were similar across the three US newspapers. As shown in table 3, a positive mean value refers to a positive opinion orientation towards a certain country, and a negative mean value refers to a negative opinion orientation of a certain country. Although USA Today (M=.48, SD=.51) had a lower mean value compared to the other two newspapers, there was no significant difference in the range of supportive

viewpoints of the US across the three US groups. More importantly, all the US newspapers presented rather positive opinion orientations towards the US (M=.61, SD=.50). Therefore, H2a is supported. H2b predicted that the fraction of news articles with a positive point of view on China would be the same across the three Chinese newspapers. Findings show that the three Chinese newspapers articles presented substantial positive viewpoints towards China (M=.82, SD=.40). Even though Guangming Daily reported a higher value (M=.98, SD=.16) than the other two Chinese newspapers, there was no significant difference across the three Chinese newspapers. H2b is supported. H2cpredicted that the fraction of news articles with a positive point of view towards Taiwan were the same across the three

Taiwanese newspapers. The three Taiwanese newspaper presented a highly positive opinion orientation of Taiwan (M=.82, SD=.43). Moreover, the Taiwanese newspapers reported no significant difference among the three groups. Therefore, H2c is strongly supported.

Secondly, as China and the US have been competing in the South China Sea region, two hypotheses were formulated. H2d predicted that all the US newspapers would report the same fraction of a negative point of view on China. This hypothesis was supported by the findings that the three US newspapers presented strongly negative opinions of China (M=.-63, SD=.49), with the Washington Post reporting the highest negativity. H2d is thus supported. H2e predicted that all the Chinese national newspapers would have a similar fraction of news articles with a negative point of view on the US. Findings show that the

(26)

articles from the three Chinese newspapers were largely negative towards the US (M=.-83, SD=.42). The findings reported no significant difference in their viewpoints towards the US. H2e is thus strongly supported.

Thirdly, three hypotheses were formulated at the country level. H2fexpected that the US, Chinese, and Taiwanese newspapers would present different fractions of points of view towards the US. The findings show that the US newspapers’ coverage were modestly

supportive towards the US (M=.58, SD=.51), while the Taiwanese newspapers’ coverage was slightly supportive (M=.12, SD=.64) and the Chinese newspapers’ coverage was heavily skeptical towards the US (M=-.73, SD=.52). As the last three rows for the US column in table 3 indicate, the three groups reported to significantly differ from each other. H2fis strongly supported. H2g predicted that the US, Chinese, and Taiwanese newspapers would present different fractions of points of view towards China. The Chinese newspapers’ coverage was substantially supportive towards China (M=.58, SD=.51), while the US (M=-.60, SD=.51) and Taiwanese newspapers’ coverage (M=-.31, SD=.61) were rather negative. In addition, the three groups reported significant differences among each other, and therefore, H2g is strongly supported. Lastly, H2h predicted that the US, Chinese, and Taiwanese newspapers would present different fractions of points of view towards Taiwan. The Chinese media coverage in the sample only had two observations to assess the Chinese articles’ viewpoints towards Taiwan, because the Chinese newspapers did not mention Taiwan altogether. Therefore, the Chinese and Taiwanese newspaper comparison was eliminated from the analysis. Meanwhile, the US and Taiwanese newspapers were both supportive and the Taiwanese newspapers were significantly more supportive (M=.79, SD=.48) towards Taiwan compared to US newspapers (M=.20, SD=.50). H2his partially supported. To

conclude, the hypotheses on the presented viewpoints in the news content at the country level largely held true.

(27)

It is important to mention that, when the p-value changed from (<.01) to a more stringent level (<.001), it did not result in any change in the above results. The conclusions on the use of authoritative sources and presented viewpoints in the news content remained identical both within and across countries.

(28)

Table 4

Logistic Regression of the Ratios of Authoritative Sources Across the Three Countries

Independent Variable Coeff. SE Country (reference: US) China -1.31*** .20 Taiwan -1.08*** .21 Type of article (reference: News) Feature -.61* .24 Editorial .46 .20

Character of sourced message (reference: Quoted)

Paraphrased .17 .21

Both .56* .24

Topic

(reference: Territorial Claims)

Natural resources -.30 .22

Military presence .57** .22

Piracy -.90 .46

Region management -.33 .26

Other economic development n/a n/a

Year -.04 .02

Pseudo R2 = .10 N = 1,409

Note. * refers to the reference category in each independent variable. * = p < .05, ** = p < .01, *** = p < .001

Coeff. = Coefficiency. SE = Standard Deviation. n/a = not available due to a limited number of observations.

Table 5

Marginal Effects for the Probability of the Fraction of Authoritative Sources

Margin SE 95% Conf. Interval

Country

US .92*** .01 .89 .94

China .75*** .02 .70 .80

Taiwan .79*** .02 .75 .83

Note. N=1,411. SE = Standard Error. * refers to the reference category in each independent variable. * = p < .05, ** = p < .01, *** = p < .001

(29)

Based on the findings, H1e was rejected, which deserves further explanation. Why do the US newspapers have a heavier reliance on authoritative sources compared to the

authoritarian Chinese newspapers? Is there any other factor playing a vital role? The literature of new institutionalism and indexing theory do not have a strong topical focus. However, in the South China Sea dispute, there are many sub-topics reported in the media coverage (as shown in table 1). For instance, there is reason to believe that the ratio of authoritarian sources is higher in military presence stories – as not many sources have access to relevant information - compared to, for example, economic focused reports, which allow media to bring in experts. To control for the impact of several factors, such as topics and types of articles, a multivariate logistic analysis was conducted, predicting whether a source is authoritative (1) or non-authoritative (0).

As presented in table 4 and table 5, the findings of the multivariate logistic analysis present mixed results on the fractions of authoritative sources across the three selected countries. Between the US and the newspapers from the other two selected countries, even when controlling for the types of articles, the ways in which sources are used, and the topics of the articles, the US newspapers reported significantly more authoritative sources. Keeping the control variables at the mean value, the predictions of the model indicated that the

probability that the US newspapers use authoritative sources is 92%, which is 16.7% higher than the Chinese newspapers and 12.7% higher than the Taiwanese newspapers. Thus, even after other influences were controlled for, the ratios of authoritative sources to

non-authoritative sources were still significantly different between the US newspapers and Chinese newspapers. The results thus demonstrate the robustness of the finding that the US newspapers use more authoritative sources than the Chinese and Taiwanese newspapers. However, between Chinese and Taiwanese newspapers, the predicted probability that a source is authoritative did not differ significantly between the Chinese and Taiwanese

(30)

newspapers (as shown in table 5). The mixed results indicate that H1c is partially rejected. One explanatory factor is the characteristic of the issue: the South China Sea dispute is a domestic issue to both the Chinese and Taiwanese newspaper, but a foreign issue to the US newspapers. This will be further explained in the conclusion.

Regarding the controlled variables, the topic indeed suggests that military stories have more authoritative sources than other topics. The findings also suggest that feature stories use less authoritative sources than news stories and opinion pieces. The duration of time does not have significant impact on the fraction of authoritative sources. In sum, this research investigated the homogeneity hypotheses of the ratio of authoritative sources in the news articles and the hypotheses on the differences in the points of view in the news content. Most of the hypotheses were confirmed, that is to say, there is a high level of homogeneity in the use of authoritative sources and opinion orientation in the news articles across newspapers within a country, while there are significant differences across countries.

Conclusion and Discussion

This research tested the homogeneity hypothesis of news content, which originates from the new institutionalism theory. The homogeneity hypothesis expects a similar fraction of authoritative sources and an identical balance of the presented viewpoints in media

coverage across news outlets within a country. Scholars’ interest in new institutionalism and indexing theory has so far been largely focused on the Western democracies (e.g. Cook, 1998; Sparrow, 1999; Bennett, 1990). This research took a first step in filling this knowledge gap, by applying the new institutionalism theory, and more specifically, the homogeneity hypothesis, to non-Western countries. A content analysis of media coverage of the South China Sea dispute was conducted across three countries, namely the US, China, and Taiwan. By and large, results supported the expectation of homogeneity of news content, in both Western and Asian contexts. The homogeneity hypothesis on both the fraction of

(31)

authoritative sources and the balance of the presented viewpoints in the newspaper articles were largely confirmed at the newspaper level and the country level.

Regarding the homogeneity hypothesis on the fraction of authoritative sources, a high level of homogeneity was found across news outlets within all three examined countries. That is to say, the newspapers within a country used the same amount of authoritative sources, which are the “societal indicators of achievement” (Sparrow, 1999, p. 126). The findings were in accordance with the new institutionalism concept that within a similar political and regulatory environment journalists tend to produce similar news content

(Tuchman, 1973; DiMaggio & Powell, 1983; Cook, 2006; Sparrow, 2006). At the same time, as indicated in both new institutionalism and indexing theory, news media tend to rely

heavily on authoritative sources. This expectation was confirmed by the empirical analysis. The three US newspapers reported greater identical reliance on authoritative sources, as compared to the Chinese and Taiwanese newspapers. The use of a limited set of authoritative sources leads individual journalists to produce similar news, which avoids upsetting or

destabilizing the news organization’s own commercial, professional, or informational interests (Sigal, 1973; Sparrow, 1999).

From a cross-country comparative perspective, this research revealed that newspapers across countries similarly show a heavier reliance on authoritative sources than on non-authoritative sources. However, newspapers across countries do exhibit different levels of dependence on authoritative sources. It is not surprising that in China, with its authoritative political system and media landscape, authoritative sources are very often used in the national newspapers. Yet the US newspapers, which situate in a more independent environment from the state, relied even more on authoritative sources than the Chinese newspapers. Even after several other influences were controlled for, the US newspapers still reported a greater level of reliance on authoritative sources, while the Chinese newspapers and Taiwanese newspaper

(32)

did not report significant difference in this matter. One explanation points to the influence of the type of news issues in the newspaper coverage. The South China Sea dispute is a

domestic issue to the Chinese and Taiwanese newspaper, but a foreign issue to the US newspapers. The US newspapers’ orientation towards officialdom can be higher on foreign policy issues, where there is less access to non-authoritative sources (e.g. opposition groups, local citizens who are directly involved in the South China Sea dispute) (Bennett, 2012; Hemmings, 2011). Moreover, foreign conflicts involve less familiar conditions to the news media and high national security stakes, which may give officials more room to define reality as they see fit in the media. Therefore, in this high-stake foreign policy issue, powerful actors talk with the loudest voice across the three examined US newspapers (Bennett et al., 2007). The mixed results demonstrate the substantial influence of news issues on news media’s reliance on authoritative sources (Cook, 2006).

Regarding the homogeneity hypothesis of the presented points of view in news content, more nuanced findings were found. Firstly, all newspapers’ coverage reported a favorable point of view towards their respective national government. The findings confirmed the expectation from both the new institutionalism and indexing theory that the opinion orientation presented in the newspapers tend to reflect the political stances of their government authorities. The news outlets’ heavy reliance on authoritative sources, which mostly favor their national governments, leads to a similar opinion orientation to national authorities across news media within a country (Cook, 1998; Bennett, 1990; Hagen, 1993). Secondly, due to the reliance on authoritative sources, all the US and Chinese newspapers reported similar viewpoints towards other countries (positive, neutral or mixed, or negative). For instance, the US newspapers presented a rather high level of skepticism towards China, while the Chinese newspapers reported even stronger negativity towards the US. This reflects the current situation in which the US and China are direct competitors for power in

(33)

the South China Sea region. This paper found more nuanced findings across the Taiwanese newspapers, in which there were more differences in the presented viewpoints of the US, China, and also Taiwan. For example, all the Taiwanese newspapers presented negative viewpoints towards China, but there are some differences in the degrees of negativism towards this country. Liberty Times articles presented the most negative views of China compared to the other two Taiwanese newspapers. This finding reflects the media landscape in Taiwan. The three selected Taiwanese newspapers have very distinct and divergent

political stances, which are highly influenced by their owners’ political orientations (Lin, 2008). The Taiwanese newspapers reported critically on China and positively on the US. This corresponds with the fact that Taiwan and China have conflicting claims in the dispute, while Taiwan and the US are allies.

Apart from the two main hypotheses, the findings also revealed the distribution of the national origin of sources to be in accordance with the indexing theory, which expects that the news media implicitly index the dynamic power sphere (Bennett, 1990; Bennett et al., 2006; Lawrence, 2006). Even though Taiwan has conflicting territorial claims to China and is directly involved in the dispute, Taiwanese sources were considerably less mentioned than the US sources in both the US and Chinese newspapers. This can be explained by the fact that Taiwan is less influential in determining the development in the South China Sea dispute than the US.

Limitations and Suggestions for Future Research

This research has several limitations. Firstly, although the news events reported by the media are considered to be an important aspect of homogeneity (Cook, 2006), this aspect was not examined. This is because only the South China Sea dispute was included in this research. The characteristics of issues can have a significant impact on the patterns of media indexing within a country and across countries (Bennett, 1990). For instance, as discussed

(34)

earlier, the US newspapers may rely more heavily on authoritative sources in foreign policy issues, because they have more difficulties in collecting original (e.g. local observer) data on the event. Consequently, the conclusions are mostly specific to this case, and not

generalizable to other topics (e.g. domestic issues, issues with no dispute involved). The effect of issues on the homogeneity hypothesis of news content should be further studied.

Secondly, the three countries included in this research all have strong claims in the dispute. The selection of countries may have a strong influence on the results. This limitation could have been addressed by including additional countries, which have more neutral political stances in the dispute, and by testing whether their performances are somewhere in the middle.

Thirdly, a limitation of this research in testing the homogeneity hypothesis on the news content is the selection of the medium: only newspapers were included in the sample. New institutionalists suggest a similar set of media practices and news content across modalities (e.g. newspapers, televisions, radios) (Cook, 2006; Entman, 2006). This thesis suggests future research to apply the new institutionalism theory across different types of media.

Lastly, this research has a critical limitation in the theoretical framework: this thesis cannot explain the variances of newspaper coverage within a country and across countries in the findings. This research adopted the new institutionalism approach, which suggests several broad indicators of news content (e.g. the range of sources, points of view, and issues), and is mostly interested in explaining the similarities through the homogeneity hypothesis (e.g. Hall & Taylor, 1996; Cook, 1998). Such a theoretical framework fails to explain and assess (1) how the institutional factors (e.g. uncertainty to find timely

information) within different countries affect their news media coverage; (2) why there is variance in the indicators both within and across countries in the findings (e.g. the US

(35)

newspapers reported the highest reliance on authoritative sources); (3) how to empirically assess the interaction between the indicators, for instance, how the type of issue (e.g.

domestic or foreign issue) influences the range of sources and the presented viewpoints in the news content (Hall & Taylor, 1996; Bennett et al., 2007). Thus, this thesis suggests further research to address how to assess and explain the impact of institutionalization factors on news content.

(36)

Appendix A: Search Strings of Keywords For selecting English articles

Keywords: “South China Sea” or “Pratas Islands” or “Macclesfield Bank” or “Spratly Islands” or “Paracel Islands” or “Scarborough Shoal”

For selecting Chinese articles

“南海”or “南中国海”or “三沙”or “西沙”or “南沙” or “中沙” or “东沙” or “黄岩岛” For selecting Taiwanese articles

“南海” or “南中國海” or “南沙” or “太平島” or “東沙” or “西沙” or “永興島” or “ 中沙” or “黃岩島”

(37)

Appendix B: Codebook

Homogeneity of News in Chinese/Taiwanese/American Newspaper 30/04/2016

Aim of study

The study aims to investigate the new institutionalism approach to the news by examining the homogeneity of news articles, by testing the range of sources and points of view in the news content.

Coding instruction

The codebook consists of three sections: (1) general information about the article; (2) the range of sources; and (3) points of view of the sourced messages. Coding needs to be done following the instructions in the codebook, and is done in Excel.

Coding procedures

Step 1: Determine if an article needs to be coded

a. Coders need to read the headline and first paragraph to determine whether an article should be coded. If necessary, give the article a read-through.

b. An article will only be coded if it satisfies the following conditions: (1) an article contains one keyword of a list of keywords (e.g. the South China Sea, 南海, 南中国海); (2) South China Sea is the main issue in the article; (3) an article refers explicitly to at least one of the three nations, China, Taiwan, and the United States; (4) at least one source should be included. The articles were selected based on the list of keywords, therefore, coders only need to determine whether an article satisfies all the last three conditions (2-4); if not, quit coding the article.

c. Which articles are not included? (1) An article has no source; (2) articles from sports, advertising, and supplements sections.

(38)

Step 2: Answer questions in section A – general information of an article a. Coders need to give the article a read-through to answer the questions in the first section, including the basic information and general points of view of the article.

b. Coders are suggested to mark all the sources when reading for the first time. This will help the coders answer the questions in the following two sections.

Step 3: Answer questions in section B – the range of sources

a. The registration unit for this content analysis is at the source-level. It means that coders need to identify each source and answer all questions in Excel for each source.

b. Sources are understood as actors (e.g. government officials, national newspaper, organizations), whose voices/opinions are mentioned in the news content. Actors that are not named, but only described, but in a way that can only refer to one actor (e.g. the Chinese President) should be coded.

c. All actors that are quoted directly (“…”) or indirectly (‘Person X started that...’ or ‘According to Person Y…’) are coded as a separate, new record for which needs to put in a new line in the Excel sheet. Note that whenever an opinion of an actor is present, this is sufficient to code this as a source (‘Person Z agreed with this.’). In case of doubt, do code the source.

Step 4: Answer questions in section C – points of view

a. Coder will be asked to identify the publication’s point of view towards certain country/region: China, Taiwan and the United States.

b. An article is “supportive” towards a country when the argument is in favour of the nation’s position or interests.

c. An article is “neutral/mixed” towards a country when the argument does not declare a clear opinion towards the country, or hold a mixed point of view.

(39)

d. An article is “against/skeptical” towards a country when the argument shows disagreement or skepticism towards a country’s position or interests.

Section A: General information about the article Identification number.

Indicate the number of the article as shown in the coding sheet (e.g. 01005) Source of Article.

1 = New York Times 2 = USA Today 3 = Washington Post 4 = People’s Daily/ 人民日报 5 = Guangming Daily/ 光明日报 6 = Xinhua Daily/ 新华每日电讯 7 = China Times/ 中国时报 8 = United Daily News/ 联合日报 9 = Liberty Times / 自由時報 Date of article

Indicate the date (DD/MM/YY) when the article was published (e.g. 12/03/2015) V1 Type of article

Indicate the type of the article. It can be referred to the section where the article was published. If it is not available or not identifiable, coder needs to decide after a read-through.

1 = News article (news articles inform readers about things that are happening in the world or in the local area.)

2 = Feature article (a feature story focuses on particular people, places, and events. It goes into great detail regarding concepts and ideas of specific market interest.)

(40)

3 = Editorial, columns or opinion pieces V2 Main topic of article

Indicate the main topic of the article. If there are several topics mentions, choose the main topic.

1 = Territorial claims

2 = Natural resources (e.g. oil, natural gas reserves, fish)

3 = Military presence (e.g. maintenance of freedom of navigation, military encounter, military practices)

4 = Piracy and sea robbery

5 = Environmental pollution and resource depletion

6 = Region management (e.g. coast guards’ lives, regional disaster management) 7 = Other economic development (e.g. pan-regional economic development. EXCLUDE an article when it covers the topic of natural resources)

8 = Others (Not included in the category, or unclassifiable) V3 Article’s point of view towards China

1 = Supportive 2 = Neutral/ Mixed 3 = Against/ Skeptical

99 = Not applicable (when the country is not mentioned) V4 Article’s point of view towards the US

1 = Supportive 2 = Neutral/ Mixed 3 = Against/ Skeptical

99 = Not applicable (when the country is not mentioned) V5 Article’s point of view towards Taiwan

(41)

1 = Supportive 2 = Neutral/ Mixed 3 = Against/ Skeptical

99 = Not applicable (when the country is not mentioned) Section B: The Range of Sources

Name of source ___

Code the name of the source. It can be a name of a person or an organization. This should be coded “anonymous” if no name provided, for both people and organizations.

V6 Nationality of source

Indicate the abbreviation of the nationality of the source. For this variable, it is allowed to check nationalities on the Internet if you are not sure. Code “Anonymous” if the nationality of the source is not identifiable. Code “IT” if the source is from international or pan-regional organizations or institutions.

V7 Type of source People.

1 = Incumbent politician (party leaders and members of government, regional and local politicians, etc.)

2 = Opposition politician (party leaders and members, which belong to the party that is not part of the government.)

3 = Independent/Non-partisan politician (politicians are not affiliated to any political party)

4 = Unclassifiable politician (politicians of which political affiliation cannot be determined)

(42)

6 = Military members (people who work for the armed force of a country, e.g. military officials, pentagon officials)

7 = Expert (people who have special skills or knowledge, e.g. academic analysts, scientists, lawyers, doctors, representatives from research institutes or consultancy)

8 = Civil society representative (e.g. spokespersons for grassroots organizations, etc.) 9 = Ordinary citizen (citizens without expertise or relevant title, e.g. bystanders, laypersons, audience member)

10 = Business person (e.g. shop owner, CEO, etc.) 11 = Celebrity/Athlete/Artist

12 = Journalist (this is NOT the author of the article, but another journalist mentioned in the article)

13 = Other persons (persons that identities cannot be identified)

19 = Other official sources (persons that identities can be identified, e.g. US officials) Organizations/institutions.

14 = State organizations/institutions (organizations/ institutions that belong to the government, or owned by the government, e.g. Ministry of Foreign Affairs, Ministry of National Defense, Obama administration)

15 = Non-state organizations/institutions (e.g. NGOs, multinational corporations, religious groups, violent non-state actors, etc. EXCLUDED if it is a news organization)

16 = News organizations

17 = Other organizations/institutions Unclassifiable.

18 = Unclassifiable

Section C: Sourced message’s points of view V8 Character of sourced message

Referenties

GERELATEERDE DOCUMENTEN

It is important to remember that, here, multifractal planning strategy adheres to the following planning principles [ 1 , 2 ]: hierarchical (polycentric) urban development to

The focus is on developing robust proxies to go beyond the physical evaluation perspective, and to extract socio- economic information and functional assessment of urban areas using

Although this study has shown that this work-up likely improves the probability that patients are cor- rectly diagnosed with the underlying cause of anaemia, it is unknown whether

De kosten hiervan zijn gerelateerd aan het aantal afgeleverde lammeren. Deze kosten namen het afgelopen jaar toe met ƒ 5,- per gemiddeld aan-

De gemeente heeft behoefte aan regionale afstemming omtrent het evenementenbeleid omdat zij afhankelijk zijn van de politie en brandweer voor inzet: ‘wij hebben

This research will investigate the relationship between the regulatory focus theory and the question if decision-makers of a firm would consider to internationalize,

Het doel van dit onderzoek is om te onderzoeken in hoeverre het gebruik van CSR-communicatie op social media door supermarkten een positief effect heeft op de Consumer

Voor de respondenten met de Duitse nationaliteit was geen significant verschil gevonden wat betreft de perceptie van symbolische waarde van het Engels ten opzichte van de