MORALITY IN BETWEEN THE LINES: A CONTENT ANALYSIS OF MORAL DOMAINS IN NEWS By Lu Wang A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Communication – Master of Arts 2013 ABSTRACT MORALITY IN BETWEEN THE LINES: A CONTENT ANALYSIS OF MORAL DOMAINS IN NEWS By Lu Wang This study investigated how newspaper headlines differ in reporting morally relevant content. Analyses compared the representation of four key moral domains (care, fairness, authority, and loyalty) in newspapers that differed on three variables: (1) organizational goals (market-appeal versus maintaininggovernment-control), (2) target audience (foreign versus domestic), and (3) nation of origin (China versus U.S.). Target audience was nested within the organizational goals condition of maintaininggovernment-control. A coding scheme based on logic underlying the model of intuitive morality and exemplars was applied to four purposely selected newspapers: People’s Daily (representing governmentcontrolled newspapers in China with domestic targets), and China Daily (representing governmentcontrolled newspapers in China with foreign targets), Nanfang Metropolis Daily (representing marketappeal newspapers in China), The Washington Post (representing market-appeal newspapers in the U.S.). Results show patterns consistent with two predictions. First, among government-controlled newspapers, those targeting domestic readers were more likely to report stories featuring authority than those targeting foreign readers (H3a supported). Second, the Chinese marketappeal newspaper demonstrated a tendency to report stories featuring fairness more often than its government-controlled counterpart (H2b supported). In addition to these predicted finding, noteworthy patterns unrelated to any hypothesis indicated that a) the market-appeal newspapers tended to report more on authority than the government-controlled newspaper targeting foreign readers; and b) the market-appeal newspapers both in China and in the U.S. tended to report more on fairness than the Chinese government-controlled newspaper targeting foreign readers. TABLE OF CONTENTS LIST OF TABLES ………..………………………………………………………….……….….iv LIST OF FIGURES ……………………………………………………………….….………….v INTRODUCTION ………………….…………………………………………….…………...….1 THEORETICAL FRAMEWORK ……………………….………………………………………3 Model of Intuitive Morality and Exemplars (MIME) …………………………...…...…...3 Authoritarian and Libertarian Newspaper Systems ..…………………………...…..….....5 News Industry Reforms in China ……………………………………………...…...…..…7 Hypotheses And Research Questions ……………………...………………...……..…….9 METHOD …………………………………………..………………………..……………..…..13 Sample …………………...………………………………..……………….……….....…13 Unit of Analysis ……………….……………………..…………….………………..…..14 Independent Variables ……………….……………………..………………….……......15 Organizational goals ………….…………………….………………….……......15 Target audience ………….…………………...……………………….……......15 Newspaper’s nation of origin …………………………….………….…………16 Dependent Variables ….…………….……………………………………….………....16 Care ….…………….…………………………………...…………….…………16 Fairness ...………….…………………………………...…………….…………17 Loyalty ...………….…………………….……………...…………….…………17 Authority ……………….……..…………………………………….….….……18 Intercoder Reliability……………….……………………………………….….….……18 RESULTS…………….………………………………………………….…………….…….…19 DISCUSSION……………….…………………………………………………………………...26 Portrayal of Conservative Domains: Authority and Loyalty …………………………..27 Portrayal of Progressive Domains: Care and Fairness ………………………………... 28 Non-hypothesized Findings on Authority …………………………………...…… ...…30 Limitations ………………………………………………………………………………32 Conclusion and Future Directions…………………………………...……….…………34 APPENDIX ……………….……………………………………………………………….…….36 REFERENCES ……………….…………………………………………...……………….……49 iii LIST OF TABLES Table 1: Indicator presence agreement percent and Scott’s Pi results in terms of separate domains …………………………………….………………..……………………....18 Table 2: Chi-Square Frequency and Percent Distributions for Features in terms of Newspapers……………………………………………….……………..………..….20 Table 3: Chi-Square Frequency and Percent Distributions for Features in terms of moral domains and Newspapers…………….…………………………….….….……………........21 Table 4: Chi Square Frequency and Percent Distributions for Features of Care ………..…..30 Table 5: Chi Square Frequency and Percent Distributions for Features of Authority1 …..…31 Table 6: Chi Square Frequency and Percent Distributions for Features of Authority2 ……...32 . iv LIST OF FIGURES Figure 1: The MIME ….….….….….….….….….….……...….….….….….….….….….….….3 v INTRODUCTION Recent media industry reformation in China has led to loosened governmental control of some news organizations. The deregulation resulted in increased market appeal and financially independent operations. With the rapid emergence of market-appeal newspapers, governmentcontrolled newspapers focused on Chinese unity still remain its influence. As such, systematic variance is expected as a result of distinguished goals of these two polarized newspaper types. Chinese media scholars have been arguing that there exist qualitative differences between government-controlled and market-appeal newspapers in China. Scholars attempted to identify these differences with content analytic studies which compare these two types of papers along a variety of content domains. For example, studies have examined general differences (Li & Jia, 1999; Xiang, 1996; Yang, 1999), layout differences (Xi, 1995) and specific content differences in soft news (e.g., stories related to murder and sex; Ai, 1996; Huang, 2001). Although these studies have proven informative in many ways, research in this area would benefit from the development of a coherent scheme capable of distinguishing the central features along which the content of these papers differs. To date, one feature that several attempts to compare is the social or moral values underlying stories. Media scholars assume that moral terms or concepts are consistently present in news stories and will affect the readers (Mastro, Enriquez, Bowman, Prabhu, & Tamborini, 2012). Nevertheless, one important question has remained unanswered yet: how are the moral values reflected in news stories? Several studies have examined this issue regarding morally related content in news reporting (Jiang, 2005; Semetko & Valkenburg, 2000; Pickle, Quinn & Brown, 2002). For instance, Pickle et al. (2002) as well as Semetko and Valkenburg (2000) coded news stories for the presence of a general moral frame. Jiang (2005) coded news stories 1 using a coding scheme which only measured positive versus negative valence of moral content. However, these attempts suffer from the lack of a coherent theory that offers a systematic framework to define features of moral content. The current study expects to overcome this issue by developing a coding scheme based on a comprehensive theoretical framework of morality, the model of intuitive morality and exemplars (MIME; Tamborini, 2011, 2012), to study differences between newspaper content in major Chinese and U.S. newspapers. The MIME suggests that indicators and frame of evolutionarily-derived moral domains can help distinguish the manner in which media represents moral values. Previous MIME research demonstrated that the moral domains are relevant for news content evaluation (Bowman, Lewis, & Bryand, 2012). More theoretically, researchers have also argued that morality-relevant content is either overtly or covertly common in news stories (Lind, 1996; Zillmann & Knobloch, 2001; Zillmann, Taylor, & Lewis, 1998). This paper first reviews the MIME and MFT. Next, it reviews literature distinguishing Authoritarian and Libertarian newspaper systems as they might exist in China and the U.S. Following from this, it provides a terse and concise introduction to Chinese news industry reforms and changes in recent years, and argues for the use of the newspapers selected in this study. Finally, it offers hypotheses predicting content differences in the selected newspapers and presents a content analytical study to test those predictions. A coding scheme based on the logic of the MIME will be applied to four purposely selected newspapers: People’s Daily (representing government-controlled newspapers with domestic targets), and China Daily (representing government-controlled newspapers with foreign targets), Nanfang Metropolis Daily (one influential market-appeal newspaper in China) and The Washington Post (representing market-appeal newspapers in the U.S.). 2 THEORETICAL FRAMEWORK Model of Intuitive Morality and Exemplars (MIME) The MIME combines theories of media influence with advances in moral psychology to describe the dynamic reciprocal processes that link audiences and their environments. This broad model outlines macro and micro processes that can be used to predict cultural differences in media content (see Figure 1). The macro-level processes describe how: a) collective evaluations and selection of media can influence the subsequent mass production of content produced for audiences and b) cultural variations in an audience’s emphasis on different moral foundations will be represented in the media content produced for unique cultures. As such, the MIME provides a foundation for the study of questions regarding content differences in various media systems, including the content differences between Chinese and U.S. newspapers. Figure 1.The MIME Production of Content Adhering to Salient Foundations/Exemplar Exposure to Content Cultural Environment Exemplar Salience Domain Salience Automatic Appraisal Controlled Reappraisal Selective Exposure to Liked Content Most previous research on the MIME has focused on the short-term micro processes related to the evaluation of media content (cf. Tamborini, 2012); however, more recent research 3 has begun to examine long-term macro components of the model predicting differences expected in media systems that produce content for different cultures. For example, research examining the portrayal of moral content in Spanish and English Language soap operas (Tamborini, Enriquez, Lewis, Grizzard, & Mastro, 2011) has demonstrated the moral foundations will be featured differently in entertainment produced for diverse cultural audiences. This study adds to research on the MIME by testing hypotheses predicting differences in the moral content of news produced for various audiences in China and the U.S. By comparing news produced for different cultures, the present study makes an important contribution to research on the MIME, which specifically includes informational as well as entertaining media. The MIME’s logical structure is based in part on the understanding of intuitive morality offered by Haidt and Joseph’s moral foundations theory (MFT). The core of this understanding can be traced back to the Shweder et al.’s (1997) discussion of the “big three” ethics, which identified autonomy, community, and divinity as three universally-shared intuitive mechanisms that can drive human moral judgments. MFT also claims that there exist universally accepted intuitive drives related to human morality, but extends the list of moral drives identified in the big three to five innate and universally-shared intuitive mental systems (Haidt & Joseph, 2008). MFT argues that humans developed this set of intuitive mental systems through evolutionary processes, and categorized these intuitive mechanisms as five moral domains that include: care (developed from mammal evolution, an ability to sense and dislike the pain of others); fairness (based on processes of reciprocal altruism, the need for justice and equity); ingroup loyalty (linked to the benefits of communal living, self-sacrifice for the group); authority (derived from hierarchical social interactions, respect for legitimate leadership and traditions); purity (based on aversion toward bodily contamination and respect for the sacred, the need for an 4 elevated and more noble life; Haidt & Joseph, 2004). The care, fairness domains are generally viewed as progressive domains while the authority, loyalty and purity domains are labeled as conservative domains based on their definitions (Graham et al., 2009). The distinction between progressive and conservative domains may be relevant to our understanding of moral content in Chinese newspapers. Notably, as a result of reforms in postMao China’s market-oriented media industry, the newspaper industry in China is shifting from a totally government-owned system in which content decisions made to facilitate governmentcontrol goals to a system driven more by the desire to increase market-appeal. This shift has contributed to a unique dual-system feature found in China’s media industry where a traditional government-controlled media system coexists with a newly market-appeal media system. As these two systems are thought to be driven by different goals, we might expect to find differences in the conservative and progressive social values reflected in the content of newspapers operating under these different systems. Although there may be great ambiguity regarding the manner in which conservative and progressive social values can be distinguished, the MIME provides a coherent theoretical framework that offers a systematic method for making these distinctions. When applied to news content, this framework suggests both positive and negative indicators of specific moral domains. For instance, the care domain may be represented by a positive indicator such as the term compassion, (which suggests emotional concern towards others’ pain) or a negative indicator such as indifference (which refers to acts of not assisting another who is suffering physical or emotional pain). Potential positive and negative indicators for each domain are discussed in a later section of this paper, and lists for each are provided along with a coding scheme. Authoritarian and Libertarian Newspaper Systems 5 The conservative and progressive values reflected in the two media systems suggested above can be understood in line with distinctions between the Authoritarian and Libertarian newspaper systems discussed in early literature on the Four Theories of the Press (Siebert et al., 1956). Specifically, conservative values are more aligned with the principles of Authoritarian systems whereas progressive values are more consistent with the principles of Libertarian systems. If we apply this claim to the study of moral content in news, we should expect systematic content differences in Chinese newspapers driven by the types of conservative goals often associated with government control versus newspapers driven by the types of progressive goals associated with market appeal. According to the Authoritarian theory of the press (Siebert et al., 1956), mass media are created, or should be placed, under the control of government and should not challenge or criticize on the government. In such a system, mass media works as an instrument to strengthen the power of the government instead of as a gatekeeper. The government-controlled media systems act as a mouthpiece expressing the views and values of those in control in order to enhance the authority of legitimate powers. As such, it is reasonable to argue that the media environment found in the government-controlled media system in China is parallel to that depicted in the Authoritarian theory of the press. As suggested by the MIME, such a system will have a greater tendency to report stories featuring conservative moral domains (i.e., authority and loyalty). By contrast, the Libertarian theory of the press argues that mass media should keep the government in check and have greater freedom to report stories that work towards the fulfillment of other goals such as informing, entertaining and making a profit (Siebert et al., 1956). 6 According to the Libertarian theory of the press, one important goal of media is to entertain and sell, which may be facilitated more by reporting stories that featuring progressive domains. Although the treatment of all moral domains may be important contributors to the entertainment appeal of newspapers, the care and fairness domains have been identified as particularly central to entertainment’s dramatic appeal (Tamborini et al., 2012). Therefore, it is reasonable to claim that while Libertarian media will tend to report on all domains, their attempt to entertain and to sell will lead them to focus mostly on the care and fairness domains. News Industry Reforms in China In the past twenty years, radical change in Chinese news industry has reshaped the general pattern and polarized the organizational goals of newspapers (i.e., maintaining government control versus increasing market appeal). This polarization is consistent with the MIME’s description of media systems that feature conservative versus progressive domains and how these domains would be represented in Authoritarian versus Libertarian press systems. According to the MIME, it is reasonable to argue that legitimate power (i.e., the Party) will strive to maintain social control by regulating media content to insert conservative ideologies. As such, in order to imbue society with their views, the government is likely to use newspapers and television news for propaganda purposes. Such government-controlled newspapers can be categorized as operating under an Authoritarian press system. Remnants of the traditional newspaper system that existed prior to Post-Mao economic and societal revolution remain in China today. Such government-controlled newspapers have little freedom to select news content and are under strict censorship directly from Party leaders. Any news covered on the front-page or in the politics section will always be censored by officers of the Party hierarchy before being passed down to the publishers. External factors in this case 7 have minimal impact on the content selected for government-controlled newspapers. As such, these government-controlled newspapers can be considered part of the Authoritarian press system. The logic underlying the MIME is also consistent with arguments suggesting that economic forces will drive newspapers to feature progressive ideological content. This should be true for newspapers working toward market-appeal goals, where external factors such as profit motives are more central in the determinants of content. Along with the Post-Mao economic and societal revolution, four major reforms that took place in China’s news industry shifted the news media from completely reflecting the goals of the government to increasingly reflecting market appeal (Chen & Chan, 1998; Chan, 1993, 1995; He, 2000; Yu, 1994; Zhang & Zhu, 2006). Some newspapers started to report stories featuring progressive domains while simultaneously covering stories featuring conservative domains. The fourth reform was particularly important in the development of China’s dual-system features. In the fourth reform, many official documents designed to loosen the government’s control of the news media were released. These materials included two famous “Circular on the Issuing of the Implementation Plan for the Pilot Work of Press and Publishing Reform,” and the “Circular of the State Council Concerning the Issuing of the Two Provisions Supporting the Development of Cultural Industry and Transformation of ForProfit Cultural Institutions into Enterprises in the Pilot Work of Cultural System Reform.” Following their release, loosened regulation and increased freedom to determine content encouraged the emergence of market-appeal newspapers. The emergence of the market-appeal newspaper in China may have caused the decline in the circulation of government-controlled newspapers observed from 1995 to 2001,when government-controlled newspapers underwent a 18.38% decline in circulation (from 217.17 8 million CNY to 177.25 million CNY). By contrast, newly issued market-appeal newspapers gained considerable growth during the same period (Huang, 2001). Since then, leading marketappeal newspaper organizations, such as Shanghai Daily Group (Xinmin Evening Daily) and Nanfang Media Group (Nanfang Metropolis Daily) successfully gained some degree of financial independency and coverage autonomy. Notably, although the recent development of market oriented newspapers shows a change in China’s dual-system news industry toward increased freedom to express discontent, there exist a number of nationally influential newspapers (e.g., People’s Daily) which still greatly resemble the authoritarian nature in many aspects. Today, news organizations that operate both under Authoritarian and Libertarian press forces survive in China. Hypotheses and Research Questions In light of previous discussion, when we consider the MIME’s predictions about conservative and progressive content along with theories of both the Authoritarian and Libertarian press, it is reasonable to expect that compared to market appeal papers, governmentcontrolled newspapers in China are more likely to report stories which assist in strengthening the Party’s legitimate power. Such newspapers should also work to facilitate patriotism and societal coherence in an effort to perpetuate political stability. As such, we would expect these newspapers to report more stories that feature the authority and loyalty domains. Hence, People’s Daily, as a most influential party-owned government-controlled newspaper, is expected to include more stories that feature both conservative domains (i.e., authority and loyalty): H1. Among newspapers targeting Chinese readers, headlines in government-controlled newspapers are more likely than those in market-appeal newspapers to feature (a) authority and (b) loyalty. 9 Although the underlying logic of MIME indicates that media driven by market-appeal goals will be more likely to report stories that feature both conservative (i.e., authority and loyalty) and progressive domains (i.e., care, fairness), the Libertarian theory of the press suggests that newspapers in an open-market will focus more on stories that will entertain and sell. We might expect these newspapers to include more sensational topics such as those related to violence, murder, legal disputes and etc. This expectation is consistent with Tamborini et al.’s (2012) suggestion that the care and fairness domains are more germane to the appeal of media content. For instance, they suggest dramatic conflict is most often built on the care and fairness domains, which are central to disposition theory (Zillmann et al., 1998). This logic would argue that violations of the care and fairness domains are used to create the threat of harm or injustice, and that enjoyment results when compassion is shown or justice is restored in the end. As these domains are central to dramatic appeal, we might expect market-appeal newspapers to take advantage of these features and more often include progressive domains (i.e., care and fairness) related content in their coverage: H2 Among newspapers targeting Chinese readers, headlines in market-appeal newspapers are more likely than those in government-controlled newspapers to feature (a) care, and (b) fairness. With Deng’s “open-door” policy in 1978, China started an era of seeking positive stature and enlarging its influence internationally (Luther, & Zhou, 2005). The government founded China Daily in 1981 which is the only Party-owned English-language newspaper targeting international audiences. China Daily attempts to inform the western world about China’s new policies and to improve China’s international reputation. People’s Daily is the “sister” version of China Daily. It was first published in 1948 and is a Party-owned newspaper targeting domestic 10 readership. The selection of news covered in China Daily and People’s Daily is under strict and direct control of the Chinese Communist Party. As such, both newspapers act as a mouthpiece for the Party and are completely government controlled. Comparing China Daily and People’s Daily can help us identify content differences between government-controlled newspapers with distinct target audiences. Notably, freedom of speech in media is highly valued by many of the democratic nations which are targeted by China Daily. Moreover, media organizations in those nations are more likely to operate under a Libertarian press system than an Authoritarian press system. As such, news reporting in these nations is expected to be more balanced in terms of reporting conservative and progressive domains. In order to reach these audiences, we might expect the Party to allow China Daily to mimic a democratic reporting style in a manner that challenges conservative domains (i.e., authority and loyalty). Such expectations are consistent with the MIME’s argument that successful media producers tend to offer content which is consistent with the values of target audiences. Moreover, these expectations are supported by previous research showing that news frames used in the New York Times were more similar to China Daily than People’s Daily (Luther and Zhou, 2005). In light of the MIME’s logic and previous findings, this paper hypothesizes: H3: Among Chinese government-controlled newspapers, headlines in those targeting domestic readers are more likely than headlines in those targeting foreign readers to feature (a) authority, and (b) loyalty. In addition to focusing less on authority and loyalty, government-controlled newspapers targeting foreign readers (i.e., China Daily) are more likely to report issues related to care and fairness domains. Although China Daily is under direct control of the Party and resembles 11 newspapers operating under an Authoritarian press system, when compared to People’s Daily, it has greater motivation to mimic the reporting style found in a Libertarian press system in order to attract its foreign readers. As stated above, this would include a greater focus on progressive domains (i.e., care and fairness). Therefore, this paper hypothesizes: H4: Among Chinese government-controlled newspapers, headlines in those targeting foreign readers are more likely than those targeting domestic readers to feature (a) care, and (b) fairness. Issues regarding expected differences between the distinct forms of market-appeal newspapers found in China versus the U.S. are less clear. This is particularly true when we consider the ubiquitous government regulation which constrains news selection in China and defines the still limited boundaries of freedom gained by these emerging market-appeal newspapers. As stated, since freedom of speech in a major determinant of a newspapers ability to challenge authority and loyalty, the more limited boundaries of freedom found in the marketappeal newspapers of China versus the U.S. should lead to comparatively fewer negative indicators and more positive indicators of authority and loyalty. Yet in terms of presence versus absence of either individual domain, it is hard to know which type of newspaper may report more stories featuring them. Though it may be reasonable to assume that market-appeal newspapers in China will have fewer negative indicators and more positive indicators of authority and loyalty while market-appeal newspapers in the U.S. will have more negative indicators and fewer positive indicators, it is unclear whether the total number of indicators should differ. Hence we raise a research question in this case: RQ: Will headlines in Chinese market-appeal newspapers be more likely than those in U.S. market-appeal newspapers to feature (a) authority and (b) loyalty? 12 METHOD Sample The content analysis was applied to a sample of four purposively selected newspapers: People’s Daily, Nanfang Metropolis Daily, China Daily, and The Washington Post. Sample selection began with an understanding that it is impossible to find individual newspapers capable of representing all other newspapers in some of the categories examined. The four chosen were carefully and purposefully selected based on their prominence in China and the U.S. All the three newspapers in China are nationally accessible with large circulations. First, People's Daily is under direct regulation of the Party, and its agenda is closely followed by all governmentcontrolled newspapers. For this category, an argument can be made that People's Daily is well representative of all government-controlled newspapers. Second, Nanfang Metropolis Daily is one of the most influential metropolitan newspapers in China and has been known for the audacious and inquisitive reporting style found in several market-appeal Chinese newspapers. In this regard, although no pure market-appeal newspapers exist within the Chinese government’s strictly regulated media system, Nanfang Metropolis Daily is a reasonable choice to represent market-appeal newspapers in Mainland China. Third, China Daily is the only official government-controlled newspaper targeting foreign readership and thus was selected for this comparison purpose. Finally, The Washington Post, which is generally considered to be one of the most influential metropolitan U.S. newspapers, was selected to be compared with the marketappeal newspaper in China. One composite week of newspaper issues (i.e., Sunday through Saturday) was created from a stratified random sample of all issues published in the four newspapers during the 2011 calendar year. In other words, one of all Mondays in the 2011 calendar year was randomly 13 selected; one of all Tuesdays in the 2011 calendar year was randomly selected; and etc. In total, seven days were selected from the entire year to construct the composite week. All issue from the four newspapers published on the selected days were included in this sample. Content analysis were applied to headlines of stories in these issues excluding advertising content, editorials, pictorial presentations, entertainment, readers' feedback, economic and business discussions, and similar materials. Electronic versions of the selected newspapers were retrieved from online databases. Unit of Analysis The unit of analysis in this study was the headline of each news story. Among the principal functions of newspaper headlines, summarizing the content of the article and indicating the writer's attitude to the story are clearly crucial. (Morley, 1998) Therefore, headlines should be representative to the whole story. To test that and validate this decision, a set of 54 stories was sampled and coded. The coders coded both the headlines and the first five paragraphs of each story. The use of the first five paragraphs to represent the content of the entire article was verified by research of Fico and Cote (2002). A comparison was made between the framing of the moral domains in the headline and the framing of the moral domains in the first paragraphs. The results of this coding displayed that the framing of moral domains in headlines were consistent with the framing of moral domains in the first five paragraphs. This finding supports the use of the headline as the unit of analysis. In addition to the fact that headlines are representative of story content, an argument for the use of headlines can be made for the impact of headlines alone. Stamm and Jacoubovitch (1980) argued that people read twice as many headlines as units of text. The fact that headlines 14 are representative of the entire stories and that people tend to gain information from reading only headlines indicates the value of using headlines as units of analysis. Independent Variables This content analysis attempted to compare newspaper headlines in terms of organizational goals (government-control versus market-appeal), target readership (foreign versus domestic), and the newspaper’s nation of origin (China versus U.S.). Organizational goals. Organizational goals were defined operationally as whether or not the newspaper organization has an editor directly appointed by the government and receives filtered news stories from the government before publishing. It was assumed that non-directly state owned newspapers have the freedom to report news content that will increase market appeal whereas directly-owned newspapers only produce news content that maintains government control. People’s Daily is funded by the Party, clearly under regulation of the Party, and its news selection is pre-determined by an official appointed by the Party. Although Nanfang Metropolis Daily cannot avoid censorship, it has considerable freedom in determining how they cover stories and strong motives to cover them in a manner that will improve market appeal and decrease financial pressure. Target audience. The target audience variable applies only to Chinese newspapers in the government-controlled condition. Target audience is defined operationally in terms of whether or not the goals of government-controlled newspaper stated in their website identify that the paper is designed to attract only a foreign audience. Both People's Daily and China Daily are under direct government regulation and meet the criteria for inclusion in the government-controlled condition. The goals of China Daily, as 15 stated in their website, identify foreign readers as their target group. People’s Daily makes no statement about its target group, but is widely accepted as targeting a domestic readership. Newspaper’s nation of origin. The nation of origin is operationally defined in terms of where the paper was published (China versus U.S.). Dependent Variables Four out of five moral domains defined in the MFT were coded as dependent variables in this study. The four moral domains are care, fairness, authority, and loyalty. Purity was not included in this study because of a concern for the lack of conceptual clarity and the ability to observe it represented in media content. Based on the work of Haidt (2004), the remaining four domains are conceptually defined as follows: care is the ability to sense and dislike the pain of others; fairness is a sense of reciprocal altruism and the need for justice and equity; ingroup loyalty is favoring the benefits of communal living and selfsacrifice for the group; authority is respect for legitimate leadership and traditions. In this content analysis, framing of each domain was coded and operationalized as the follows. Each domain will be coded in terms of indicator presence. Indicator presence is coded as (+1) positive (-1) negative or (99) not present. Operational definitions are given as follow. Care. Positive indicators of care are operationally defined as (a) Compassion: the act of showing a deep awareness of and concern for another's physical and emotional suffering; (b) Care: the act of offering or providing physical and emotional assistance. This includes protecting and supporting others, physically and emotionally. Negative indicators of care are operationally defined as (a) Harm: the act of causing another’s physical or emotional pain, or planning to put another in jeopardy of physical or emotional pain; (b) Indifference: the act of choosing not to assist another who is suffering physical or emotional pain; (c) Cruelty: the act of enjoying the 16 pain or distress of others regardless of whether the pain/distress is functional or dysfunctional; or feeling no remorse towards the dysfunctional pain or distress of others. Fairness. Positive indicators of fairness are operationally defined as (a) Honesty: The act of following an implicitly or explicitly stated, socially accepted agreement at personal cost; (b) Procedural justice: The act of using a transparent, societally accepted rule-based procedure of decision making to resolve conflict; (c) Distributive justice: Fairness in the distribution of rights or resources which is further separated to (1) Equality: Acting in a manner such that the benefits people receive is based on equally meeting everybody’s needs. (2) Equity: Acting in a manner such that the benefits people receive is proportional to their contribution; and (d) Retributive justice: The act of punishing a wrong doer at a level that is accepted by the society as commensurate to the wrong doing. Negative indicators of fairness include (a) Dishonesty: The act of violating an implicitly or explicitly stated, socially accepted agreement; (b) Procedural injustice: The act of violating a transparent, societally accepted rule-based procedure of decision making; (c) Distributive injustice: unfairness in the distribution of rights or resources which is also further separated to (1) Inequality/favoritism: acting in a manner such that the benefits provided meet some individual or group’s needs more than another individual or group’s needs. (2) Inequity: acting in a manner such that the benefits people receive is not proportional to their contribution; (d) Retributive injustice: the act of punishing a wrong doer at a level that is unacceptable by the society standards as commensurate to the wrong doing. Loyalty. Positive indicators of loyalty include (a) Solidarity: Showing solidarity/ cohesiveness/ harmony towards a specific group; (b) Martyrdom: Sacrifice of any individual interests for the group; (c) Bigotry: suspiciousness or bias against the outgroup. Negative indicators of loyalty include (a) Departure: giving up ingroup membership (b) Betrayal: 17 behaviors that sacrifice or have the potential to sacrifice ingroup interests for the self or other groups’ benefits; (c) Treason. Authority. Positive indicators of authority include (a) Obedience/Deference: acts showing respect for the legitimate power of a leader or any group or institution to which the person belongs (political, religious, social, business, educational, etc.); (b) Traditionalism: acts showing respect for traditional customs and social norms of any group or institution to which the person belongs. Negative indicators of authority include (a) Disobedience: Acts showing disrespect for the legitimate power of any group or institution to which the person belongs (political, religious, social, business, educational, etc.); (b) Defiance: Acts showing disrespect for the leaders of any group or institution to which the person belongs; (c) Progressivism: acts showing disrespect for traditional customs and social norms of any group or institution to which the person belongs. Intercoder Reliability Intercoder reliability was reported in terms of agreement percentage and Scott’s Pi. Two graduate students majoring in communication independently coded the sample. Table 1. Indicator presence agreement percent and Scott’s Pi results in terms of separate domains Indicator Presence Agreement Indicator Presence Scott’s Pi Care 96% 0.90 Fairness 86% 0.79 Loyalty 96% 0.82 Authority 90% 0.84 Training occurred over an eight-week period and was conducted on separate samples of headlines selected from the same newspapers. The reliability test was conducted on a sample of 50 headlines randomly selected from the four newspapers. Results are presented in this table below. 18 RESULTS Across the four newspapers, 1386 headlines were coded in total. Among these, 23.2% (n = 355) were from People’s Daily, 20.8% (n = 317) were from Nanfang Metropolis Daily, 30.6% (n = 468) were from China Daily, and the remaining 16.1% (n = 246) were from The Washington Post. Since the samples used in this study were large enough to produce statistically significant results from modest differences that might have limited importance, it is worth noting that all the hypothesized findings on differences across newspapers or domains are significant at p = .001. Moreover, in terms of “practical significance,” all significant findings are associated with at least a 5% difference between the percent scores of the categories compared. Examination of the data began by computing descriptive statistics on the frequency with which the four separate newspapers featured any of the four moral domains in headlines. The 2 chi-square comparing all four newspapers was significant, χ (3, N = 1386) = 52.01, p < .001. Separate chi-square comparisons revealed that government-controlled newspapers targeting foreign readers featured moral domains less often than any all other newspapers, including those 2 that were government-controlled targeting domestic readers, χ (1, N = 823) = 24.10, p < .001, 2 2 market-appeal in China, χ (1, N = 785) = 45.65, p < .001, and market-appeal in the U.S., χ (1, N = 714) = 19.09, p < .001. No significant differences were observed among any of the other newspapers (see Table 2). Following the examination of any moral-domain presence in the four separate newspapers, similar analyses examined the frequency with which the four separate domains were featured in headlines across all newspapers combined. 19 Table 2. Chi-Square Frequency and Percent Distributions for Features in terms of Newspapers Newspaper Chinese Government Domestic Target People's Daily present Frequency (Percent) A 181 (51.0%) absent 174 (49.0%) Chinese Government Foreign Target China Daily present 159 (34.0%) 309 (66%) Chinese Market-Appeal Nanfang Metropolis Daily present U.S. Market-Appeal The Washington Post present absent absent absent B A 185 (58.4%) 132 (41.6%) A 125 (50.8%) 121 (49.2%) Note: Scores with no superscript in common are significantly different at p <.001 by chi-square tests. Overall, authority was featured in headlines more often than care and fairness, which were featured more often than loyalty. The chi-square comparing all four domains was 2 significant, χ (3, N = 740) = 103.91, p < .001. Separate chi-square comparisons revealed that (a) 2 2 authority was featured more often than care, χ (1, N = 462) = 8.87, p = .003, fairness, χ (1, N = 2 468) = 7.19, p = .007, and loyalty, χ (1, N = 336) = 107.44, p < .001; and (b) both care and 2 2 fairness were featured more often than loyalty, χ (1, N = 272) = 58.37, p < .001, and χ (1, N = 278) = 62.68, p < .001 respectively. No significant difference was observed between care and fairness (see Table 3). Following the initial analyses and examination of descriptive statistics, additional analyses were conducted to test the predictions represented in the four hypotheses and answer the research question. To accomplish this goal, chi-square analyses were conducted to examine the 20 presence or absence of each moral domain among headlines from the four different newspapers (Chinese government-control domestic target, government-control foreign target, market-appeal in China, market-appeal in U.S.). Separate analyses were conducted for all newspaper pairs on each individual domain. Table 3. Chi-Square Frequency and Percent Distributions for Features in terms of moral domains and Newspapers Newspaper Frequency (Percent %) Care Chinese Government present 42 Domestic Target absent 313 People's Daily (11.8) (88.2) Fairness B (14.4) 51 304 (85.6) Chinese Government present 68 Foreign Target absent 400 China Daily (14.5) 33 (7.1) 45 (9.6) (85.5) 435 (92.9) 423 (90.4) present 54 (17.0) 76 (24.0) 72 absent 263 (83.0) 241 (76.0) present 35 U.S. Market-Appeal absent 211 The Washington Post (14.2) (85.8) 45 201 (18.3) Chinese Market-Appeal Nanfang Metropolis Total present 199 (14.4) absent 1187 (85.6) b Loyalty A AB (81.7) ab 17 338 (4.8) (95.2) 31 (6.6) 437 (93.4) (22.7) 12 (3.8) 245 C 205 (14.8) 1181 (85.2) Authority A (26.8) 95 260 (73.2) (77.3) 305 (96.2) 51 195 (20.7) 13 233 (5.3) (94.7) B A A (79.3) a 263 (19.0) 1123 (81.0) c (5.3) 73 1313 (94.7) Note: Row comparisons are represented by upper case superscripts. Column comparisons are represented by lower case superscripts. Pairs with no superscript in common differ significantly by chi-square tests at p=.001. Overall, the findings from these analyses supported: (a) H2b’s prediction that fairness representations would differ in China’s government-controlled versus market-appeal newspapers, and (b) H3a’s prediction that authority representations would differ among Chinese government- 21 controlled newspapers targeting domestic versus foreign readers. No significant differences were found in line with other predictions. Results are presented in the table below. Hypothesis testing began by examining the authority and loyalty domains. Hypothesis 1 predicted that among newspapers targeting Chinese readers, headlines in government-controlled newspapers are more likely than those in market-appeal newspapers to feature authority and loyalty. Hypothesis 3 predicted that in Chinese government-controlled newspapers, headlines in those targeting domestic readers are more likely than those targeting foreign readers to feature authority and loyalty. The research question asked whether there will be differences in the extent to which headlines in Chinese versus U.S. market-appeal newspapers feature authority and loyalty. Separate chi-square analyses conducted on authority for different newspaper pairs revealed significant differences only for comparisons associated with H3a. Consistent with the predictions of H3a, among Chinese government-controlled newspapers, those targeting domestic readers (People’s Daily) were more likely to report stories featuring authority than those 2 targeting foreign readers (China Daily), χ (1, N = 823) = 42.04, p < .001. With regard to H1a, 2 though scores were in the predicted direction, differences did not reach significance: χ (1, N = 672) = 1.47, p = .225. Therefore, for authority, only H3a was supported. In terms of the research question, chi-square results failed to reveal significant differences between market-appeal newspapers in China (Nanfang Metropolis Daily) and market-appeal newspapers in the U.S. (The 2 Washington Post), χ (1, N = 563) = 0.32, p = .573. Unexpectedly, headlines for market-appeal newspapers both in China (Nanfang Metropolis Daily) and the U.S. (The Washington Post) featured authority more often than government-controlled newspapers targeting foreign readers 22 2 (China Daily). Notably, these differences were significant in both cases, χ (1, N = 785) = 25.56, 2 p < .001, χ (1, N = 714) = 17.12, p < .001, respectively. Following analyses on authority, similar chi-square analyses conducted on loyalty for different newspaper pairs failed to reveal significant differences. First, though the pattern observed for H1b was in line with the prediction that loyalty would be featured more often in China’s government-controlled newspapers than market-appeal newspapers, analysis failed to 2 produce significant results, χ (1, N = 672) = 0.41, p = .523. Second, in contrast to the prediction of H3b, patterns showed that among government-controlled newspapers, those targeting domestic readers are less likely to feature loyalty in headlines than those foreign targeting 2 domestic readers. Yet chi-square revealed no significant difference, χ (1, N = 823) = 1.24, p = .266. Therefore, for loyalty, neither H1b nor H3b was supported. Finally, regarding the research question, among market-appeal newspapers, headlines in U.S. newspapers reported more stories featuring loyalty than those in China newspapers. However, once again, the 2 difference was not significant, χ (1, N = 563) = 0.73, p = .392. Following analyses on authority and loyalty, chi-square analyses were applied to examine hypotheses regarding the care and fairness domains. Hypothesis 2 predicted that among newspapers targeting Chinese readers, headlines in market-appeal newspapers would be more likely than those in government-controlled newspapers to feature care and fairness. Hypothesis 4 predicted that among Chinese government-controlled newspapers, headlines in those targeting foreign readers would be more likely than those targeting domestic readers to report stories that feature care and fairness. 23 Chi-square analyses conducted on care for different newspaper pairs revealed no significant differences. Although patterns were consistent with H2a’s prediction that marketappeal newspapers in China would feature care more often in headlines than their government2 controlled counterparts, the difference was not significant, χ (1, N = 672) = 3.70, p = .054. Similarly, though patters were consistent with H4a’s prediction that among China’s governmentcontrolled newspapers, those targeting foreign readers would feature care more often than those 2 targeting domestic readers, again, the difference was not significant, χ (1, N = 823) = 1.27, p = .26. Therefore, for care, neither H2a nor H4a were supported. Chi-square analyses conducted on fairness for different newspaper pairs revealed significant differences associated with H2b and H4b. Notably, however, while the pattern associated with H2b was consistent with predictions the pattern observed for H4b was opposite of that predicted. As predicted by H2b, market-appeal newspapers in China featured fairness 2 more often than government-controlled newspapers in China, χ (1, N = 672) = 10.09, p = .001. In contrast to the prediction of H4b, among Chinese government-controlled newspapers, fairness was featured less often in newspapers targeting foreign readers than those targeting domestic 2 readers, χ (1, N = 823) = 11.79, p = .001. Therefore, for fairness, only H2b was supported. Finally, it is noteworthy to report that non-hypothesized differences associated with fairness were observed. Not only was fairness in Chinese government-controlled newspapers targeting foreign readers featured less often than in those targeting domestic readers, it was also 2 featured less often here than in market-appeal newspapers both in China and in the U.S., χ (1, N 2 = 785) = 45.27, p < .001, χ (1, N = 714) = 20.94, p < .001, respectively. In essence, Chinese 24 government-controlled newspapers targeting foreign readers featured fairness less often than any other condition. 25 DISCUSSION This content analytic study aimed to test predictions that newspapers with (a) different organizational goals, (b) different readership, and (c) a different nation-of-origin would tailor their stories in terms of four moral domains (i.e., care, fairness, loyalty, and authority). The general media environment in China offers a good opportunity to explore MIME-based predictions regarding the manner in which newspapers operating under authoritarian versus libertarian press climates might differ in their portrayal of morality-relevant stories. It also provides an opportunity to systematically examine how market-appeal versus governmentcontrolled newspapers in China differ in their coverage of progressive and conservative topics while attempting to simultaneously satisfy readers and avoid punishment from the Chinese government. Moreover, beyond comparing newspapers operating under different media systems in China, market-appeal newspapers in Chinese versus the U.S. were compared to see if differences in country of origin would influence their reporting on loyalty- and authority- related issues. Overall, the findings produce general support for predictions that authority would be represented more in media produced for audiences in Authoritarian systems than Libertarian systems fairness would be represented more in Libertarian than Authoritarian systems. The results show patterns consistent with some predictions. First, among government-controlled newspapers, those targeting domestic readers were more likely to report stories featuring authority than those targeting foreign readers (H3a supported). Second, the Chinese marketappeal newspapers demonstrated tendency to report stories featuring fairness more often than government-controlled newspapers in China (H2b supported). In addition to these predicted finding, noteworthy patterns unrelated to any hypotheses indicated that a) the market-appeal newspapers tended to report more on authority than the government-controlled newspaper 26 targeting foreign readers; and b) the market-appeal newspapers both in China and in the U.S. tended to report more on fairness than the Chinese government-controlled newspaper targeting foreign readers. Interpretation of these findings begins with a discussion of hypotheses related to authority and loyalty followed those related to care and fairness. Following this, the findings associated with the research question are examined along with a discussion of unpredicted findings related to authority. The section concludes with a discussion of the study’s limitations and directions for future research. Portrayal of Conservative Domains: Authority and Loyalty Support for predictions that among government-controlled newspapers in China, those with a domestic target would feature authority more than those with a foreign target (H3a) can be understood as consistent with MIME-based predictions that target-audience values will shape the representation of moral-domains in media content. It appears that even when newspapers operate under forms of censorship resembling an authoritarian media system, those targeting a foreign audience may mimic libertarian media practices and balance the report of moral domains. In line with this interpretation, fewer stories featuring authority were found in headlines of China Daily than People’s Daily. This understanding is consistent with earlier research examining the MIME’s indication that media producers tailor stories to the values of their target audience. However, whereas previous research has supported this claim for entertaining media content, these findings provide initial evidence that the domains perceived most important by target readers will shape news stories produced for different audiences. Failure to support the prediction that government-controlled newspapers would report stories featuring loyalty more often suggests several possibilities. One possible explanation is simply that MIME’s underlying logic is flawed. Another possibility could be the thriving 27 globalization process which caused the fuzziness in defining ingroup versus outgroup members. The definition of loyalty in MFT suggests that people would sacrifice for their ingroup interests and act against outgroup members. However, as members of different cultures become more integrated into a global community, the lines separating different groups become blurred. This is particularly true for widely accessible newspaper attracting large circulations, such as those used in this study. Portrayal of Progressive Domains: Care and Fairness Support was also found for predictions that among newspapers targeting Chinese readers, headlines in market-appeal newspapers would feature fairness more often than governmentcontrolled newspapers (H2b): Support for this hypothesis can be understood as consistent with previous MIME research suggesting that the market-appeal nature of dramatic conflict is most often built on the care and fairness domains (Tamborini, 2012). Though all newspapers try to attract audiences, the libertarian theory of the press describes media whose central goals include attracting and entertaining audiences in order to profit. In line with this understanding, the findings here that market-appeal newspapers are more likely to feature indicators of the fairness domain in their headlines are consistent with claims that moral domains can be used to enhance forms of dramatic conflict in order to attract audiences. In contrast to support found for predictions regarding fairness representations in newspapers targeting Chinese readers, no significant differences were observed in market-appeal versus government-controlled newspapers for care (H2a). Moreover, among Chinese government-controlled newspapers, no significant differences in care were observed between those targeting foreign versus domestic readers (H4a). 28 One interpretation of these findings is that the logic underlying the differences predicted for care in this study is flawed. As care is an important moral domain, it is reasonable to expect that all types of newspapers may tend to report stories featuring care. However, it is also possible that existing differences in line with logic underlying these predictions exist, but were not made apparent by the coding scheme used in this study. Evidence of this alternative explanation is demonstrated in an examination of the separate positive and negative indicators of care. The coding scheme used in the current study looked only at indicator presence and, as such, combined both positive and negative indicators to form a single measure of care presence. When examined together, no significant differences in care are apparent across newspapers. However, when positive and negative indicators were examined separately, significant differences were observed between market-appeal and government-controlled newspapers. In this additional analysis, headlines in the government-controlled newspapers targeting a domestic audience (People’s Daily) and a foreign audience (China Daily) were combined to form a composite category for government-controlled newspapers; while headlines in the market-appeal newspapers in China (Nanfang Metropolis Daily) and in the U.S. (The Washington Post) were combined to form a composite category for market-appeal newspapers. A chi-square test was applied to these two composite categories to examine the positive versus negative indicators of care. The results are displayed in table 4. These findings demonstrated that negative indicators of care were more common in the two market-appeal newspapers than the two government 2 controlled newspapers, χ (3, N = 119) = 12.95, p < .001. One interpretation of this is that market-appeal newspapers may tend to report more negative indicators of care to attract audiences and sell their papers. This interpretation is consistent with previous research showing that harm (i.e., negative indicator of the care domain) may be a critical predictor of audience 29 exposure (Lachlan, Tamborini, Eden, Bowman, & Grizzard, 2012), According to this study, violations of harm are important in the creation of dramatic conflict, and central to the appeal of entertaining media. The findings in the present study suggest the possibility that the same is true for news. Table 4. Chi Square Frequency and Percent Distributions for Features of Care Newspaper Type Positive N (%) Negative N (%) Total N Government Control People’s Daily & China Daily 69 (62.7%) 41 (37.3%) 110 Market Appeal Nanfang Metropolis Daily & The Washington Post 33 (37.1%) 56 (66.3%) 89 Non-hypothesized Findings on Authority Several non-hypothesized patterns of note were observed in analyses on the authority domain. First, among market-appeal newspapers, those publishing in China (i.e., Nanfang Metropolis Daily) appeared to report stories featuring authority more often than those publishing in the U.S. (i.e., The Washington Post). Even though the difference is not significant, it is still noteworthy when we consider that The Washington Post has a strong political focus in its news selection. Common sense suggests that news focusing on political issues will likely involve many positive or negative indicators of authority. As such, it is surprising to see a Chinese newspaper with a “soft” focus report more stories featuring authority than a U.S. newspaper with a political focus. Additional analyses revealed that the U.S. market-appeal newspaper reported 2 more negative indicators than the Chinese market-appeal paper, χ (1, N = 123) = 9.61, p = .002 (see Table 5). As such, while the authority domain was reported more often overall in the Chinese market-appeal newspaper, “challenge or disrespect for legitimate power” was more 30 common in the U.S. newspaper. This can be understood in light of the general media climate in China. Even after critical revolutions in the China’s newspaper industry, censorship is still ubiquitous and the threat of aftermath punishment still has a “chilling effect.” Therefore, it is plausible to suggest that the so-called “market-appeal” newspapers in China do not have the same level of freedom experienced by their western peers. Table 5. Chi Square Frequency and Percent Distributions for Features of Authority Newspaper Type Positive N (%) Negative N (%) Total N Market-appeal in China Nanfang Metropolis Daily 43 (59.7%) 29 (40.2%) 72 Market-appeal in the U.S. The Washington Post 16 (31.4%) 35 (68.6%) 51 Just as we can explain the surprising finding that the market-appeal newspaper in China reported more on authority than the market-appeal newspaper in the U.S., we can also explain the surprising finding that authority was reported less often in the government-controlled newspaper targeting foreign readers than the market-appeal newspapers both in China and in the U.S. Although it seems illogical at first glance, the finding can be understood to be consistent with an interpretation that Authoritarian systems show greater respect for legitimate power than Libertarian systems. Evidence consistent with this claim can be seen in post hoc analyses examining the positive versus negative indicators of authority found in the Chinese governmentcontrolled newspaper compared to the market-appeal newspapers in China and in the U.S. If logic underlying the Four Theories of the press is accurate, we would expect newspapers operating under an Authoritarian system to report more positive and less negative indicators of authority than those operating under a Libertarian system. Additional post hoc 31 analyses examining the positive and negative indicators of authority were conducted on the previously described composite scores combining the two government-controlled newspapers (People’s Daily and China Daily) and the two market-appeal newspapers (Nanfang Metropolis Daily and The Washington Post). Analyses show that the government-controlled newspapers tend to report more positive indicators and fewer negative indicators than the market-appeal 2 newspapers, χ (3, N = 263) = 46.83, p < .001 (see Table 6). This finding is consistent with claims that newspapers driven by the government control goals of Authoritarian system are likely to show greater respect for legitimate power than newspapers driven by the market appeal goals of Libertarian system. Table 6. Chi Square Frequency and Percent Distributions for Features of Authority Newspaper Type Positive N (%) Negative N (%) Total N Government-Control People’s Daily & China Daily 122 (87.1%) 18 (12.9%) 140 Market-appeal Nanfang Metropolis Daily & The Washington Post 59 (48.0%) 64 (52.0%) 123 Limitations One limitation is the fact that only one newspaper was selected to represent each category. As explained in the method section, this is understandable for the two Chinese government controlled papers, as China Daily is the only Chinese government newspaper targeting foreign readers directly owned by the party, and People’s Daily sets the agenda for all Chinese government newspaper targeting domestic readers. This is not the case for the market appeal newspaper. Nevertheless, the newspapers chosen were not selected without cause. Nanfang Metropolis Newspaper, was selected to represent market-appeal newspapers in China and only 32 The Washington Post was selected to represent market-appeal newspapers in the U.S. These two newspapers were purposively sampled because of their prominence. Nanfang Metropolis Newspaper is one of the earliest emerging market-appeal forces in China and currently most successful one (Huang, 2001). Its influence, especially among the young readers, is so strong that its content is usually compared to the Party newspapers. The Washington Post is a well-known and nationally influential newspaper in the U.S. which is rooted in the political capital of the U.S. This provides us a unique opportunity to compare a serious and slightly politics-oriented U.S. newspaper to a newly emerged “market-appeal” newspaper in China. Differences between these two newspapers may represent true differences between the China and the U.S. cultures. To overcome this limitation, future studies are suggested to include multiple newspapers to determine whether the patterns observed here with one influential newspaper of each type are also found when random sampling is applied. Another limitation of this study deals with the difficulties and challenges of creating a valid scheme to code moral domains in media content. This difficulty is increased when we consider differences in reporting morally-relevant content across different audiences, across media systems, and across different cultures. Notably, this issue highlights the benefits of applying a systematic and theoretical logic in such content analytical studies. The fact that MFT suggests a coherent scheme capable of distinguishing the central features of different moral domains is highly valuable. Instead of attempting to distinguish moral content through exemplars which may vary across cultures, by explicating the underlying features that distinguish universal moral domains, MFT offers a more solid foundation to help identify moral-relevant features in media content. The MIME is relatively new and provides a systematic way to organize research on morality in media. Undoubtedly, more efforts are needed to develop a valid coding scheme 33 that can be applied reliably across various forms of communication. Meanwhile, the high reliabilities of using this coding scheme which were showed in this study suggest promising progresses. Conclusion and Future Directions The present study investigated how newspaper headlines differ in reporting morally relevant content. Analyses compared the representation of four moral domains in newspapers that differed in their organizational goals, target audience, and nation origins. This study attempted to improve on previous content analyses that have examined morally relevant content in media by applying a coding scheme based a coherent theory of morality (MFT) and a model explicating the manner in which media and morality are interrelated (MIME). The results of this analysis are valuable for their ability to provide new insight regarding the manner in which moral domains are represented in news stories produced by different media organizations. Taken as a whole, these findings suggest that (1) Chinese government-controlled newspapers tend to report stories featuring authority more often than market-appeal newspapers in China; (2) within the government-controlled newspapers, those targeting domestic readers tend to report more on authority those targeting foreign readers. Future studies should focus on the valence rather than the mere presence of each moral domain. This type of research will add more to our knowledge of how newspapers differentially frame morally relevant content in reporting for different audiences. For instance, according to the MIME, the extent to which a domain is supported or negated in the content of different media systems an indicator the domain’s salience in the culture for which the content is produced (Tamborini, 2012). Research coding and comparing the domain valence in the news produced for different audiences would be a better test of MIME-based prediction regarding the long-term 34 processes that govern the dynamic reciprocal relationship between media and morality. Furthermore, guided by the MIME, researchers should also explore how difference in the framing of individual moral domains impacts the salience of specific moral domains in audience members. The long-term components of the MIME predict that repeated exposure to content that consistently focuses on and positively frames specific domains will alter the pattern of domain salience found in members of that audience group. Although the research presented in this current study is incapable of testing this part of the MIME, it begins to lay a foundation for future research efforts in this area. . 35 APPENDIX 36 APPENDIX CODING PROTOCOL Introduction Research Goals This content analysis protocol attempts to assess differences in how morality is reflected in headlines from newspapers that vary in terms of (a) organizational goals, (b) targeting audience, (c) nation of origin and five moral domains. Headlines from four newspapers purposely selected to vary along these dimensions will be coded to indicate the presence of positive or negative indicators of each moral domain. Chinese media scholars argue that there are qualitative differences between government-controlled and market-appeal newspapers in China. Although previous studies have proven informative in many ways, research in this area would benefit from the development of a coherent scheme capable of distinguishing the central features along which the content of these papers differs. To date, attempts suffer from the lack of a coherent theoretical that offers a systematic framework to define features of moral content. This current content analysis attempts to overcome this issue by applying the model of intuitive morality and exemplars (MIME: Tamborini, 2011, 2012) to study differences between newspaper content in major Chinese and U.S. newspapers. Conceptual Definitions Organizational goal is defined conceptually here in terms of the motives that govern editorial decisions regarding news content. These motives are categorized as either (a) maintaining government control, or (b) increasing market appeal. Maintaining government control refers to editorial decisions driven by the organization’s desire to maintain social control by imbuing society with ideologies favorable toward the governing party. Increasing market appeal refers to editorial decisions driven by profit motives. 37 Newspaper target is defined as the group of readers form whom the content is tailored. Tailoring content implies that editorial decisions on news content and presentation style are made with the reading habits and preferences of these audience members in mind. Nation of Origin is defined as the nation where the newspaper is published. The moral foundations examined in this study are defined as follows in terms adapted from of the MFT (Haidt & Joseph, 2004). Care: This foundation is developed from mammal evolution, an ability to sense and dislike the emotional or physical pain of others. It is related to the virtues of kindness/gentleness/nurturance. Fairness: This foundation is based on processes of reciprocal altruism, the need for justice and equity. It underlies virtues of justice, proportionality, and equal rights. Ingroup loyalty: This foundation is linked to the benefits of communal living, selfsacrifice for the group. It stresses the importance of group cohesiveness and group harmony. It is activated when an individual feels that it's "one for all, and all for one." Authority: This foundation is derived from hierarchical social interactions, respect for legitimate leadership and traditions. It is active anytime people value social order and is especially common in cultures with high-power distance. Data Collection News headlines from the following four newspapers will be collected: - People's Daily, the largest government-owned newspaper in China, targeting Chinese - Nanfang Metropolis Daily, one of the largest market-appeal newspaper in China, targeting Chinese 38 - China Daily, the only official party-sponsored English newspaper in China, targeting foreigners. - The Washington Post, one of the largest conservative market-appeal newspapers in the U.S. PDF versions of the original newspapers will be retrieved from the official websites for requested dates. Only news stories reported in text will be analyzed, excluding any pictorial presentations, advertising content, editorials, readers' feedback and similar matter. Random sampling will be employed on the three newspapers published in 2011. The random selected sample is expected to represent the population of news reporting in China and in the U.S. in 2011. The estimated number of news stories is 1500 for one constructed week which will be taken by randomly selecting one Monday, one Tuesday and etc. Data Processing Procedures PDF versions of the original newspapers will be downloaded and stored electronically so that coders can easily access the materials repeatedly. Since the PDF versions contain exactly same content as print versions and carry out the same layouts, there should be only limited or even no effects of collecting stories from electronic papers instead of original print. Processing/Preparing of Content To prepare the content for coders, steps below will be followed: 1. Exclude non-news content (e.g., advertising matter, editorial, columns and etc.) 2. Label the headline of news stories based on the issue date (e.g., 1_14_1h for the first story's headline on January 14th issue) 3. Prepared headlines and paragraphs are then ready to be analyzed for moral domains as specified below. 39 Content Variable Operational Definition Organizational goals. Organizational goals were defined operationally as whether the newspaper organization has an editor directly appointed by the government and receives filtered news stories from the government before publishing (government controlled), or operates free of this government control (market appeal). For Chinese newspapers, this is operationally defined as whether this is explicitly stated in their online organization statement. For the one U.S. newspaper used here, the study begins with the presumption that this paper, like all privately owned newspapers in the U.S., operates under an open system with no government regulation on news selection; thus is categorized as a market-appeal newspaper. Target audience. The target audience variable applies only to Chinese newspapers in the government-controlled condition. Target audience is defined operationally in terms of whether or not the goals of government-controlled newspaper stated in their website identify that the paper is designed to attract only a foreign audience. Both People's Daily and China Daily are under direct government regulation and meet the criteria for inclusion in the government-controlled condition. The goals of China Daily, as stated in their website, identify foreign readers as their target group. People’s Daily makes no statement about its target group, but is widely accepted as targeting a domestic readership. Newspaper’s nation of origin. The nation of origin is operationally defined in terms of where the paper was published (China versus U.S.). Four moral domains. Each domain will be coded in terms of indicator presence. Indicator presence is coded as (+1) positive (-1) negative or (99) not present. Operational definitions are given as follows. Indicator Presence 40 Care positive. 1. Compassion: the act of showing a deep awareness of and concern for another's physical and emotional suffering; e.g. a. When watching others suffering, shed tears, cry… b. "Helen, Are you OK? Do you need any help?" c. "Helen, I know you are suffering, I feel your pain." 2. Care: the act of offering or providing physical and emotional assistance. This includes protecting and supporting others, physically and emotionally; e.g. a. Anonymous girl helped the injured out of the damaged car and took them to the nearest hospital b. give help when others need c. Cuddling your friend who was dumped Care negative. 1. Harm: the act of causing another’s physical or emotional pain, or planning to put another in jeopardy of physical or emotional pain; e.g. a. Hit, beat, kick, etc. b. "You are a loser!" 2. Indifference: the act of choosing not to assist another who is suffering physical or emotional pain; e.g. a. No pedestrians stopped to help the injured girl. b. When watching others getting hurt/suffering, walk away or choose not to help. c."You're dumped? Whatever, I'm busy." 41 3. Cruelty: the act of enjoying the pain or distress of others regardless of whether the pain/distress is functional or dysfunctional; or feeling no remorse towards the dysfunctional pain or distress of others; e.g. a. "I love the smell of human blood. I'm excited! It's fun to kill!" b. when watching others suffering, laugh or show happiness in some other ways c. Two boys tortured the kitten for hours just for fun. Fairness positive. 1. Honesty: The act of following an implicitly or explicitly stated, socially accepted agreement at personal cost (e.g., keeping a promise, following the rules, telling the truth…); e.g. a. He kept his promise. b. "I will never lie for money!" c. reveal some truth at personal cost 2. Procedural justice: The act of using a transparent, societally accepted rule-based procedure of decision making (e.g., the legal system, or agreed upon mediator) to resolve conflict; e.g. a. Go to Court b. "Let the judge pass judgment!" c. Appealing to the law d. Decisions based on legal procedures 3. Distributive justice: Fairness in the distribution of rights or resources a. Equality: Acting in a manner such that the benefits people receive is based on equally meeting everybody’s needs; e.g. 42 i. Regardless of social status, all kids should have the right of receiving appropriate education ii. Equal rights under the law b. Equity: Acting in a manner such that the benefits people receive is proportional to their contribution; e.g. i. The boss rewards hardworking employees. ii. Everybody gets their fair share 4. Retributive justice: The act of punishing a wrong doer at a level that is accepted by the society as commensurate to the wrong doing; e.g. a. This criminal who killed a girl deserves severe penalty. b. The boy who broke the window had to pay for it. c. An eye for an eye, tooth for tooth d. The child who got caught smoking cigarettes was grounded. Fairness negative. 1. Dishonesty: The act of violating an implicitly or explicitly stated, socially accepted agreement (e.g., telling a lie…) e.g. a. Susan called in sick even though she felt fine. b. Break promises/lie c. Susan knows that two tons of steel will be enough but she tells a lie to her manager so that she may earn extra money 2. Procedural injustice: The act of violating a transparent, societally accepted rule-based procedure of decision making (e.g., the legal system, or agreed upon mediator) to resolve conflict; e.g. 43 a. Violating the rules b. The decision was supposed to be made by the community but the boss insisted to decide on himself 3. Distributive injustice: Unfairness in the distribution of rights or resources a. Inequality/favoritism: acting in a manner such that the benefits provided meet some individual or group’s needs more than another individual or group’s needs; e.g. i. "Serve that couple first. They must be movie stars." b. Inequity: acting in a manner such that the benefits people receive is not proportional to their contribution. e.g. i. Students from capital cities receive extra 20 points for the entrance exam. 4. Retributive injustice: the act of punishing a wrong doer at a level that is unacceptable (too much or too little) by the society standards as commensurate to the wrong doing e.g. a. The guy murdered his wife but only got 6 months in jail b. The boy ate ice cream against his father's wishes and was beaten severely Ingroup loyalty positive. 1. Solidarity: Showing solidarity/ cohesiveness/ harmony towards a specific group (e.g., nation, team, gender, race, class, etc); e.g.: a. Following and supporting your group’s decision (even if it is in conflict with other goals) b. Seeking societal harmony c. Singing a school fight song/national anthem 44 d. Wearing clothes that show the logo or colors of a group e. Verbally defending or getting angry at attacks on your group 2. Martyrdom: Sacrifice of any individual interests for the group; e.g.: a. The firefighter injured when saving the public goods. b. Going to war for your country c. Fighting your families’ battles 3. Bigotry: suspiciousness or bias against outgroup; e.g.: a. Verbally, physically or symbolically attacking outgroup members because they are outgroup members b. Mary works as a waitress in a restaurant. She claims that the food from another restaurant tastes bad even though most people think the two restaurants are similar. c. The Smith hated their next-door neighbors. Anytime something went missing or broken, they blamed on them. Ingroup loyalty negative. 1. Departure: giving up ingroup membership a. The basketball player quit the team 2. Betrayal: behaviors that sacrifice or have the potential to sacrifice ingroup interests for the self or other groups’ benefits; e.g. a. The employees sold the company’s secrets to the competitor. b. The lead singer left the band behind when he was offered a contract to be a solo performer. 45 3. Treason: behaviors that sacrifice or have the potential to sacrifice the interests of one’s nation for the self or another nation’s benefits; e.g. a. The soldier sold the nation’s secrets to another nation. Authority positive. 1. Obedience/Deference: acts showing respect for the legitimate power of a leader or any group or institution to which the person belongs (political, religious, social, business, educational, etc.); e.g. a. Verbally praise the leader b. Following the orders of leaders 2. Traditionalism: acts showing respect for traditional customs and social norms of any group or institution to which the person belongs; e.g., a. Firework shows on July 4th. b. The Jewish man covered his head when he entered the temple. Authority negative. 1. Disobedience: Acts showing disrespect for the legitimate power of any group or institution to which the person belongs (political, religious, social, business, educational, etc.); e.g., a. Question, criticize or challenge religious beliefs. b. People protest against the government. 2. Defiance: Acts showing disrespect for the leaders of any group or institution to which the person belongs; e.g. a. Question, criticize or challenge the president b. People protest against the mayor 46 3. Progressivism: acts showing disrespect for traditional customs and social norms of any group or institution to which the person belongs; e.g. We should abandon arranged marriage Coder Selection/Training Coders will be trained to understand the moral domains from MFT and to be capable of coding various references to individual domains. Literature will be provided for group reading projects at the beginning of training from which coders will learn basic knowledge of the theoretical concepts and become familiar with the conceptual definitions of each moral domain. Coder training proceeds in the following steps: 1) Relevant literature reading 2) Protocol Read-through 3) Collective training with real examples 4) Coding ten sample stories independently 5) Discussing disagreements 6) Adjusting coding instructions 7) Repetition of steps 2-4 until collective agreement is reached Coding Procedures Coders must code stories independently under quiet and none-disturbing conditions. Eating, watching videos/TV, surfing the Internet, listening to music or similar behavior will be prohibited while coding. Coders will be asked to review the protocol every time before the actual coding task and check the operational definitions frequently if needed. Coding results will be recorded by excel files with numbers of news stories clearly listed. Each coding period should not exceed 2 continuous hours or 3 hours with 15 minutes rest in between. 47 Reliability and Validity Assessment Reliability and Validity assessment will be undertaken for each moral domain: Reliability Assessment Example headlines will be randomly selected from the real news reporting to assess reliability. Coders will be given 50 example headlines each time to code independently and results will be recorded with coders' names and numbering of headlines. Scott's Pi will be calculated for each category based on the coding results with expectation of achieving .75 or above. Simple agreement percentage will also be reported. Validity Assessment Face validity is established by using common logical presumption and the moral foundations dictionary provided by MFT (Haidt & Joseph, 2004). The operational definitions of each foundation logically reflect the conceptual definitions. For example, behaviors supporting the concept "authority" include actions like honoring the norms and traditions of a society and faithfully following the legitimate leadership. There are other ways of operationally defining the concepts but the current operational definitions should be acceptable to scholars and easily seen to be reflecting the conceptual definitions. For example, "equal treatment of citizens" or "appealing for equal education opportunities" logically fall under the category "Fairness" while "killing abandoned dogs" or "refusing help the wounded" logically reflect the concept of "Care." However, concurrent validity is difficult to establish since there has been little research done on morality in content analysis. Similar measures and replicable standards are rarely found in current literature. 48 REFERENCES 49 REFERENCES Ai, F. (1996) Successful Experiment of Critical Reports in the “West China City News”. Press Circles, 4, 17-18, 20. Benson, R., & Hallin, D. C. (2007). How states, markets and globalization shape the news. The French and US national press, 1965-97. European Journal of Communication, 22(1), 27-48 Bowman, N. D., Lewis, R., & Bryand, M. (2012). The morality of May 2, 2011: A content analysis of US headlines regarding the death of Osama bin Laden. Paper presented at the 62nd annual meeting of the International Communication Association, Phoenix. Chan J. (1993). Commercialization without independence: trends and tensions of media development in China. In J. Cheng & M. Brosseau (Eds.), China Review 1993 (pp. 25012521). Hong Kong: Chinese University Press. Chan, J. (1995). Calling the tune without paying the piper: the reassertion of media controls in China. In C. K. Lo, S. Pepper & K. Y. Tsui (Eds.), China review 1995 (pp. 5.1-5.16). Hong Kong: Chinese University Press. Chen, H. L., & Chan, J. (1998). Bird-caged press freedom in China. In J. Cheng (Ed.), China in the post-Deng era (pp. 645-667). Hong Kong: Chinese University Press. Fico, F., & Cote, W. (2002). Partisan and structural balance of election stories on the 1998 governor’s race in Michigan. Mass Communication & Society, 5(2), 165-182. Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and Conservatives Rely on Different Sets of Moral Foundations. Journal of Personality and Social Psychology, 96, 1029-1046. Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, K., & Ditto, P. (2010). Mapping the moral domain. Journal of Personality and Social Psychology, 101, 366-385. Groseclose, T., & Milyo, J. (2005). A social-science perspective on media bias. Critical Review, 17(3-4), 305-314. Haidt, J. and Joseph, C. (2004). Intuitive ethics: How innately prepared intuitions generate culturally variable virtues. Daedalus, 133, 55-66. Haidt, J., Koller, S. H., & Dias, M. G. (1993).Affect, culture, and morality, or is it wrong to eat your dog? Journal of Personality and Social Psychology, 65, 613-628. He, Z. (2000). Chinese Communist Party press in a tug-of-war: a political-economy analysis of the Shenzhen Special Zone Daily. In Chin-Chuan Lee (Ed.). Power, money, and media: communication patterns and bureaucratic control in cultural china (pp.112-151). Evanston, IL: Northwestern University Press. 50 Huang, C. J. (2000). The development of a semi-independent press in post-mao China: an overview and a case study of Chengdu Business News. Journalism Studies, 1(4), 649-664. Huang, C. (2001). China's state-run tabloids: The rise of "city newspapers". Gazette, 63(5), 435450. Jiang, S. (2005) Ideological Influences and News Content Variation in Chinese Newspapers. A paper submitted for International Communication Association Conference. Lachlan, K. A., Tamborini, R., Eden, A., Bowman, N. D., & Grizzard, M. (2012). The influence of morality subcultures on the acceptance and appeal of violence. Journal of Communication, 62(1), 136-157. Li, B.Y. & Jia, Y.H. (1999) My Market is Just Next To Your Door: The Rise of City Newspapers. News Front 2. Lind, R. A. (1996). Care and Justice in Audience Evaluations of Ethics in TV News. Journal of Mass Media Ethics, 11, 82-94. Luther, C. A. & Zhou, X. (2005) Within the Boundaries of Politics: News Framing of SARS in China and the United States. Journalism & Mass Communication Quarterly 82(4), 857－ 872. Mastro, D., Enriquez, M., Bowman, N. D., Prabhu, S. & Tamborini, R. (2012) Morality subcultures and media production: How Hollywood minds the morals of its audience. In R. Tamborini (Ed), Media and the moral mind. London, UK: Routledge, 75-92. Pickle, K., Quinn, S. C., & Brown, J. D. (2002). HIV/AIDS coverage in black newspapers, 19911996: Implications for health communication and health education. Journal of Health Communication, 7(5), 427-444. Reinecke, L., Tamborini, R., Grizzard, M., Lewis, R., Eden, A., & Bowman, N. D. (2012). Characterizing mood management as need satisfaction: The effects of intrinsic needs on selective exposure and mood repair. Journal of Communication, 62, 437-453. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68-78. Semetko, H. A., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news.Journal of Communication, 50(2), 93-109. Shoemaker, P. J., & Reese, S. D. (1996). Mediating the message: theories of influences on mass media content (2nd ed.), White Plains, NY: Longman. Shweder, R.A., Much, N. C., Mahapatra, M., & Park, L. (1997). The “big three” of morality 51 (autonomy, community, divinity) and the “big three” explanations of suffering. In A. Brandt & P. Rozin (Eds.), Morality and Health (pp. 119-169). New York: Routledge. Siebert, F. S., Peterson, T., & Schramm, W. (1956). Four theories of the press: The Authoritarian, Libertarian, social responsibility, and soviet communist concepts of what the press should be and do University of Illinois Press. Song, J., Shao, G., & Wu, J. (2010). Responding to regulatory, market, and information technology changes: Chinese newspapers in the transitional era. Global Media Journal, 9(16). Tamborini, R. (2011). Moral intuition and media entertainment. Journal of Media Psychology: Theories, Methods, and Applications, 23, 39-45. Tamborini, R. (2012). Model of intuitive morality and exemplars. In R. Tamborini (Ed), Media and the moral mind. London, UK: Routledge, 43-74. Tamborini, R., Enriquez, M., Lewis, R. J., Grizzard, M., & Mastro, D. (2011, May). A content analysis of moral foundations presented in Spanish and English language soap operas. Paper presented at the 61st annual meeting of the International Communication Association, Boston. Xi, W.J. (1995) New Thinking of New Evening Newspapers. Press Circles, 5,31-33. Yang, W.P. (1999) Socialist Popular Newspapers: On the Nature of City Newspapers. News Front, 6. Zhao, Y. Z. (1998). Media, market, and democracy in China: between the Party line and bottom line. Urbana: University of Illinois Press. Zhao, Y. Z. (2000). From commercialization to conglomeration: the transformation of the Chinese press within the orbit of the Party state. Journal of Communication, 50 (2), 3-26. Zhang, E., & Fleming, K. (2005). Examination of characteristics of news media under censorship: A content analysis of selected Chinese newspapers SARS coverage. Asian Journal of Communication, 15(3), 319-339 Zillmann, D., & Knobloch, S. (2001). Emotional reactions to narratives about the fortunes of personae in the news theater. Poetics, 29, 189-206. Zillmann, D., Taylor, K., & Lewis, K. (1998). News as nonfiction theater: How dispositions towards the public cast of characters affect reactions. Journal of Broadcasting and Electronic Media, 402, 153-169. Zillmann, D. (2002). Exemplification theory of media influence. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (2nd ed., pp. 19-41). Mahwah, NJ: Lawrence Erlbaum. 52