1 EMERGING POWERS OF INFLUENCE: THE RISE OF NEW MEDIA IN CHINA By F ANGFANG GAO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR O F PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Fangfang Gao
3 To my dear parents and beloved homeland China
4 ACKNOWLEDGMENTS I have many people to thank for making this milestone possible. I would first like to acknowledge my advisor and th e chair of this dissertation, Dr. Cory Armstrong, for her support and guidance throughout the doctoral program. She was such a dedicated advisor who led me through the whole process and helped me gain confidence as an academic researcher. I could not ask f or a better mentor and chair. I am fortunate enough to have an excellent committee. Professor Melinda McAdams opened the door of new media study for me and enlightened me on this interesting research area. Her endless aspirations toward life and research motivate my future life. Dr. Kim Walsh Childers directed my voyage in the realm of mass communication research and generously shared her wealth of knowledge with me. She sincerely understood my struggles as an international student and helped me to overcom e different barriers in pursuing my graduate study Dr. Ben DeVane challenged me to think about the intersection of the digital technology and mass communication. The ir help, insights, and patience were invaluable. I wish to thank all of them and many supp ortive professors, staff, friends, and colleagues during my time at the University of Florida. Finally, I am grateful for the support of my family. M y parents, although thousands of miles away, provided me immense support that I cannot even find proper words to express my appreciation. With their deep faith in their only daughter, I would continue challenging myself and liv e a confident life.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES ........................................................................................................ 10 ABSTRACT ................................................................................................................... 11 CHAPTER 1 INTRODUCTION .................................................................................................... 13 Media in China ........................................................................................................ 13 Tainted Milk Formula Scandals ............................................................................... 16 Need for Current Study ........................................................................................... 20 Theoretical Framework ........................................................................................... 22 Theoretical and Practical Contributions .................................................................. 26 Purpose and Overview of the Study ....................................................................... 29 2 LITERATURE REVIEW .......................................................................................... 31 AgendaSetting Theory ........................................................................................... 31 Evolution of AgendaSetting Theory ................................................................. 31 SecondLevel Agenda Setting and Framing ..................................................... 36 Inter media Agenda Setting .............................................................................. 40 Citizen Journalism .................................................................................................. 43 Development of Citizen Journalism .................................................................. 43 Public Sphere and Civic Participation ............................................................... 50 Public Forums .................................................................................................. 53 Characteristics o f online public forums ...................................................... 54 Patterns of participation ............................................................................. 56 Blogs ................................................................................................................ 59 Development of blogs ................................................................................ 59 Hyperlinks and multimedia ......................................................................... 61 The Influence of Citizen Journalism ................................................................. 63 Media in China ........................................................................................................ 65 Traditional Media in China ................................................................................ 65 Development of the Internet in China ............................................................... 67 New Media in China ......................................................................................... 69 Public forums in China ............................................................................... 70 Blogs in China ............................................................................................ 72 Food Safety Issues in Media Coverage .................................................................. 72 Story Focus and Frames .................................................................................. 74
6 Tones and Sources .......................................................................................... 77 Synthesis ................................................................................................................ 78 The Present Study .................................................................................................. 80 Conceptual Defini tions ..................................................................................... 81 Research Questions ......................................................................................... 82 3 METHODS .............................................................................................................. 87 Content Analysis ..................................................................................................... 87 Sampling ................................................................................................................. 88 Newspaper Sampling ....................................................................................... 88 Forum Sampling ............................................................................................... 93 Blog Sampling .................................................................................................. 94 Variables ................................................................................................................. 97 Intercoder Reliability ............................................................................................. 103 4 RESULTS ............................................................................................................. 104 Story Focus ........................................................................................................... 105 Frame ................................................................................................................... 109 Tone...................................................................................................................... 111 Source .................................................................................................................. 113 Pattern of Participation .......................................................................................... 115 Multimedia and Hyperlinks .................................................................................... 125 Influence of Author ................................................................................................ 127 5 DISCUSSION AND CONCLUSIONS .................................................................... 130 Coverage of the Tainted Formula Scandals .......................................................... 131 Story Focus and Frame .................................................................................. 131 Tone and Source ............................................................................................ 135 Influence of Gender and Blogger Type ........................................................... 138 The Growth of New Media .................................................................................... 142 Par ticipation .................................................................................................... 142 Multimedia and Hyperlinks ............................................................................. 147 Implications ........................................................................................................... 148 Th eoretical Implications .................................................................................. 148 Practical Implications ...................................................................................... 150 Limitations ............................................................................................................. 153 Future Research ................................................................................................... 155 APPENDIX A NEWSPAPERS AND ONLINE FORUMS FOR STUDY ........................................ 158 B RESULTS OF INTERCODER RELIABILITY ........................................................ 159 C CODEBOOK FOR NEWSPAPERS ...................................................................... 161
7 D CODEBOOK FOR ONLINE FORUM POSTS ....................................................... 164 E CO DEBOOK FOR BLOG POSTS ........................................................................ 168 LIST OF REFERENCES ............................................................................................. 172 BIOGRAPHICAL SKETCH .......................................................................................... 189
8 LIST OF TABLES Table page 1 1 Major food safety incidents in China during the first decade of the 21st century ................................................................................................................ 17 3 1 Rotated component mat rix for factor analysis of story foci in the coverage of the tainted formula incidents ............................................................................. 107 3 2 Story focus in the coverage of the tainted formula incidents ............................ 107 3 3 Major story focus in the coverage of the tainted formula incidents ................... 108 3 4 Frames in the coverage of the tainted formula incidents .................................. 109 3 5 Dominant frame in the coverage of the tainted formula incidents ..................... 110 3 6 Tones in the newspaper and blog coverage of the tainted formula incidents ... 112 3 7 Tones in the coverage of the tainted formula incidents in online forums .......... 112 3 8 One way ANOVA examining sources used in the coverage of the tainted formula incidents .............................................................................................. 113 3 9 Independent sample T test of story focus and reader participation in blogs ..... 117 3 10 Independent sample T test of major focus and reader participation in blogs .... 118 3 12 Independent sample T test of dominant frame and reader participation in blogs ................................................................................................................. 121 3 13 Independent sample T test of story focus and reader participation in forums .. 123 3 14 Independent sample T test of major focus and reader participation in forums 124 3 15 Responsiveness of discussion in blogs and online forums ............................... 125 3 16 Homogeneity of discussion in bl ogs and online forums .................................... 125 3 17 Multimedia and hyperlinks in blogs and online forums ..................................... 126 3 18 Gender differences of authors in the newspaper coverage of tainted formula incidents ........................................................................................................... 128 3 19 Gender differences in the blog coverage of tainted formula incidents .............. 128 3 20 Influence of blogger type in the coverage of tainted formula incidents ............. 129
9 A 1 Chinese newspapers selected for analysis ....................................................... 158 A 2 Chinese online public forums selected for analysis .......................................... 158 B 1 Intercoder reliability for each variable (Krippendorff's Alpha) ............................ 159
10 LIST OF FIGURES Figur e page 2 1 Proposed framework for study ............................................................................ 81 D 1 A n example thread in online discussion forums in China. ................................ 165
11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EMERGING POWERS OF INFLUENCE: THE RISE OF NEW MEDIA IN CHINA By Fangfang Gao December 2011 Chair: Cory Armstrong Co chair: Kim B. Walsh Childers Major: Mass communication China is one of the fastest growing Internet markets in the world. The Internet has been changing peoples lives and rev olutionizing the media landscape in China. New media platforms online, such as blogs and public forums, have been built up and adopted by millions of people. E xposure to new technology and devices has made computers, Internet, and mobile phones indispensable parts of life for many people, especially young people, in China. Compared with traditional media such as newspapers, radio and television, the Internet and Internet related new media platforms bring a fairly different media experiences to Chinese peopl e, creating new channels for people to get information and enabling them to express, organize, and influence public opinion in unprecedented ways. By watching, reading, and talking to each other, Chinese Internet users generate collective intelligence and build public discourse that is different from the traditional media discourse initiated by the government. This new phenomenon in Chinas online sphere could usher in a new era of public engagement and a more transparent and democratic society.
12 Based on t he agendasetting theory and the concepts of new media and citizen journalism, this study investigated the emerging aspects of new media and citizen journalism in China from an agendabuilding perspective, examining the theoretical significance and practic al implications of new media and citizen journalism in China. Specifically, this study focused on how Internet users have been using online public forums and blogs to express and exchange their opinions during a public health crisis the tainted milk formul a scandals explor ing the ways in which citizen journalism and new media manifested themselves at a time period full of uncertainty and public tension. To better understand the roles and characteristics of new media in the Chinese society, this content ana lysis of newspaper, online forum, and blog coverage of the tainted milk formula scandals examined the differences between the old and new media platforms, analyzing the discrepancies between public discourse from new media and government discourse from tra ditional mainstream media. Further, by examining the inter media agendasetting effects between new media and traditional mainstream media, the study investigated how citizen media and traditional mainstream media may influence each other in the ways they covered the public health crisis. The theoretical and practical implications of the findings are discussed.
13 CHAPTER 1 INTRODUCTION The rapid growth of the Internet has changed the media landscape in China by bringing in multiple new media platforms suc h as blogs, public forums, and micro blogs. Unlike traditional media in China, new media platforms allow people to express their opinions and hear from others more easily and freely. The rise of citizen engagement in information production and dissemination creates a new realm for grassroots public discourse, providing broader implications for the flow of free information in Chinas controlled media environment. The tainted milk formula scandals from 2008 to 2011, constituted a public health crisis that at tracted attention from all over the world and was widely covered and discussed in all Chinese media platforms including mainstream newspapers, blogs and online forums This case provides a good opportunity to look at the growing influence of new media in C hina. By examining the differences and correlations between the public discourse generated by new media and the government discourse shown in mainstream newspapers, the findings will shed light on how the media landscape is changing especially in the development of new media and citizen journalism, in mainland China. Media in China The past three decades have witnessed tremendous social, political, cultural and underlying ideological changes in the Peoples Republic of China, also referred to as mainland C hina (Feng & Wu, 2009). Since the late 1970s, the country has been more open to the outside world, putting the priority on economic reforms. The policy changes have led to a transition from a command
14 economy, in which government was the leading economic power, to a market oriented economy with an emphasis on a free market or laissez faire policy. The market orientation, accompanied by political decentralization with more local autonomy, has prompted the emergence of a civil society and an incremental democr atization in China (Huang & Chen, 2009; Yu, 2002). The economic and political reforms have also contributed to changes in the external and internal environment of Chinese media. In the 1980s, the Chinese Communist Party (CCP) decided to allow newspapers, m agazines, television, and radio stations ( except major party newspapers and party affiliated magazines, TV and radio) to be commercialized and dependent on advertising revenues instead of being financed exclusively by the government (Press Reference, 2010) This requires and enables media in mainland China to transform from traditionally government own ed businesses into relatively independent market entities, although all of the news media are still owned by the state. Print media, including newspapers and magazines, are still under the regulations of the General Administration of Press and Publication of the Peoples Republic of China. Moreover, the change of policies on media industries in China created a fundamental change in audiences and the advertising market. Mainstream media, including newspapers, have been driven to attract more readers and advertisers. Less strictly controlled by the government than party newspapers, metro newspapers have begun to pay more attention in their news reporting to what average citizens rather than government authorities are interested in (Huang & Chen, 2009; Yu, 2002). In addition,
15 challenged by the rapid development of new information technology, Chinese media have started the process of digitalization. For example, more and more newspapers began to set up newspaper websites, to utilize the Internet for source finding, interviews, internal information sharing, copy editing, etc. The development of the Internet in China has been rapid with a dramatic expansion in the num ber of Internet users. Internet based media, such as blogs, online forums, social networking sites, and microblogging, being the driving force for the development of citizen media in China, have become more and more popular among Chinese Internet users (C NNIC, 2010). Traditional media have been paying more attention to online citizen media and the voices from the public, citing more user generated content. Government authorities have begun to feel the impact of the emerging citizen media. During the taint ed milk formula scandals more information and opinions about the scandals were released online by the public. In response, government authorities had to provide answers/solutions to the questions or criticism from citizens. This path of information flow i s revolutionary. Previously in China, with only the existence of traditional newspapers, radio and television, information flowed from the government through mainstream media to the public. There was limited control of information by the public. However, w ith the help of new media like public forums and blogs, informed people can get their voices out and find people who agree or disagree with them. For the first time, a public sphere independent of the government has emerged. C ivic conversations in the soci ety are facilitated. Information can flow from the public through new media to
16 mainstream media organizations and the government, enabling twoway communication instead of the government dictating communication. As a result, Chinese citizens may now have s ome influence on policy making, forcing the government to respond to social problems exposed online and take actions. In other words, such contraflow of information from people to the government could improve the transparency of the society and speed up t he development of democracy in mainland China (Chang, 2008; Yang, 2003). As the online landscape continues to evolve and mature, not only journalism practice but also the public sphere in China is gradually evolving into a new era. New media have facilitat ed more voices and have given citizens more power in the society (Gillmor, 2004). The existence of the new digital platforms has made it possible for the appearance of a Chinese public sphere. Further growth of influence is on the horizon. Tainted Milk Formula Scandals Food safety is now a growing public health concern around the world. In China, a series of food safety incidents during the first decade of the 21st century killed many Chinese citizens and degraded the international image of the country (C han & Lai, 2009; Xu, 2009) ( Table 11). Among all of the food safety scandals the tainted milk formula scandals probably received the most attention worldwide. In 2008, m elamine, a nitrogenrich industrial chemical used in plastics and fertilizers and banned in food production, was found in t he Sanlu infant formula during an inspection by the government after media exposure of the potential problem. The inspection showed that many Chinese dairy companies put melamine in their products to thicken the watered down milk and boost the
17 apparent protein content to fool standard tests, cutting production cost and thus increasing their profit s (Chan & Lai, 2009; Jacob, 2008; Liu, 2010). Infants who drink such melaminecontaminated formula for sustained periods coul d develop kidney stones and suffer from kidney failure (Zhao & Lim, 2008). Table 11. Major food safety incidents in China during the first decade of the 21st century Year Food Safety Incidents in China 2003 Poisonous ham. Pesticide Dichlorvos was used by some producers in Jinhua ham production to prevent spoilage and insect infestation. 2004 Counterfeit baby formula. Around 70 babies died of malnourishment. More than 100 other babies in Anhui Province suffered from malnutrition; Counterfeit alcoholic drinks. Industrial alcohol was blended with rice wine in alcoholic drinks. 4 men died of alcohol poisoning and 8 other men were hospitalized; 2005 Sudan I Red Dye, a chemical banned from food manufacturing in China since 1996, was found to be illegally added by some food companies; Excess iodine in Nestle milk powder, which may lead to pathological changes of the thyroid gland; Hygiene problem with Haagen Dazs products in several provinces. 2006 Pesticide residue on vegetables; Use of prohibited ch emicals ( i.e. carcinogenic nitrofuran metabolites) and drugs ( i. e. malachite green) in fishery; Counterfeit drugs ( i. e. fake Armilarisni A). Disinfectant, the chemical diglycol, etc. were found in several drugs. 2007 Alleged carcinogen (oil filtering p owder magnesium trisilicate) used in frying oil for KFC products; Melamine contaminated wheat gluten and rice protein in pet food for export. 2008 Infant milk powder formula tainted by melamine. Over 50,000 children were affected. 2009 Illegal reuse o f m aterial from the 2008 adulterations in dairy products; Leather protein was found being used in milk supplies including milk powder and other processed dairy products. 2010 Shengyuan milk formula was suspected for causing child precoci ous puberty 2011 Melamine was found again in some brands of infant formula products Later in 2009 and 2010, materials from the 2008 adulterations were found being reused illegally in dairy products. Moreover, in 2009, some dairy products,
18 inc l uding milk, milk formula, et c., were found to have been polluted by a carcinogenic protein derived from leather. Later in 2010, after drinking Shengyuan milk formula, many children began to develop symptoms of precocious puberty such as breast development and high level s of estrogen. It was suspected that Shengyuan formula was the cause of the problem, although the companies denied responsibility Earlier in 2011, melamine was found again in some infant formula products in a nationwide investigation initialized by the government. Medi a, especially new media, played an i mportant role in revealing these threats to public health. On May 20, 2008, a Chinese citizen, Yuanping Wang, posted on Tianya.cn (a popular online forum and probably Chinas biggest online public forum without attachmen t to any major news agency) that his daughters urine had beco me turbid with granules because she had been drinking the Sanlu infant formula (Chang, 2008). He made a plea for public attention regarding the tainted formula. There were more than 3,000 responses to Wangs original post. From that time, the public, especially Internet users, began to pay attention to the incident. The story did not appear in traditional media until September 11, 2008, when the Oriental Morning Post reported that Sanlu formulas one of Chinas oldest and most popular infant formulawas the cause of kidney stones for many infants around the country (Chang, 2008). Many newspapers, local and national, then started to follow up (Chang, 2008; Xin, 2010). At the same time, a nationwide investigation by the central government revealed a link between the
19 unusual outbreak of kidney stones among babies and infant formula from Sanlu (Liu, 2010). After that, the tainted Sanlu formula incident quickly unfolded to be a nationwide scandal and yet another international embarrassment. What made people even angrier was that the later nationwide inspections found that most dairy companies had the same problem illegally adding melamine to their formulas (Zhao & Lim, 2008). According to the Xinhua news agency, the biggest and most authoritative news agency in China, contaminated formula had caused at least six infant deaths, and nearly 300,000 children had been sickened after consuming contaminated formula by December 2008 ( Jacob, 2008). After the initi al exposure of the health threat citizen journalists and Internet users started reporting and discussing the formula scandals more widely (Xin, 2010). Mobilized Chinese Internet users tried to search online for more details, relentlessly exposing those who were involved in the scandals (Chang, 2008). For example, some found that Chinas Central Television (CCTV), the most authoritative TV station in China, had broadcast a program praising Sanlus quality control process on September 2, 2008. Although the v ideo was quickly deleted from the TV websites archive after the scandals some people found the programs script and posted it online, questioning the credibility of the media (Chang, 2008). Outraged people studied the website of the General Administration of Quality Supervision, Inspection and Quarantine (AQSIQ), Chinas major quality control bureau and watchdog organization on product safety issues, finding posts from parents complaining about Sanlus infant
20 formula as early as June 30, 2008 (Chang, 2008; Ma, 2008). Most complaints from parents were ignored, as Sanlu had been exempted from quality inspections by the AQSIQ until the exposure of the scandals in September (Chang, 2008). The 2008 Sanlu tainted formula scandal and the following tainted formul a incidents ( such as the reuse of materials from the 2008 adulterations in dairy products, Shengyuan estrogencontaminated milk formula, the leather polluted milk and the melamine tainted infant formula in 2009, 2010, and 2011; see Table 1 1) resulted in peoples loss of trust in dairy products made in China. Millions of Chinese Internet users posted on online discussion forums and blogs to blame the dairy companies, criticizing the government for l ax supervision. The Internet users consistent efforts inf luenced the resignation of Changjiang Li, the head of the AQSIQ forcing the government to set up new regulations for dairy products (Chang, 2008; Ma, 2008). Need for Current Study To some extent, Chinese Internet users prompt and indepth investigation i nto the formula scandals overshadowed the work of many journalists from traditional media (Chang, 2008). Online forums and blogs have become freer places for Chinese people to voice their anger, lodge complaints and expose problems, especially regarding public crises such as the milk formula scandals In other words, new media provide important platforms for public participation in civic life in China. People can perform a watchdog role when the press fails to do so. Citizen journalism is increasing, exerti ng more influence on the Chinese media landscape and gradually breaking the hegemony of traditional media. The emerging influence of new media, together with the rise of citizen engagement in
21 information production and dissemination in China recently, has been a salient issue and an under researched topic. This topic is significant and important for several reasons. First, the impact of new media and citizen journalism is clear in China. What happened during the milk formula scandals was a good case in point. The growth of the Internet opens a public sphere in which grassroots public discourse can take place, which is nearly impossible with only the traditional mainstream media under the close direction of the government. Instead of attempting to reform the existing media, Chinese Internet users can use blogs, public forums, and other new media tools to go around the government controlled media and find their own audience, hence getting people connected and informed in new ways. As a result, although the stat e control wont go away easily, it may be eroded gradually. Second, f or China, a politically Communist and culturally traditional country, the renovation brought about by th e latest new media technologies is theoretically significant and practically meani ngful. Last, understanding how the digital landscape is changing and why it is changing in China can help researchers not only be aware of the current trends in Chinese media but also envision the future of the political, economic, cultural, technological, and media aspects of Chinese society. Health communication studies have confirmed that print and broadcast media remain the publics primary source of health information (Major, 2009; Singer & Endreny, 1993), although traditional media have been criticized sometimes as contributing to the larger problems in society, such as
22 misperception of public affairs issues, lack of civic discussion, and manipulation of the public mind (Dong, Chang, & Chen, 2008). While analyzing the development of new media, it is necessary to take the role of traditional media into consideration, especially in the media environment in China, where traditional media with support from the government still hold the biggest share of the market, while the Internet is facing censorship tha t could prevent new media from becoming fully alternative media. In addition, when the society is full of anxiety and uncertainty on food safety issues such as during the tainted formula incidents traditional media can be very important actor s in shaping the public discourse online and setting the public agenda because they have a huge and established audience base as well as established authority. However, few studies have analyzed in depth the role of new media in comparison with traditional newspapers in China, examining the differences in old and new media when covering public health crises like the formula scandals in China, revealing the agendasetting process es of each medium and inter media agendasetting effect This study will be among the first to explore this issue. Theoretical Framework Agendasetting theory has been a fundamental theory in mass communication research, providing a well established theoretical framework for researchers to understand the powerful influence and the pervasive role of the media, and to analyze the formation and transfer of issues from media agendas to public agen das. Generally speaking, agendasetting theory argues that media select certain issues and report them frequently and prominently, which leads the public to perceive those issues as more important than other issues of the day
23 (McCombs, 2005). For the past four decades, the core assumption of agendasetting theory has been the transfer of issue saliencehow the filtering and emphasis of certain issues in media can influence the perceived importance of issues among the public (McCombs & Shaw, 1972). Hundreds of studies have confirmed the effects of media coverage on the formation and development of public discourse (McCombs, 2004). Basically, agenda setting conc erns three categories: 1) media agenda setting : issues discussed in media (newspapers, television, radio, Internet), factors such as media type, source, etc. that are influential in the selection of issues, how the issues become important in media; 2) public agenda setting : issues discussed as personally relevant to people and how the issues become the publics priorities; and 3) policy agenda setting : the legislative agenda of policymaking bodies and how the issues become important to policy makers (Kosic ki, 1993; Wallington, Blake, Taylor Clark, & Viswanath, 2010, p. 77). The three categories are distinct but related. In other words, media agenda setting, public agenda setting, and policy agenda setting are different in that they each represent one specif ic aspect of the society. Meanwhile, they are also tightly intertwined, as media agenda, public agenda, and policy agenda develop in concert with each other (Kosicki, 1993). For example, there is inter media agenda setting between mass media like newspaper s and television (i.e. Roberts & McCombs, 1994), and between the media agenda represented by newspapers and the public agenda represented by online forums or blogs (i.e. Roberts & McCombs, 1994; Roberts, Wanta, & Dzwo, 2002; Lee, Lancendorfer,
24 & Lee, 200 5). In addition, different media have different potential in setting agendas due to differences in their content, audience, penetration, etc. In mass communication research, both first level agenda setting and second level agenda setting have been studied (McCombs & Shaw, 1972; Kiousis & McCombs, 2004). First level agenda setting is the core theoretical notion in traditional agendasetting research, focusing on the amount and prominence of issue coverage. It has been established in first level agendasettin g studies that the medias concern with issues in the news can lead to an increase in the publics concern with those issues (Kiousis & McCombs, 2004), suggesting that the media can decide which issues the public will be aware of. In contrast to the tradit ional emphasis in agenda setting research, secondlevel agenda setting links the agendasetting process to framing, examining the influence of attribute saliencethe properties, qualities, or characteristics that describe objects in the coverageand the tone of those attributes (Ghanem, 1997). Although originally developed for the studies of politics, the agendasetting framework has been applied to many different areas, including health communication (Roger, 1996; Wallington, Blake, Taylor Clark, & Viswana th, 2010), sports journalism (Seltzer & Dittmore, 2009) and others. The importance of the agenda setting effects of mass media on health issues in the society has been well documented (Roger s, 1996). McCombs (2005) demonstrated that whether the basic agenda setting effects of news media continue in much the same fashion as the previous decades or eventually disappear because of the
25 changing media landscape, measuring these effects will remain high on the research agenda for at least the near term (p. 546) Given the appropriateness and adequacy of the agendasetting theory, the competing agenda environment of new and traditional media in China would appear to be a natural new setting for the evaluation and validation of the efficacy of agendasetting theor y by clarifying the roles and interactions of mainstream newspapers, bloggers, online forum discussants in the construction of discourses surrounding a public health crisis, predicting the future media landscape in mainland China. However, there is weaknes s in agendasetting theory. The notion of agenda setting derives more or less from a mass society perspective, acknowledging the powerful role of media in the society and implying a direction of causal influence from media to audience (Baran & Davis, 2009) Audiences differences in perception and understanding were of less of a concern in traditional agenda setti ng research (Baran & Davis, 2009). Moreover, the emergence of the Internet changed the communication landscape, challenging the defining features of mass communication since early 20th century including mass production, lack of individual control, and a limit on the number of channels (Chaffee & Metzger, 2001). Instead, the Internet is decentralized, with myriad new channels and content options avai lable. Peoples ability to choose, produce and deliver content is enhanced greatly. Blogs, citizen media, and social network sites, representing different ideological, commercial and personal niches, have changed peoples communication behaviors across the world and opened vast new territories to
26 communication researchers for the agenda setting role of new media (Chaffee & Metzger, 2001). Theoretical and Practical Contributions This study holds the potential to make a variety of important contributions to t he theory and practice of mass communication, new media research, and Chinese media research. By applying agendasetting theory to both Internet based media and traditional mainstream media in China, the study will lend insights into a practical situation and extend its boundary to a different cultural context. Because academic scholarship about new media in China is still in its infancy, it is suitable to apply agendasetting theory to a new medium that continues to grow in scope and popularity. Through ex amination of the coverage of a public health crisis in mainstream newspapers, online forums, and blogs from an agendasetting perspective, this content analysis will show how the old and new media set agendas for the public and in what ways, exploring how the public agenda came into being, examining the influence the old and new media agenda had on the public agenda and vice versa. Therefore, this investigation not only builds and extends agendasetting theory to new media platforms in a country with a dif ferent me dia system but also contributes to the assessment and progress of the theory. An imperative criterion in assessing the usefulness of a social science theory is to determine its theoretical scope of generality (Shoemaker, Tankard, & Lasorsa, 2004) If a theory is high in scope, it can be applied to various situations and settings. The value of the agendasetting theory in explaining the influence of traditional news media on audiences by transferring the media agenda and issue salience to the
27 publi c agenda in the Western media systems has been well established. The setting for the current study old and new media tools such as blogs and online forums in Chinaprovide a compelling new situation for testing the theoretical scope and assumptions of agendasetting theory. The uniqueness of the study also lies in the contrast between the old and new media and the examination of inter media agenda setting between the two medium types during a time of public health crisis. The comparison helps us not only ex plore the interplay of the old and new media but also enhances the understanding of each medium, demonstrating whether and how the different media from which people are getting their information and building their opinions are indeed different. From a theoretical perspective, adding inter media agenda setting into the current study allows for more comprehensive examination of the theory. Moreover, this study reinforces the explication of the agenda setting process to include both object and attribute salience. In other words, it examines first and secondlevel agenda setting. As previously noted, there is a need for more scholarship on new media development in China. This explanatory and predictive study helps contribute to the overall body of knowledge in new media research and Chinese media research, while introducing an established theoretical framework into the two research areas. With the hope that this issue specific quantitative study can shed light onto a broader ground, this study can help to explai n the implications of the interplay between the media systems and social structure in China today. Based on agendasetting and framing research with roots in the broader mass
28 communication field, and built from previous literature on health communication a nd international/intercultural communication, this research takes a multidisciplinary approach and incorporates perspectives from Chinese society. Additionally, this study holds the potential to make practical contributions. This research addresses an area of growing importance in the media landscape in China, analyzing the unfolding political and social implications of the Internet and new media development, a nd envisioning the short term and long term future of news and communication in China. More speci fically, this research helps to gain insights into some issues related to the Internet in China f or example, the governments intention to control information versus the publics struggle to speak freely to reveal social problems and express dissatisfaction with the government. Also, by examining the issue salience, framing styles, tones, participation patterns, sources citations, and use of hyperlinks/ multimedia in mainstream newspapers, online forums, and blogs during the milk formula scandals this study provides an overview of communication strategies and tactics used by media pr actitioners and Internet users, demonstrating their different goals as well as their different ways of thinking and communicating. While evaluating the communication strategies and tactics, this research offers suggestions to the Chinese government, media, and the public on how to adapt to the new information environment especially on how to build crisis response strategies that can cope effectively with public health crises spec ific to Ch inese society in the digital age.
29 Purpose and Overview of the Study The main purpose of the study is to apply agendasetting theory in examining the discourses surrounding the milk formula scandals from 2008 to 2011, exploring the influence of th e emerging new media and citizen journalism in mainland China, hence predicting future development of the media landscape in China. The focus of the study is on two distinctive discourses : 1) public discourse resulting from the citizen participation in inf ormation/opinion production/dissemination and civic discussion on new media platforms, including online forums and blogs in China; and 2) government discourse disseminated by the mainstream Chinese newspapers. More specifically, this quantitative content a nalysis emphasizes : 1) the formulascandal related content posted in online public forums and blogs by citizens, which represents the public discourse developed within the online public sphere; and 2) the formulascandal related content released by the mai nstream newspapers, which represents the government discourse. Examining what the two discourses were and how they came into being throughout the exposure and development of a public health crisis helps us to better understand the roles of different media in China and to construct comparisons and correlations between the old and new media. For discussions about the milk formula scandals in Chinese online public forums, the story focus tone, sources, size of participation, and patterns of discussion are an alyzed. For formulascandal related blog posts written by Chinese Internet users, including health experts, parents, average citizens, or journalists, the story focus frames, tone, sources, participation patterns and the use of hyperlinks and multimedia a re analyzed. In comparison, the story focus
30 frames, tone, and sources in mainstream Chi nese newspaper coverage of the tainted formula scandals are also examined, so as to gain insights into the roles and potential differences of media in China. In the fol lowing chapters, Chapter 2 examines the theoretical constructions and literature review in five sections: agendasetting theory, citizen journalism, media in China, food safety issues in media coverage, and synthesis. Based on the previous studies, researc h questions are proposed at the end of Chapter 2. Chapter 3 covers the research methodcontent analysis used for the study. Chapter 4 reports the statistical results for each research question. Chapter 5 analyzes the findings, discussing the theoretical contributions, practical implications and the limitations of current study, and suggesting directions for future research.
31 CHAPTER 2 LITERATURE REVIEW Agenda S etting Theory Evolution of Agenda S etting Theory Cohen (1963) pointed out that mass media may not be successful much of the time in telling people what to think, but it is stunningly successful in telling its readers what to think about (p. 13). Agendasetting theory holds that the news media can set the agenda for public thought and discussion through their choices of what stories to consider newsworthy and how much prominence and space to give the stories (McCombs & Reynolds, 2002, p. 1). The core concept of the theory is the transfer of the issue salience from the media agenda to the public agen da (McCombs, 2004; McCombs & Ghanem, 2001; McCombs & Shaw, 1993; McCombs, Shaw, & Weaver, 1997). Nearly a century ago, the intellectual forefather of agendasetting theory, Walter Lippmann (1922) argued that mass media can paint the pictures in our heads and serve as a conduit to the world outside (p. 4). The role of the press is like the beam of a searchlight that moves restlessly about, bringing one episode and then another out of the darkness into vision (Lippmann, 1922, p. 364). He realized that, rather than from direct experience, the public normally learns about news or issues from secondhand sources such as the press; the news and issues chosen by the press often then become the issues with the highest importance in peoples perception (Lippmann, 1922). Lippmanns extrapolations were later tested by Max McCombs and Don Shaw in their empirical study (McCombs & Shaw, 1972). After McCombs and Shaws ground-
3 2 breaking study of the role of the media in the 1968 presidential campaign in Chapel Hill, No rth Carolin a the agendasetting theory was first introduced in 1972 (McCombs & Shaw, 1972). McCombs and Shaw (1972) hypothesized an agendasetting process from the mass media coverage of a political campaign, in which media emphasis (i.e. frequency of co verage, length, and placement) were disseminated to the public with a rough ranking (agenda) of what the important issues were over time. Theoretically, the media agenda would be known and internalized by the public until the public priority ranking of iss ues mirrored the media emphasis on issues (McCombs & Shaw, 1972). McCombs and Shaw (1972) tested the hypothesis with a two pronged study: They first asked undecided voters what they thought the most important issues in the campaign were, then conducted a content analy sis of the campaign coverage in nine state, local, and national media. The study found a strongly positive correlation between the media agenda and the public agenda, supporting the hypothesized media effect on what issues people think are important (McCombs & Shaw, 1972). In other words, issue salience in public agenda is greatly influenced by the amount of media coverage of the issues ( McCombs & Shaw, 1972). Since the seminal study of agenda setting by McCombs and Shaw, hundreds of empirical studies of the theory have delved into its different dimensions. The vast majority found support for the influence of media on peoples perceived importance of issues. For example, in a study of public opinion trends during the 1960s, there was a correlati on of .78 between news
33 coverage and peoples perceptions of the most important issues of the time (Funkhouse, 1973). When examining agenda setting for civil rights issues between 1954 and 1976, a correlation of .71 between the media content and public opinion was revealed (Winter & Eyal, 1981). More recently, it was found that print and broadcast media coverage of the economy may have serious consequences on consumer economic evaluations and expectations, particularly when the coverage is negative (Hester & Gibson, 2003). Based on a national poll and a content analysis of network newscasts, it was found that the more media coverage a foreign nation received, the more likely people in the United States were to think the nation was vitally important to U.S. interests (Wanta, Golan, & Lee, 2004). For online media, Ku, Kaid, and Pfau (2003) found that there were interactions between political websites, news content, and public opinion, demonstrating the significant agendabuilding function of the Internet in shaping and determining the issues most salient to the general public. Agendasetting effects have also been evident outside the United States. Through comparisons of six major concerns on the public agenda with local news coverage in Spain, a high degree of c orrespondence was revealed between patterns of news coverage and the public agenda about local problems (Canel, Llams, & Rey, 1996). Based on a panel survey in Sweden, Shehata (2010) found that the agendasetting effects of media were present because the percentage of people naming unemployment as the most important political issue increased by 17% as it dominated news coverage during the 2006 Swedish National Election campaign. In Japan, agendasetting effects of media on national policy making
34 and policy adoption for smoking control were observed (Sato, 2003). Overall, when generalizing the agendasetting theory to different contexts, most researchers have confirmed that topics emphasized in news coverage and the amount of coverage given to the topics in news media play an important role in transferring the media agenda to the public agenda (McCombs & Ghanem, 2001; McCombs & Reynolds, 2002). Thus, a lack of attention given to a topic by the media decreases the likelihood that the topic will be on the public agenda (Wanta & Wu, 1992). In the process of agenda setting, from media agenda to public agenda, there are factors that can influence the agendasetting effects the need for orientation and obtrusiveness of issues (Miller, 2005). People vary in their need for orientation. Because higher levels of interest and uncertainty produce higher levels of need for orientation, people who are more interested and uncertain about the issues are most likely to be influenced by media (Miller, 2005). It also has been proposed that agenda setting effects should be strongest for unobtrusive issues, with which the audience has not had direct experience, because they must rely on media for information about those issues (Miller, 2005). Gandy (1982) suggested going beyond agenda setting to determine who sets the media agenda, how and for what purpose it is set, and with what impact on the distribution of power and values in society (p. 266). Some researchers have gone beyond the relationship between media and public agendas, adding another component the policy agendainto research (Dearing & Rogers, 1996; Edwards & Wood, 1999). The policy agenda is defined
35 as those issues to which political institutions (i.e. government, the president) give serious consideration and devote subs tantial attention (Edwards & Wood, 1999). Researchers have argued that political institutions policy agenda is inherently responsive to media coverage (Edwards & Wood, 1999), and is intertwined with both the media agenda and the public agenda (Dearing & R ogers, 1996). In China, where all the media are statedowned and the media content is controlled by the government, the media agenda is almost the same as the policy agenda or government agenda. There are several dimensions of agenda setting, including fir st level agenda setting, secondlevel agenda setting, and inter media agenda setting. First level agenda setting is the most traditionally studied level, which posit s that, intentionally or not, media use objects or issues to influence the public. Further, each of the objects/issues on an agenda has numerous attributes, which are the characteristics and properties that describe the object/issue. Both the selection of objects for attention and the selection of attributes to describe those objects are powerful agendasetting tools (Bryant & Oliver, 2008). So, beyond the agenda of objects/issues, or first level agenda setting, there is secondlevel agenda setting, which focuses on the attributes or characteristics of the objects/issues. Unlike first level and s econdlevel agenda setting, which both emphasize salience transfer from the media to the public, inter media agenda setting stresses salience transfer among different media. So far, only first level agenda setting has been discussed in this section. The following two sections will discuss second level agenda setting and inter media agenda setting in depth.
36 SecondLevel Agenda Setting and Framing Merging traditional first level agendasetting research with framing research, secondlevel agendasetting resear ch shifts the focus from the mere transfer of issue salience to include the transfer of attribute salience (Kiousi s & McCombs, 2004; McCombs, 1994). It is necessary to distinguish between first level and secondlevel agenda setting. First level agenda sett ing is the contribution of overall media to public assessment of issue importance (what to think about), while secondlevel agenda setting concerns the way an issue is framed in the media (how to think about it) (McCombs, 2004). For example, first leve l agenda setting is concerned about how the amount of media coverage on a political campaign can influence peoples awareness of the campaign, while second level agenda setting emphasizes how the media coverage of certain attributes of the political campai gn can affect peoples perception of the campaign. Therefore, secondlevel agenda setting is also referred to as attribute agenda setting, describing how media link attributes such as properties, qualities, characteristics, and tones to objects or issues, which consequently influences the agenda of issue attributes in the public (Ghanem, 1997; Wanta, Golan, & Lee, 2004). Numerous studies have found support for the assumptions of secondlevel agendasetting theory. During the Spanish national elections, McCombs, Lla mas, Lopez Escobar and Rey (1997) found that media coverage of the attributes of political candidates was highly correlated with the publics ranking of the important attributes of candidates. Examining media coverage of George W. Bush and John McC ain during the 2000 Republican primary in New Hampshire,
37 Golan and Wanta (2001) suggested that media agenda of candidate issues and attributes was significantly associated with voters evaluations of the candidates. In an experimental study of candidate attributes, Kiousis, Bantimaroudis and Ban (1999) found subjects impressions of candidate personality traits mirrored media portrayals of those traits. Craft and Wanta (2004) found that the two attributes that received most extensive news coverage in media coincided with the two attributes of the September 11 terrorist attacks that news consumers were most concerned about. Studying the attributes of foreign nations, specifically how negatively and positively the foreign nations were portrayed in media, it w as revealed that the more negative coverage a nation received, the more likely people in the United States were to think negatively about the nation (Wanta, Golan, & Lee, 2004). Taking both first and secondlevel agenda setting into consideration, T akeshi ta and Mikami (1995) found significant evidence for the transfer of both issue salience and attribute salience from the media to the public in Japan. Examining how issues and attributes of issues are presented by media in second level agenda setting links the agenda setting theory to the notions of framing (Kiousis, Mitrook, Wu, & Seltzer, 2006; McCombs, 1997; McCombs & Ghanem, 2001). Framing is a central organizing idea for making sense of a relevant event, suggesting what is at issue, showing persistent patterns of cognition, interpretation, and presentation (Gamson & Modigliani, 1989, p. 3). Entman (1993) explained that to frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to
38 pr omote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described (p. 52). Translating it using agendasetting terms, framing is the selection of a restricted number of thematically rel ated attributes for inclusion in the media agenda (McCombs, 1997, p. 6); to frame is to describe issue attributes that characterizes the defining features of the issue (McCombs, Lopez Escobar, & Llamas, 2000). Therefore, topic selection and framing work t ogether to create agendasetting effects the frequency with which a topic is mentioned probably has a more powerful influence than any particular framing mechanism (a focal point of news presentation), but framing mechanisms could serve as catalysts to fr equency in terms of agendasetting (Ghanem, 1997, p. 12). In other words, while repeatedly mentioned topics may influence public opinion, the way how these repeated topics are outlined is also influential. For example, newspaper coverage of tainted milk scandals described the same particular scenes in many stories depicting thousands of babies were sick because of the tainted milk formula. One might assume that such coverage from many newspapers nationwide encouraged rescue/relief responses from the government and dairy companies. There are two dimensions of secondlevel agenda setting research: substantive and affective attributes. Substantive attributes involve information about qualities of newsmakers, while affective attributes involve positive, neutral, and negative qualities, which link to the positive, negative or neutral tones of media coverage in framing research (McCombs, Llamas, Lopez Escobar, & Rey, 199 7 ). McCombs, Lopez Escobar, and Llamas (2000) found that public
39 perceptions of substantive and affective attributes were closely related to media coverage of those attributes; people linked their positive or negative evaluations of political candidates to media coverage. Who sets media agenda, or the sources of media agenda, is another important i ssue to look at in agendasetting research. The selection of sources by news media is important, because sources have the power to shape the news reports and influence the media agenda (Berkowitz, 1987; Messner & Distaso, 2008; Salwen, 1995; Weaver & Elliott, 1985). Scholars in the early 1980s began to ask who sets the medias agenda ( Bryant & Oliver, 2008). Among the various factors that shape the media agenda, source is one of the key factors. Zoch and VanSlyke Turk (1998) stated that the choice of sources can influence medias construction of reality: news is not necessarily what happens, but what a news source says has happened (p. 763). It is necessary to examine the sources that set the media agenda, which in turn can set the public agenda. McCombs ( 2005) pointed out that the sources of the media agenda have become increasingly important in a news environment that is dramatically changing through the Internet. Not only individual sources such as politicians, public officials, public relations practit ioners and other individuals, but also organizational sources such as the government, health agencies, and commercial companies influence the process of agenda setting. For example, examination of T he New York Times and T he Washington Post during the 1960s and 1970s showed that almost one fifth of their news stories were based on press releases and press conferences
40 (Bryant & Oliver, 2008). A special organizational source is the media organization. The interactions and influence of various media on each other, a phenomenon called inter media agenda setting, plays an important role in the ultimate shaping of the media and public agenda (McCombs, 2004). Inter media Agenda Setting Inter media agenda setting was first noticed in the standardization of news stori es among newspapers using wire services (Roberts & McCombs, 1994; White, 1949). Gold and Simmons (1965) reported influences on patterns of coverage in Iowa newspapers by the Associated Press wire service. Later, the agendasetting effects of leading newspapers on other newspapers and television networks were revealed. Reese and Danielian (1989) found that T he New York Times played an initial agendasetting role for other newspapers and television networks when covering the drug issue during 1985 and 1986. T rumbo (1995) demonstrated that elite newspapers had a significant agendasetting effect on the national television networks concerning the coverage of global warming from 1985 to 1992. Roberts and McCombs (1994) found that local television coverage reflect ed the issue agenda of local newspapers during the 1990 gubernatorial campaign. Extending to the new media environment, inter media agendasetting was also evident. Online citizen media challenges as well as supplements the established media as the sole holder of truths (Bryant & Oliver, 2008; Meraz, 2009). Yu and Aikat (2006) found convergent agendas between online news sites, including Google News and Yahoo News, and the traditional print and broadcast media such as T he New York Times Washington Post CN N, and
41 MSNBC. In South Korea, it was also revealed that the leading online newspaper exerted influence on the issue agendas of both other online newspapers and the online wire service (Lim, 2006). The agendasetting role of traditional media on online bul letin boards was noted. Examining the online media coverage of selected issues from five news outlets during the 1996 fall political campaign, Roberts, Wanta and Dzwo (2002) found that discussion of the selected issues in the online bulletin boards was cor related with news media coverage, with time lags varying from one day to seven days, showing that news media have been setting the agenda for bulletin board discussions, which to some extent served as a surrogate for the public agenda. A similar finding of an inter media agendasetting effect from newspapers to online bulletin boards was observed in South Korea, too (Lee, Lancendorfer, & Lee, 2005). Studies demonstrated that the agendasetting role is assumed by not only mass media but also by opinion leaders in interpersonal networks; for instance, the agendasetting effects of mass media on the public agenda could be reinforced by interpersonal communication (Yang & Stone, 2003; Wanta & Wu, 1992), as happens in online bulletin boards where messages from me dia are filtered and strengthened through interpersonal discussions. For blogs, they are not only influenced by traditional media agendas but also able to exert influence on mainstream medias agendas. Using the CBS 60 Minutes segment about George W. Bush s military service in Texas Air National Guard as a case study, Dylko and Kosicki (2006) argued that traditional journalists relied on information from blogs; blogs therefore were able to push
42 stories onto the mainstream medias agenda. Farrell and Drezner (2008) found that top political blogs can socially construct agendas that act as focal points for mainstream media. During the 2004 presidential election, Cornfield, Carson, Kalis and Simon (2005) found that weblogs can create an intense buzz on certain i ssues, which are transferred into traditional media. Through a comparative content analysis of selected national blogs and news outlets, Lee (2006) found that there was a positive correlation between rank ordered issues in both the blog and media agendas. Messner and Distaso (2008) discovered that mainstream newspapers have been legitimizing blogs as sources while being major sources for blogs. Consequently, a news source cycle between the traditional media and blogs was created, and information passed back and forth from medium to medium (Messner & Distaso, 2008). In addition, it was revealed that in top independent political blogs and elite newsroom blogs, traditional elite medias agenda setting monopoly power was diluted by online media such as Yahoo New s, MSNBC and citizen media like blogs, YouTube, etc. (Meraz, 2009). Generally speaking, mass media can and do influence the perceived importance of issues among the public, providing people with information to use in their online discussions and stimulati ng interest in topics among the public. Internet users take agendasetting one step further t hey bring the topics that interest them from mass media to online discussions in online forums, citizen blogs, social networking sites, hence making the issue more prominent on the agenda of the society (Roberts, Wanta, & Dzwo, 2002).
43 Citizen Journalism Development of Citizen Journalism Citizen journalism also referred to as grassroots journalism, is defined as the act of a citizen, or group of citizens, playing an active role in the process of collecting, reporting, analyzing and disseminating news and information (Bowman & Willis, 2003, p. 9). It is a philosophy of journalism and a set of practices that are embedded within the everyday lives of citizens, and media content that is both driven and produced by those people (Atton, 2003). The roots of citizen journalism can be traced back to the time of Ben Franklin, Thomas Paine and the anonymous authors of the Federalist Papers in the 18th Century. At that tim e, Ben Franklins Pennsylvania Gazette was civicminded; pamphleteers such as Thomas Paine printed their own publications to disseminate what they called common sense (Gillmor, 2004). Like citizen journalism today, the anonymous authors of the Federalist P apers analyzed the proposed Constitution and argued the fundamental questions of how the new republic might work for the public, which became essentially a powerful conversation that helped make a nation (Gillmor, 2004, p. 2). In recent history, citizen contributions have been an important part that cannot be neglected in media. There are letters to the editor and oped articles written by nonjournalists in newspapers; the rise of talk radio and call in shows on TV gave average people opportunities to share their opinions with a large audience (Glaser, 2006). In the late 1980s, the advent of desktop publishing made it possible for people to design and print their own publications, although the distribution was still limited (Glaser, 2006). The arrival of the World Wide Web in the 1990s removed the
44 bottleneck of distribution, making it free or inexpensive to publish to wide audience; the subsequent development of new technologies allows people to use mailing lists, forums, personal home pages, blogs, etc., to share their thoughts with the world (Gillmor, 2004; Glaser, 2006). Individuals can capture and share personal experiences or issues perceived to be in the public interest with others without following rules of traditional media or paying a price for their activities (Moyo, 2009). This gives rise to crowdsourcing and participatory journalism. Beginning from the early 1990s, the idea of getting citizens involved in the newsgathering and news production process grew from isolated instances to a more concer ted effort by media organizations (Muthukumaraswamy, 2010, p. 49). Since then, previously passive audience has increasingly contributed to news production and dissemination throughout the world. According to the Pew Center for Civic Journalism, some form of civic journalism was practiced by at least 20% of U.S. newspapers in almost every state and every region from 1994 to 2000 (Friedland & Nichols, 2002). In the 2000s more and more newspaper or TV websites have been opened up to reader comments or emails from readers to the journalist of a bylined story, allowing readers to react to, criticize, praise or add to whats published by professional journalists (Outing, 2005; Project for Excellence in Journalism, 2007). Many newspapers have begun online forums in which readers can participate and incorporated reader blogs into professional news websites; some newspapers have started journalists/editors blogs to share the inner working of the newsroom with readers to demonstrate transparency and have a public
45 d ialogue with readers (Outing, 2005). Such traditional media attached forums and blogs increase the interactivity of the mass media overall by widening opportunities for reader to reader and reader to media communication (Schultz, 2000). Some newspapers hav e encouraged readers to contribute or collaborate with reporters, doing socalled opensource reporting, in which readers who are knowledgeable on the topic were asked to contribute their expertise, provide guidance to the reporters, or even do actual r eporting which will be included in the final journalistic product (Outing, 2005). In addition, many newspapers or TV stations have started to use work from citizens to add to their media content. For example, mainstream media sites run by BBC and MSNBC ac cepted photos and video taken by ordinary citizens with their camera phones for the reporting of the London bombings on July 7, 2005 (Glaser, 2006). Today, we have iReport on CNN, iCaught on ABC, uReport on Fox, and FirstPerson on MSNBC. Such practices ar e termed crowdsourc ed journalism collaborative journalism, networked journalism, or participatory journalism. C rowd sourc ed journalism refers to outsourc ing the journalism work traditionally performed by professional journalists or news institutions to an undefined, generally large group of people in the form of an open call (Muthukumaraswamy, 2010). Networked journalism, similar to collaborative journalism, means that professional and amateur journalists work together, linking to each other to share fact s, questions, answers, ideas and perspectives in news reporting (Beckett, 2008; Shirky, 2009). In networked journalism, the process of news making instead of the final news products becomes even more important (Beckett, 2008).
46 News is no longer an authorit ative political institution. Journalists are actually sharing their ability to tell news with people outside the news institutions (Robinson, 2009). Similarly, participatory journalism allow s users to contribute to journalistic processes with technologies available, giving them greater agency or authority to influence news making (Singer, Domingo, Heinonen, Hermida, & Paulussen, 2011). Nevertheless, how journalist and user generated content become combined on news sites differs greatly (Mitchelstein & Bocz kowski, 2009; Thurman, 2008). Yet, the division between news production and news consumption in citizen journalism is converging (Mitchelstein & Boczkowski, 2009). Generally speaking, citizen journalism, participatory journalism, and crowdsourced journali sm refer to the idea that people without professional journalism training can use the tools of new technologies (i.e. digital cameras, videophones, etc.) and the global distribution of the Internet (i.e. public forums, blogs, etc.) to create, augment or fact check media on their own or in collaboration with others (Glaser, 2006). However, there are nuances among them. In the classic form of crowd sourced journalism, people recount not much more than their daily observations ( Muthukumaraswamy, 2010) With th e focus of crowdsourced journalism on collective intelligence, expertise may or may not be required for it. Participatory journalism emphasizes the collaborations between the public and the journalists In participatory journalism, although i ndividual cit izens are allowed to perform some of the communication functions previously controlled by media,
47 journalists still have higher level of control in news production (Domingo, Quandt, Heinone, Paulussen, Singer, & Vujnovic, 2008). The classic form of citizen journalism means that people do what professional journalists do, investigating and reporting originally on public issues on platfor ms such as: 1) discussion groups including mailing lists, newsgroups, bulletin boards, online forums, and chat rooms; 2) weblogs; 3) collaborative publishing such as Wiki and Slashdot; 4) peer to peer applications such as Instant Messaging (IM), Short Message Service (SMS) and social networking sites; 5) legacy media sites launched by newspaper/broadcast organizations as places where users dominate the content, which are different from their official news websites dominated by the work of professional journalists; 6) independent citizen journalism sites owned by citizens and content created by individual citizens ( Bowman & Will is, 2003; Schaffer, 2007). T he practice of such citizen journalism emphasizes first person, eyewitness accounts by participants; a reworking of the populist approaches of tabloid newspapers to recover a radical popular style of reporting; collective and anti hierarchical forms of organization which eschew demarcation and specialization (Atton, 2003, p. 267), challenging the professionalized and institutionalized practice of mainstream media. One feature of citizen journalism is that even average people can capture news and distribute it globally on their own. Apparently, the Internet is the most important vehicle for this. There are some milestones in the development of independent citizen media online, including the advent of personal websites, the launch of open news publishing systems (i.e. Indymedia, Ohmy N ews, Wikinews,
48 etc.), the arrival of peer to peer applications (i.e. Flickr YouTube, Digg, Twitter, etc.), and the introduction of some popular online tools (i.e. online forums, blogs, etc.). Additionally even without the Internet people can still contribute to news gathering and reporting with technologies such as cell phones (i.e. during the Uganda election and Haiti earthquake) ( Goldstein & Rotich, 2008; Heppler, 2010) The major characteristic of the opennews publishing systems is the free and collaborative nature of news production. Instead of being locked into the standard paradigm of news production and viewership, online users can acti vely engage in the citizen news sites that allow for open participation. Indymedia, OhmyN ews and Wikinews are good cases in point. Having their debut after the World Trade Organization (WTO) meeting protests in Seattle in 1999, Independent Media Centers, popularly referred to as Indymedia, are grassroots news websites where anyone can share photos, text and video with other activists and the world (Muthukumaraswamy, 2010). Another pioneering effor t i n citizen journalism is OhmyN ews, which was launched in 2000 in South Korea (Glaser, 2006). With the motto every citizen is a reporter, Ohmy N ews is a popular news site dominated by readers, as the majority of news stories there were written and submitted by audience member (Glaser, 2006). Being part of Wikipedia, the famous online encyclopedia, Wikinews is a citizen journalist site, start in 2004, that allows anyone with Internet access to publish and edit journalistic content directly on its site without prior authorization or registration,
49 relying on the collaboration of people to correct mistakes and guarantee accuracy (Thorsen, 2008, p. 935). The popularity of peer to peer applications such as Flickr and YouTube add new aspects photos and videoto ci tizen journalism. It is not always possible for professional journalists to be where news events take place. However, with the help of cameras, video cameras, and cell phones, ordinary citizens can bring to the publics attention user generated multimedia content on issues that otherwise may go unreported (Moyo, 2009). Begun in February 2004, Flickr is a Yahooowned photosharing website that is evolving significantly. Started in 2005, YouTube now hosts millions of user generated video and video blogging, i n which users express their opinions about current events, politics, and entertainment. Some have recognized YouTube as nothing less than a cultural, sociological, and economic transformation (Garfield, 2006, p. 224). Other applications such as podcasting, Twitter, etc., also provide ways for citizens to deliver content they create or redeliver content from others (Gillmor, 2010). Digg and Newsvine are interesting metajournalism sites as an alternative model of citizen journalism, in which users post news stories from external websites including mainstream news websites, blogs, and amateur content (Goode, 2009). Similar to YouTube, they aggregate and rank user generated and user selected content, defining news and journalism in an innovative way. These c itizen journalism sites provide models for emerging news production, consumption and participation practice, helping to reconfigure the public sphere (Goode, 2009).
50 Publishing a blog or dialogues in online public forums are examples of self expression and means of establishing ones identities in a broad sociocultural context ( Bryant & Oliver, 2008). Moreover, the blogging or online discussion activities have added avenues for public discourse that have proven to be effective in influencing the political, media, and social landscape in Western society from time to time ( Bryant & Oliver, 2008). Online public forums and blogs are the focus of this study. Their development and characteristics will be examined later. Overall, no matter what form it takes, citiz en journalism or citizen media can play a strong role in public life, reviving and magnifying the public sphere. Public Sphere and Civic Participation Originally conceived by Habermas (1989), the public sphere is described as a realm of social life that ex ists between the private sphere and the state sphere in which people come together to discuss public issues and form public opinion. Drawing on Habermas (1989), Schudson (2003) argued that the press by itself is not democracy and does not create democracy (p. 198). Rather, journalism was recognized as opening up and providing a public sphere in which an issue can be debated, and as a communal force (Schudson, 2003, p. 212). Historically, there are strong ties among public sphere, communities, and journalism (Dailey, Demo, & Spillman, 2008). However, the idea of the public sphere is no longer limited to the traditional media such as newspapers, television, or radio today. It has been extended to the Internet, because the Internet is a potential public sph ere where debate and discourse take place with equality and openness (Hurrell, 2005; Li, 2010). More specifically, Brants (2005)
51 argued that the interactivity embedded in Internet based communication helps to foster dialogues and deliberations, which are t he cornerstones of a well functioning public sphere. Despite voices saying that the Internet may discourage people from civic engagement as television does because peoples attention, especially youths attention, can be easily sidetracked to entertainment on media (Putnam, 2000), most people believe that the Internet can provide a sphere to foster public discussion so as to create largescale public awareness of problems in the society and promote civic participation to work on solutions (Dahlberg, 2001; Kwak, Williams, Wang, & Lee, 2005; Zhou, Chan, & Peng, 2008). Three groups were identified in such online communication: 1) communitarians who promote shared values and community spirit; 2) liberal individualists who facilitate the expression of individual interests; 3) the deliberative group who expand the public sphere based on rational critical discourse (Dahlberg, 2001). Xenos and Moy (2007) found there were direct effects of Internet use on information acquisition and use; whats more, contingent on levels of political interest, there was also a relationship between Internet use and contemporary civic or political engagement. Bryant and Oliver (2008) revealed that online information seeking and interactive civic conversation were the two major reasons w hy individuals sought civic participation through the Internet. Shah, Cho, Eveland, and Kwak (2005) demonstrated that the Internet, as a source of political information and a sphere for public expression, complemented traditional media to foster political discussion and civic messaging in the society. Featuring a high level of
52 interactivity between participants, blogs and online public forums show signs of being virtual communities with network connections to other Web pages, blogs, or online forums (Bryant & Oliver, 2008). In fact, citizen journalism as in blogs or online forums has an advantage over traditional journalism in that it can enhance public participation and foster civic engagement. Through the facilitated expression and conversation, online public forums and blogs expand and renew the concept of public sphere (Gillmor, 2010). Previously, without the existence of the Internet and Internet facilitated platforms, the public sphere was weak in China due to the fact that 1) there was limited free sp ace in traditional media or other public forums that allowed for articulation of social issues and sharing of information; 2) there were almost no civic organizations that can could as a routinized social base against state power for citizens; and 3) organized public protest is under state control (Yang, 2003). With the emerging online venues where people could either be discussants in groups or audiences of public discussions, t here is a growing possibility of an online public space in China (Li, 2010). Despite Chinese governments Internet controls, online citizen media have revived and magnified the public sphere in China, renovating one of the oldest civilizations in the world. The society or the public sphere, traditionally dominated by government disco urse through mainstream media, now has been penetrated by public discourse from new media platforms like public forums and blogs. Zhou, Chan and Peng (2008) claimed that the Internet is increasingly becoming a sphere for Chinese citizens to discuss public affairs, in that it provides users with interactive services such as
53 online forums and blogs, which can encourage information exchange or opinion expression. Examining issues being widely discussed and the participation patterns of Internet users in online public forums and blogs helps us to gain insights into the formation, characteristics, and potentials of the new online public sphere in China. Public Forums O nline public forum s are conceptualized as a conversational arena or a virtual communication spac e, in which people can come together to talk with each other, discuss issues, form opinions or plan action, hence allowing for many to many asynchronous communication for the public (Morris & Ogan, 1996; Schneider, 1996). In fact, electronic discussion groups, preceding the appearance of the Web, have been around since late 1970s in the forms of Usenet discussion groups, newsgroups, conferences, and forums (Light & Rogers, 1999). Electronic bulletin board systems (BBS), also referred to as online bulletin boards, electronic bulletin boards, or bulletin board services, are one of the earliest forms of online public forums. Ward Christensen, who coined the term bulletin board system, began preliminary work in 1978 on what would eventually become CBBS (Comput er Bulletin Board System), the first Bulletin Board System ( Schweitzer, 2005). A BBS is defined as an interactive, online, electronic database system that helps users exchange information conveniently through the Internet (Dewey, 1987, p. 1). The 1980s and 1990s were the heyday of the BBS (Dewey, 1987) Later different types of online forums have come into being, although BBSes remain in use around the world.
54 Characteristics of online p ublic f orums Online public forums can be free and open to everyone, a lthough there are also closed and private systems As described by Roberts, Wanta, and Dzwo (2002), a n online public forum normally can be categorized into many sub forums based on subjects, needs and user preferences. In each sub forum there are various threads dealing with different topics concerning the subject of interest. Users can choose to only read the messages, to start a new thread by posting a new message, or to respond to the original post or posts from other participants in a developing thread (Antaki, Ardvol, Nez, & Vayreda, 2005). Each post is marked with a date and time. All the posts are public and visible to others. Conversations take place if different users continue posting in the forums. Because they offer good places to find people with similar interests and hobbies, online public forums are able to attract users from various backgrounds and gain popularity in different countries. Five potential reasons were identified to explain why online public forums have been popular: they are low cost; they are fast at sending and retrieving information; it is easy to post messages in online forums; and there are many topics, issues, and interests available in online forums; there is a large audience for the online forums (James & Wotring, 1995). Interactivity and deliberation of users are vital elements in online public forums (Morris & Ogan, 1996; Zhou, Chan, & Peng, 2008). Interactivity is reflected by the responsiveness of communication, in which the roles of the message sender and receiver are equal and interchangeable and the communicative power is more symmetrical than with oneway communication (Kiousis, 2002; Schultz, 2000). Deliberation involves recognizing, incorporating,
55 and rebutting the arguments of others as well as justifying one s own (Zhou, Chan, & Peng, 2008). A public forum should be able to hold certain level of interactivity and certain degree of deliberation in the discussions among users. Interactivity between users is an essential component for online public forums because the basic function of online public forums is for people to share or negotiate their opinions about various social issues with others (Schneider, 1996). The more the users participate in communication in online forums, the more reciprocity and conversations will be fostered, which is the cornerstone of a well functioning public sphere (Brants, 2005; Schneider, 1996). Deliberation is another essential component for online public forums. The deliberative component refers t o the idea that public decisionmak ing can be achieved by argumentation offered to and by participants committed to the values of rationality and impartiality (Wright & Street, 2007). For discussions to be useful in a democracy, opinions need to be debated. Deliberative forums can not only create public space within which citizens can express their respective viewpoints but also afford them the opportunity to elaborate on their positions by being forced to confront individuals who may raise opposing arguments (McLeod, Scheufele, & Moy, 1999, p. 744). In this sense, online public forums are similar to Group Decision Support Systems, in which anonymous participants experience more freedom to express their own opinions and criticize others ideas without fear of reprisal. The consequent substantive conflicts may facilitate consensus finding pro cesses and enhance the decisionmaking (Dubrovsky, Kiesler, & Sethna, 1991).
56 Patterns of participation There are various topics for discussion in online public forums. Users can post news, information, or commentary. Unlike faceto face interactions, users are anonymous in online discussions. Wright and Street (2007) developed three categories of messages in online public forums: 1) seed (to spark new discussions), seek ( to make a query ) or provide ( to of fer information or content to the forum ) ; 2) reciprocity to respond to a previous message or to include opinions or information from sources outside of forums (Barak & Gluck Ofri, 2007); 3) validation to offer arguments or reasons for the validity of the participants positions to achieve rationally motivated agreement. Chen and Chiu (2008) classified messages displayed in online public forums as: 1) contributions that introduce new ideas or information into the discussion; 2) repetitions that repeat the previous content, indicating ones understanding and often agreement with the earlier poster; 3) null content including compliments (i.e. Good answer!), simple evaluations (i.e. no), general questions (i.e. What?), and off topic messages (i.e. advert isement). Three structural features of public discussion network size, discussion frequency, and network heterogeneity have been identified in online forums (Kwak, Williams, Wang, & Lee, 2005). Network size, an indicator of the forums popularity, is an im portant factor in attracting users. However, due to different experiences and expectations, participants may have different levels of engagement in online forums (Light & Rogers, 1999; Schultz, 2000). In fact, it was revealed that, compared to the general traffic, a small number of hardcore active participants contributed the bulk of the discussion to the online forums
57 (Light & Rogers, 1999). A large number of forum users are lurkers who act as the audience for the discussion by only hitting the thread and reading the messages without leaving any reply, which is called periphery participation (Wu, 2008). In an analysis of Chinese online forums, Wu (2008) found that every 100 hits of a thread generated one to three replies on average; the majority of the hits 86 to 99 hits was by lurkers. Responsiveness or interactivity between users is an essential component for online discussion, indicating the importance of the message being discussed (Chen & Chiu, 2008). Yet, active participation and periphery participation in discussion are both important to research because such analysis reveals the participation pattern of users and reciprocity of information. The notion of diversity or homogeneity of discussion is critical for researchers to understand the public discourse because, for discussions to be useful in a democracy, a range of opinions should be debated. Through the back andforth of disagreement and discussion, publ ic consideration, consensus and decisions can be made (Dubrovsky, Kiesler, & Sethna, 1991; McLeod, Scheufele, & Moy, 1999). Unlike faceto face contacts, people are more likely to express their disagreement with others in online anonymous interactions (Chen & Chiu, 2008). The homogeneity of discussion is normally evaluated by whether and how later p articipants support or oppose the ideas posted before (Zhou, Chan, & Peng, 2008). Researchers also found that, in online discussions, the earlier messages can affect later messages in terms of evaluations, knowledge content, social cues, personal informati on, and elicitation (Chen & Chiu, 2008).
58 More specifically, a disagreement in previous messages increased the likelihoods of disagreement displays in later messages (Chen & Chiu, 2008). Audience agency is another important factor in online participation (Mitchelstein & Boczkowski, 2009). Giddens (1984) viewed that human beings act intentionally as human agency, intrinsically involving with society and actively entering into its constitution. The repetition of the acts of human agency reproduces the structur e; further, human agency and social structure are in an interactive and reciprocal relationship (Giddens, 1984). Applying Giddens structuration theory to online discussion, Internet user s as audience agency contribute to the development of online discussi on and participation, meanwhile the structure of online participation influence human interaction. Similarly, social practice theory proposed that each individual carries a habitus, which is a set of techniques, references, and beliefs, based on their past experiences and socialization processes; each individuals habitus generates strategies that form the basis of their actions (Bourdieu, 1993, p. 72). Further, many people share repertoire of ways of doing things, hence creating collective habitus; also, t he interactions between human agency develop sets of meaning, discourse, practice and ways of thinking within particular contexts (Bourdieu, 1993, p. 72). As a result, culture, collective identity, and individuality are constructed and enacted. According t o social practice theory, we can understand that Internet users behave individually and interact with each other as audience agency, developing shared understandings and creating norms, collective values, and
59 culture in online participation. The online context operates as a lens through which they see the world and through which their practice are mediated. Blogs Blogs, or weblogs, began as often highly personal online periodical diaries, posted by individuals in reverse chronological order, telling stories allowing reader comment and sometimes offering multimedi a and hyperlinks (Deuze, 2003). U ser generated content in blogs can be presented by texts, pictures, videos, audio, and/or hyperlinks. Anyone can publish at any time and communicate with readers thr ough blogs ( Tremayne, 2007) As a result, blogs become a personal publishing system through which anyone theoretically can have access to a worldwide audience (Gillmor, 2004). Development of blogs There has been an explosion of blogs on the Internet since the early 1990s, when blogs first appeared (Dailey, Demo, & Spillman, 2008). In the late 1990s, there were no more than 100 blogs with limited influence in the society (Johnson & Kaye, 2004 ). By 2006, 8% of American adults (12 million) blogged and 39% of t he online population in the United States (57 million) read blogs (Lenhart & Fox, 2006). Blogs have established themselves as an important part of online culture. Blogs can be understood as a new form of communication that shifts the control of informatio n from the hands of mainstream media to the audience, allowing ordinary citizens to express views and engage in public conversation (Gillmor, 2004). There have been some landmarks in the development of blogs and their connection with journalism. In 2002, S enator Trent Lott had to resign because his racist comment in public was caught and kept alive by bloggers,
60 which led professional journalists and the general public to pay enough attention to it (Bowman & Willis, 2003). In 2004, Rathergate bloggers pursuit of CBSs use of forged documents in its reporting on President Bushs National Guard service resulted in the launch of an investigation by CBS and the resignation of Dan Rather from the news anchor desk (Bowman & Willis, 2003). Eyewitness bloggers such as Salam Pax gave stunningly detailed early accounts of the war in Iraq, on which mainstream media failed (Glaser, 2006). In addition, bloggers have been given press passes since the 2004 U.S. political conventions (Lane, 2010). All of the above examples show that bloggers have been given unprecedented opportunities for personal expression. Individuals who do not have any professional journalistic training or background can get engaged with the process of collecting, reporting, analyzing and disseminating news and information ( Tremayne, 2007) Linking this to the notion of citizen journalism, blogs facilitate a decentralized, bottom up approach to news reporting by helping readers become active partners in news production and dissemination (Gillmor, 2004). In this sense, blogs represent a democratization of traditional journalism (Gillmor, 2010). They have been regarded as a new means for individuals to bypass media gatekeepers to participate in public discourse on important social issues with access open t o all (Deuze, 2003). Regardless of the controversy surrounding blogs, either prejudice against them as a journalistic form from professionals or recognition of their merits (Thurman, 2008), blogs have become citizen media that are independent of and someti mes challenging to mainstream
61 media organizations. Preliminary researchers have attempted to explicate blogs as a social phenomenon, focusing on the blogged content and the social function of blogs on journalistic and public discourse (Deuze, 2003). It was shown that, representing the civic discourse, blogs empowered the public by encouraging expression and conversation so as to expand and renew the concept of public sphere (Dailey, Demo, & Spillman, 2008). Being either opinionated, informational, or both, blogs cover a wide range of topics and styles. There are domains such as personal journals, opinions, voyeuristic observations, journalistic reporting or gossip in blogs, although most blogs remain firmly rooted in personal disclosure ( Bryant & Oliver, 2008; Lenhart & Fox, 2006). Blogs share some characteristics such as open participation, interactivity, multimedia, and hyperlinks. Users can post comments that are published alongside the bloggers original post ( Thurman, 2008). With reader comments, the int eractivity or twoway communication between bloggers and readers as well as between readers and readers facilitates conversation (Kiousis, 2002). Hyperlinks and multimedia are the other two important features of blogs that will be discussed in the following section. Hyperlinks and multimedia Hyperlinks are references or navigation elements linking to some other internal content within the same domain or some external content that is located elsewhere on the Internet (Deuze, 2003). By linking to more background materials instead of including all of them in the primary text, hyperlinks make online stories more contextual (Tremayne, 2004). Applying network theory to the use of hyperlinks in online news stories and the war blogosphere, researchers
62 have two major findings: first, the use of hyperlinks in online stories and blogs has been increasing, which is consistent with network theory; and second, there was a mechanism of preferential attachment story topics had a significant effect on the linking practice (Tr emayne, 2004; Tremayne, Zheng, Lee, & Jeong, 2006). International relations stories included a higher number of hyperlinks than other stories (Tremayne, 2004). Moreover, Tremayne, Zheng, Lee, and Jeong (2006) found that ideology, post type, gender, and blogger background can influence the outgoing and incoming hyperlinks in blogs. It is important to examine hyperlinks in blogs because hyperlinks can direct readers attention to what bloggers believe is important in a certain social context. There are two basic types of hyperlinks in blogs, including internal hyperlinks and external hyperlinks. Internal hyperlinks are hyperlinks that direct readers to content in the same blog or website. External hyperlinks are hyperlinks directed to Web pages outside the blog or website. Bloggers can create networks by sending hyperlinks that direct readers to other websites and attracting readers into their blogs (Tremayne, 2004). Hyperlinks can link to different sources of content such as legacy news sites, citizen news sit es, or original documents, etc. They also can link to different types of content such as textual content and multimedia content. Different sources or content directed to through hyperlinks shows the bloggers understanding of the issues in coverage as well as their abilities in using new media technology (Carpenter, 2010). Multimedia refers to the application of photographs, video, audio, slideshows and graphic information ( Carpenter, 2010). A study from the Project
63 for Excellence in Journalism (2007) foun d that the use of video, photo, or graphics in stories on news organizations was still limited, although on the rise. Based on a quantitative content analysis, Carpenter (2010) found that, in comparison with online newspaper publications, online citizen journalism sites such as blogs were more likely to feature interactive and multimedia components. Hyperlinks and multimedia features in online news articles or blogs are important because they can help to give readers better understanding of stories and easi er navigation of websites (Tremayne, 2007). The Influence of Citizen Journalism Citizen journalism has been playing an important role in influencing public opinion and engaging citizens in public affairs (Gillmor, 2004; Muthukumaraswamy, 2010). The rise o f citizen journalism has been recognized by many researchers as heralding a new era where the power to define news is decentralized (Gillmor, 2004, 2010; Moyo, 2009). Schaffer (2007) maintained that citizen media is emerging as a form of bridge media, linking traditional media with forms of civic participation (p. 3). Some typical characteristics of citizen journalism include information sharing, collaboration among users, interactivity, etc. (Gillmor, 2004; Johnson & Wiedenbeck, 2009). Conversation in di fferent forms such as comments and hyperlinks lie in the center of citizen journalism, as it expands twoway communication and provides opportunities for citizen inputs (Johnson & Wiedenbeck, 2009; Lacy, Riffe, Thorson, & Duffy, 2009). Equipped with new technology tools such as SMS, email, and blogs, citizens across the world are able to challenge or augment mainstream media coverage of issues relevant to their lives. The activism of citizen journalism and its impact on
64 traditional media have been shown in events such as the September 11 terrorism attack, the 2004 Asian tsunami, the London bombings, Hurricane Katrina and presidential elections in the United States, Australia, South Korea, and Zimbabwe (Moyo, 200 9; Xin, 2010). Ordinary people helped in the worldwide reaction and relief efforts to the tsunami and flooding in Southeast Asia in late 2004 as well as Hurricane Katrina in the United States in 2005 (Glaser, 2006). Citizen journalism, which challenges the professionalized and institutionalized practic es of mainstream media (Atton, 2003), i s of special importance in a country like China where mainstream media are still under tight control, while social conflicts are intensifying, as in the milk formula scandals Based on an analysis of four case studies of citizen journalism practice in China, Xin (2010) concluded that the impact of citizen journalism on Chinese mainstream media is multifaceted, as citizen journalism is sometimes used by mainstream media as a news source as well as an alternative channel to distribute politically sensitive information. Aside from the challenges brought by citizen journalism to mainstream media, researchers have found that citizen news/blog sites can serve as complements to daily newspapers (Lacy, Duffy, Riffe, Thorson, & Fleming, 2010). Based on a systematic study of the day to day practices of citizen journalists, Reich (2008) came to a similar conclusion that citizen journalism can serve as a vital complement to mainstream journalism, instead of its substitute. Reese, R utigliano, Hyun, and Jeong (2007) revealed that, although blogs managed to retain citizen voices that are not shown in traditional media when covering
65 current events, most citizen blogs are still an important secondary market for news from mainstream media organizations. Xin (2010) found that citizen journalism is unlikely to be a driving force in promoting social changes compared to what mainstream media can do in China, but citizen journalists can work together with mainstream media to make the information flow in China more transparent. It was argued that online discussions could probably drive the agenda of official media in China (Qiang, 2004). Media in China Traditional Media in China Since Chinas Reform and Open Policy in the late 1970s, China has been growing into an important global player, emerging as a significant economic and cultural factor in the world. Along with the huge growth in economics, there have been great changes in Chinese media systems. The reform of Chinese media in the 1980s was c haracterized by the withdrawal of government subsidies and pushing on media into market economics, which gave birth to the unique media system that media are dependent on advertising revenues while also being regulated and content controlled by the governm ent (Polumbaum, 1990). Some argued that there were underlying conflicts of interests in this system conflicts between the political priorities of the Party and the principles of the commercialized news media, or clashes between the communist propaganda nor ms and journalists pursuit of free speech (Sparks, 2010). In other words, it is hard for media to simultaneously attract a mass audience while keeping strict controls of media content as is required by the state (Sparks, 2010). Regardless of the inner ten sion for media and media practice, the scope and aims of
66 reporting have shifted and expanded. The political and social implications of media reform are still unfolding as the government, media, and the public are adapting to this media environment. General ly speaking, all the media in China, including newspapers, magazines, radio, and TV are state media. The entry into the World Trade Organization (WTO) in 2001 did not change the stateowned nature of media in China, although some outside investors have bee n allowed to own no more than 49% share of some media (Messner & Garrison, 2006). With a population of more than 1 .25 bi llion, China now has more than 2,000 newspapers with a total annual circulation of more than 26 million, more than 3,200 television stat ions and 700 radio stations (Press Reference, 2010). Since the media reform in the late 1980s, there have been basically three categories of newspapers in China: 1) party newspapers as the throat and tongue of the party that reflect the partys voice; 2) trade/professional newspapers with emphases on one area of human life such as business, fashion, art, etc., 3) metro daily newspapers that are more close to peoples daily lives and have relatively more freedom for reporting, although still under the close wa tch of the government (Press Re ference, 2010). There are differences in party affiliation, government control, reporting style, and coverage focus between different types of newspapers. The party newspapers and metro newspapers are the main subjects o f this study. The media reforms also have been applied to radio and television. Since the late 1980s, radio and TV stations in China have been transferring from
67 government agencies to enterprises, albeit stateowned enterprises, embracing commercial management through the business of advertising and programs rights (Liu, 2010). Similar to newspapers, they maintain political functions alongside their commercial operations (Liu, 2010). Except news and current affairs, there is also a large amount of popular entertainment programming in TV broadcasts in China (Sparks, 2010). For example, there is a popular program called Super Girl, which is modeled after American Idol. Most major newspapers and TV stations have launched websites, hosting a variety of activ ities such as online forums/communities, blogs, article/program archive, etc. (CNNIC, 2006). Development of the Internet in China Computers were first introduced to China in the mid1980s and the Internet debuted in China in 1986 (Wei, 2009). The Internet is developing fast in China, as part of the countrys rapid modernization process. The Pew Internet and American Life Project reported that there were 137 million Chinese Internet users at the end of 2006, 165 million by mid2007, and a whopping 210 milli on by the beginning of 2008 (Fallows, 2008, p. 5). According to the China Internet Network Information Center (CNNIC), a nonprofit organization and the state network information center of China, by the end of 2009, the number of Chinese Internet users had reached 384 million, 28.9% up from 2008; by June 2010, there were 420 million Internet users in China, the largest online population in the world (CNNIC, 2010). The penetration rate of the Internet has grown to 31.8%. Among the entire online population i n China, 364 million (98.1%) were broadband users (CNNIC, 2010). There were nearly 10% more male Internet
68 users than female Internet users in China, at a ratio of 54.8: 45.2 for males and females (CNNIC, 2010). The largest user group is aged 30 or above and accounts for 41% of the whole online population in China (CNNIC, 2010). Information acquisition is a major reason for Chinese individuals going onlinea bout 80% of Chinese Internet users read news online during the first half year of 2010 (CNNIC, 2010). In another words, online news users have reached 307.69 million in m ainland China (CNNIC, 2010), which makes the Internet an important frontier where journalists must compete. Online music, online news and search engines are the top three Internet applications (CNNIC, 2010). In addition, online audio/video, blogs, social networking sites, forum/BBS, and online shopping were also very popular among Chinese Internet users (CNNIC, 2010). The arrival and development of the Internet in China holds the potential to create a freer public sphere. Researchers have stated that, although the Chinese government still holds tight control over media in China, especially over politically undesirable media content, due to the development of market economics and the Reform and Open Policy started in the late 1970s, there has been increasing openness and freedom for Chinese media (Lee, 2000). First, the Internet provides an alternative information source for Chinese people, so users may be able to search and find information that is unavailable from traditional media or learn about issues otherwise not covered in mainstream media (Sparks, 2010; Wang, 2007). In addition, the World Wide Web makes it possible for Chinese citizens to access foreign news or news coverage of China by foreign media
69 directly, hearing different voices and remarks (Burger, 2009). Enough information is indispensable for a public discourse to build and for public sphere to exist. Second, the Internet provides ways for Chinese people to easily participate in democratic discourses intentionally or unintentionally. Some independently minded journalists use the Internet to find sources and to publish stories that would otherwise be difficult to develop in mainstream media (Sparks, 2010). For ordinary citizens, Zheng and Wu (2005) argued that in China the Internet is perhaps the single most important avenues to criticize government policies and to participate in politics (p. 525). Admittedly, content distributed online is monitored in China. For example, the Great Firewall is used to filter out politically sensitive information that may challenge the Communist Party or undermine national unity and social stability (Zhang, 2006). However, people may use irony, parody, or innuendo to say what they want to say ; informed Internet users can read between the lines, hence enjoying a higher level of freedom (MacKinnon, 2008). There have been lively debates online that otherwise wont be found in a form of public discussion and expression (Sparks, 2010). Wang (2007) revealed that the huge number of online comments and discussions generated during some events in China, such as the anti Japan protest in 2005, have built powerful public opinion and exerted pressure on the government. New Media in China As the online com munity in China has grown steadily at doubledigit rates over recent years, new media tools are making great stride in China (Nielsen China Forum, 2010). Chinese Internet users listen to music, read the latest news,
70 search online, talk to each other in onl ine forums, connect with friends through social networking sites, play online games, and view user generated content (i.e. blogs, home vi deo, etc.). Generally speaking, local players have dominated the digital landscape in China. Some popular online applic ations include Web portals such as Sohu and Sina, which reports the highest online traffic in China (from Alexa.com), search engine Baidu (similar to Google), videosharing sites such as Tudou and Youku (the equivalents of YouTube), social networking site Xiaonei (the equivalent of Facebook), online public forums like Ti a n ya with a variety of topics on automobiles, arts, entertainment, politics, economics, sports, politics, etc., and blogging services such as Sina Blog. Citizen media in China have been off ering an increasing number of rich and vigorous stories that have attracted nationwide attention and debate, while traditional media were limited by their routinized and institutional forms of newsgathering and news presentation, showing a narrow understan ding of what news is (Chang, 2008). The technologies have led to the development of citizen participation, which promotes the spurring and sharing of information and views without the iron control from the government, hence bringing in liberalization to jo urnalism and mass communication (Moyo, 2009). Although the government attempts to curb postings about sensitive issues, especially during socially sensitive periods, Chinese Internet users can always find ways to get around the censorship. It is hard to totally suppress the Internet. Public forums in China Online public forums and blogs have been viewed as ideal places for people to express their opinions and interact with others in China (Wang, 2007).
71 By June 2010, nearly one third (31.5%) of Chinese Internet users had participated in online discussions on a variety of online forums/bulletin boards (BBS); more than half (55.1%) of Internet users were bloggers (CNNIC, 2010). Consistent with what CNNIC found, the survey by Nielsen China Forum (2010) revealed that about one third of users have taken part in online discussion forums and BBS (bulletin board) sites, which are one of the most politically active place in Chinese cyberspace; more than half of all online Chinese users regularly blogged and used social media to voice their opinions about entertainment, relationships, sports, politics, technology, etc. Online public forums are very popular in China due to their low cost, easy access, and the abundance of information (Zhou, Chan, & Peng, 2008). People in China can discuss a wide range of topics and issues in online public forums (Yang, 2003). Anyone can start a discussion by posting a new message and starting a new thread, which then grows as others reply to the original message. A large number of audienc e members can read simultaneously and give feedback immediately. The online public forums can be operated by individuals, commercial companies, or government agencied (Qiang, 2004). At any given time, there are literally tens of thousands of users active i n online forums, reading news, searching for information, and debating current events (Qiang, 2004). Compared with the traditional mainstream media, there are more critical remarks about Chinese government policies/practices and various social or political problems in online public forums in China (Yang, 2003).
72 B logs in China Blogs were first introduced into China in 2002 (CNNIC, 2006). By the end of 2009, the number of blog users in China reached 221 million with an annual increase of 36.7%; more recently, the number of bloggers grew to 231 million (CNNIC, 2010). In their blogs, authors post stories, upload video and audio content, receive comments from readers and sometimes respond to these comments. The function of blogs has been changing from merely pers onal exhibition to information dissemination among larger populations, to share news that is happening now in real time (Bowman & Willis, 2003). Blogs, along with micro blogs and social networking sites, have become a driving force for the development of c itizen media in China (CNNIC, 2010). A good example to show the growing influence of blogs in China is the coverage of a rare Chinese tiger in 2007. When a rural farmer claimed that he took a photo of a tiger that was thought to be extinct, government aut horities and mainstream media went wild for it. Numerous headlines and awards mentioned the farmer. However, some bloggers posted their thoughts that the photo could be manipulated, which meant that the recovery of the rare tiger was falsified. Thousands o f bloggers followed up, debating over the truth of the issue. Soon, a blogger posted the same image of the tiger from a calendar he had, concluding the debates in the blogosphere and the false coverage in mainstream media. Food Safety Issues in Media Cover age BSE, E. coli 0157:H7, dioxins, hormones, antibiotics, and melamine, to name just a few, are now familiar topics circulating globally and sparking public concern. People nowadays are paying more attention to food safety issues as
73 there have been more food safety problems exposed by media and more inadequacies in the regulatory systems found under public scrutiny (Frewer, 1999; Gwendolyn, 2009). Issues concerning food safety and public health could quickly transform into highly politicized arenas due to public scrutiny (Gwendolyn, 2009). In China, food safety issues are more prominent than in other countries. From January to May 2010, 108 food poisoning incidents occurred, with 56 deaths and 2,452 people being poisoned across the country within five months (Chen, 2010). In 2010, the Ministry of Health in China issued its first announcement warning about food safety since the countrys Food Safety Law was passed (Chen, 2010). Considering the potential for severe and fatal illness from consuming those tainted food products, effective communication of food safety issues is critical to public health (Fleming, Thorson, & Zhang, 2006). It is believed that the news medias role in providing health information to the public has grown substantially (Major, 2009), alt hough inadequate and inaccurate reporting of public health threats were found repeatedly (Lowrey, Evans, Gower, Robinson, Ginter, McCormick, & Abdolrasulnia, 2007; Roche & Muskavitch, 2003; Voss, 2003). Not only traditional media such as newspapers and tel evision, but also new media such as blogs, social networking sites, etc., serve as sources of health information for people. According to the International Food Information Council (IFIC) (2010), 32% of people said they received their food safety informati on from the Internet, and 28% said they received theirs from newspapers (28%). Also, 54% of people surveyed reported that media reports would prompt them to make
74 changes related to food safety, while 8% reported that blogs or social networking sites were m ost likely to prompt their food safety changes (IFIC, 2010). Story Focus and Frames Media coverage of health issues and their selections of topics and frames can heighten the salience of issues and impact the publics health beliefs, attitudes, and behaviors (Gwendolyn, 2009). For public health crises, there are various topics, themes, and frames used in news coverage. Based on an examination of news reports on health topics, including sudden acute respiratory syndrome (SARS), mad cow disease, West Nile vir us, etc., researchers found that risks were more likely to be mentioned rather than prevention or treatment (Berry, Wharf Higgins, & Naylor, 2007). Examining TV network news coverage of the nationwide recall of the E. colicontaminated spinach in 2007, Nucci, Cuite, and Hallman (2009) found that 40% of the stories focused on the status of the investigation into the cause of the contamination that led to the spinach recall, and 20% of the stories primarily emphasized the recall (notice or warning). They cat egorized the story focus in the stories into two major types: 1) public health content including disease symptoms/information, information on the spread of contamination (infected/deaths), state(s) with confirmed cases of illness, type of spinach identifie d as contaminated, guidelines for handling spinach, and at risk populations; 2) investigationrelated content including who was potentially implicated as responsible for the outbreak, possible causes for the outbreak (intentional), possible causes for outbreak (unintentional), and whether the story directed the viewer to additional information (Nucci, Cuite, & Hallman, 2009, p.
75 246). They found that more public health content were information about the spread of the contamination, while the majority of stor ies did not provide any information about the type of spinach contaminated; for investigation related content, it was revealed that more stories were about investigation status rather than information critical for public health (Nucci, Cuite, & Hallman, 2009). Analyzing the birdflu or H5N1 related, content in newspapers and public discussion forums from 1997 to 2006, Hellsten and Nerlich (2010) found that newspaper articles focused more on solutions such as vaccination and drug developments, suggestions and discussion on how to contain the potential pandemic as well as political decisions regarding pandemic preparedness, while the online public discussion covered a wider range of topics, including basic information about bird flu, catastrophic prediction co nnecting to biblical images of Armageddon, or alternative views with humorous undertones. For West Nile virus, it was found that major North American newspapers in 2000 exhibited mixed success in providing information on symptoms and risk reduction measures, while performing poorly in providing risk information such as that related to serious infection, illness, and mortality (Roche, & Muskavitch, 2003). Liu (2010) analyzed the news coverage of the Sanlu tainted formula scandal by the Xinhua News Agency, the leading news agency in China, and showed that the coverage focused on solutions such as law enforcement actions and providing good medical care to sickened babies, while avoiding addressing the governments failure to prevent the crisis from happening and staying away from reporting on peoples grief and moral assessments.
76 In coverage of public health crises or food safety issues, frames can provide cues about how to define news and how to interpret the communicative acts within the specific context ( Entm an, 1993). Analyzing media coverage of controversial scientific issues such as stem cell research, Nisbet, Brossard, and Kroepsch (2003) found that the most dominant frames included strategy/conflict, ethics/morality, policy background, and scientific back ground. However, strategy/conflict frame was more likely to appear as a dominant frame, while other frames such as ethics/morality were more likely to appear as secondary and complementary frames, showing the focus of newspaper coverage over the issue (Nis bet, Brossard, & Kroepsch, 2003). In an analysis of newspaper coverage of Hurricane Katrina, Dill and Wu (2009) found that government failure and the unimaginable destruction were the most predominant frames. In addition, the uncertain future frame, the di stress frame, and the attribution of blame frame were also common (Dill & Wu, 2009). Ex amining how T he New York Times framed three public health epidemics (mad cow disease, West Nile virus, and avian flu), it was revealed that two frames were employed cons istently: 1) the action frame that emphasized preventative and treatment actions; and 2) the consequence frame that focused on the influence of the diseases such as victims, cost, and social impact (Shih, Wijaya, & Brossard, 2008). Frames such as economic consequences, attribution of responsibility, conflict, human interest, and severity were found in newspaper coverage of SARS (Luther & Zhou, 2005). Examining the Xinhua News Agencys coverage of the Sanlu tainted formula scandal three major frames were identified: 1) government reaction frame; 2)
77 treating sickened babies frame; and 3) punishment frame, among which the government reaction frame was dominant (Liu, 2010). Tones and Sources Tone can be understood as the capability of words to show inner feeling, meaning, or judgment on the issues being covered. There is a global attitude measure that can be used to measure the tones of news coverage, including six aspects: negative/positive, foolish/wise, worthless/valuable, unacceptable/acceptable, bad/good, and unfavorable/favorable (Burgoon, Miller, Cohen, & Montgomery, 1978). The most frequently used measurement for tones is negative, neutral and positive. Sources that provide information have been recognized to have a major influence on news coverage (Gan, 1979). Journalists rely on sources for interpretation. In China, public health information such as food safety alerts, issue updates, mass food recall, etc., is always disseminated to the public through press releases or press conferences by the governmen t agencies responsible for supervising food safety. Studies have shown that information subsidies such as press releases from the government play a vital role in structuring media agendas, especially in times of public health crises (Berkowitz & Adams, 199 0), which is probably due to the authority of the government and the resources in their hands to develop and disseminate information (Gans, 1979). Nucci, Cuite, and Hallman (2009) found that the role of government information subsidies in disseminating inf ormation is critical to public health. Other than government sources, health agencies, industry, and university sources, often represented by spokespersons, were frequently cited as well
78 (Nucci & Kubey, 2007). It was also found that, except within the cont ext of treatment, when nonexperts and experts were cited proportionally, expert sources were always cited more frequently than nonexpert sources in the discussion of health topics (Berry, Wharf Higgins, & Naylor, 2007). Research in science reporting has shown that a greater quantity of scientific sources and experts such as scientist, research reports, and health professionals (Ramsey, 1999) lend greater credibility to news stories. With different sources such as the government, health agencies, industry, universities, independent organizations, experts and nonexperts, it allows for multifaceted coverage of public health crises. Synthesis So far in this chapter, previous literature about agendasetting theory, citizen journalism, media in China, and medi a coverage of food safety issues, has been reviewed. Weaving these ideas together for the purposes of this study, this study aims to applying the agendasetting theory to an analysis of traditional media discourses and online public discourses surrounding the tainted milk formula scandals in China. Agendasetting theory holds two fundamental assumptions: 1) people get their news from a finite number of news sources or media outlets; 2) people are influenced by media (McCombs, 2004, 2005). However, in the new media environment, there are numerous channels that people can access to obtain information; different media platforms may have very different perspectives when reporting the same issue; messages received by people are more likely to be individually tailored and actively chosen by the audience (Chaffee & Metzger,
79 2001). As the number of news outlets increases, public attention scatters, and personalized agenda develops, it is possible that, in place of a collectively shared agenda, fragmented and competing media agendas, and therefore diverse public agendas, will emerge (Chaffee & Metzger, 2001). The tainted milk formula incident s in this study w ere of great social importance and controversy. Due to the development of online news sites social network site s, and citizen media, as well as the individualized patterns of media consumption in China, it was expected that there would be diverse online discourses over the issue, as people develop various individual agendas and express different viewpoints about the taint milk. Previous researchers have tracked the news stories most often linked to by bloggers (issue salience) and compared the results to stories favored by traditional media, determining that bloggers constructed an alternative agenda within the blogosphere (Delwiche, 2005). Given the growth of online media in China and the different characteristics and functions of traditional media and new media, it was predicted in this study that newspaper agenda about the taint milk formula representing government and traditional media would not be the same as the public agenda about the tainted formula on discussion forums and blogs. The public may present their own view of what news matters, playing the role of opinion leaders in constructing public discourse ( Delwiche, 2005). In this way, the traditional agendasetting model can be modified by inserting Internet users, no matter forum discussion participants or bloggers, between the public and the press as opinion leaders (Delwiche, 2005). Comparing agendas fro m discussion
80 forums, blogs, and newspapers about the tainted milk formula helps the researcher to find out the agendasetting effects of each medium, investigating their roles in shaping the discourses about the public health crisis. Further, the public may actively interact with media in placing issues on the daily agenda (Zhou & Moy, 2007). The interplay of the public agenda and media agenda changes the focus of agendasetting theory from what issues the media tell people to think about to what issues peo ple tell the media they want to think about (Chaffee & Metzger, 2001). By examining sources in newspaper articles, online forum posts, and blog post, about tainted milk formula scandals the way how online public opinion made the original event a prominent one in traditional media and heard by the mass population, as well as the way how mainstream newspapers influence the ongoing online discussion about the incident s, can be outlined. With all these considerations, the agendasetting theory and the notions of citizen journalism, media in China, and public health coverage were put into a systematic investigation in this study. The Present Study This study analyzes the discourses related to the tainted milk formula scandals in Chinese newspapers, online forums and blogs. The general research questions for this study are: What role did new media play in China during a public health crisis, the tainted formula scandals from 2008 to 2011? How was it different from or related to the function of traditional media in China? Were the public discourse and the government/media discourse about the tainted formula incident s in China the same? Who were the agenda setters?
81 Below is the proposed framework for research, to compare public discourse and media/governme nt disco urses (Figure 21). Figure 21. Proposed framework for study Conceptual Definitions Applying agendasetting theory to the research, this study focused on several aspects of agenda setting: story focus (topic attribute), frame, tone (affective attribute), source, the pattern of participation and multimedia use. Story focus is the subject matter most emphasized by media concerning the formula scandals (Nucci, Cuite, & Hallman, 2009). The choice of story focus and the amount of coverage for different story f ocus are major indicators of issue salience. Frames are basic cognitive structures that guide the representation and perception of reality ( Entman, 1993) Audiences use the explanations given by media/bloggers their selection and presentation of information to understand complex issues and make decisions on their own (Entman, 1993; Gamson & Modigliani, 1989). Different frames may perform different functions such as defining problems, diagnosing causes, making moral judgments, and suggesting remedies ( Entman, 1993) Conceptually speaking, ton e is the quality and modulation of the words depicting the tainted formula incident s with the potential to show the
82 communicators inner feeling, meaning, or judgment on the issue. Source refer s to those who are quoted i n articles, discussion forums, and blogs. Information in media is collected from various sources, which are believed to influence media content in tremendous, and sometimes subtle, ways The patterns of participation in online forums and blogs include two dimensions, interactivity and deliberation. Interactivity is the responsiveness and communication among forum users or between bloggers and readers, while deliberation is the way forum users and bloggers incorporate and rebut others opinions, and justifyi ng their own arguments. Research Questions This study is aimed at analyzing the discourses surrounding the milk formula scandals in Chinese newspapers, online forums, and blogs, gaining insights into the growing contributions and importance of new media, a nd exploring the implications for journalistic practice and public sphere development in China. Based on agendasetting theory and the concrete theoretical constructs in the reviewed literature, research questions were proposed. Story focus, or issue attri bute, is an important variable to analyze in secondlevel agendasetting research. It is important to know what attributes of the issue, the tainted milk formula scandals were emphasized in newspapers, discussion forums, and blogs. So the story foc i in di fferent medias coverage of the tainted formula scandals were examined and compared. In addition, there could be several issues or foc i being mentioned in a single article or post. However, there was only one major focus, which was the dominant theme of th e story It was
83 interesting to see which f oc i were mentioned in the story and what the major focus was Thus, the following research questions were posited: RQ1 : What were the story foc i in the online forum posts, blog posts, and newspapers in China concer ning the tainted formula scandals ? RQ1a : What w as the major focus in each online forum post, blog post, and newspaper article ? RQ1b : How did story foci and major focus differ across newspapers, online forums, and blogs? F rame is another variable that is linked to secondlevel agendasetting studies. When covering the tainted milk formula scandals the presentation of frames by newspapers and bloggers may reflect their understanding of the issue and the priorities they think should be emphasized. Similar to story focus the story frame also could change over time. To understand the framing of the tainted milk formula scandals in newspapers and blogs, the following research questions were asked: RQ2 : What frames appeared most frequently in the formulascand al related newspaper articles and blog posts? RQ2a : What w ere the dominant frames in blog posts and newspapers? RQ2b : How did framing differ between newspaper and blog coverage of the formula scandals ? Tone is an important factor to influence the presentation and perception of the issues. As for the tainted milk formula scandals it is necessary to analyze the major tone of the formulascandal related content in newspapers, blogs and online public forums, examining whether there are differences between m edia. So, the following research questions were proposed:
84 RQ3 : What were the major tones of the online forum posts, blog posts, and newspapers concerning the tainted formula scandals ? RQ3a : How did tones differ among newspapers, online forums, and blogs? Sources could exert tremendous influence on the process of agenda setting. The sourcing patterns of media in China during the public health crisis are a point of interest in the study. Considering inter media agendasetting effects, how citizen media and mainstream newspapers cite each other may provide a way for the researcher to define the roles of old and new media in China. Therefore, to uncover the common sources used in newspapers, blogs and online forums concerning the tainted formula scandals and t he differences in sources among newspapers, blogs and forums, the following research questions were outlined: RQ4 : What were the most frequently cited sources in the online forum posts, blog posts, and newspaper articles when covering the tainted formula scandals ? RQ4a : How did sources differ among newspapers, online forums, and blogs? One of the most significant features of new media is participation, which include two dimensions, interactivity and deliberation. Interactivity refers to the responsiveness and communication among users, while deliberation is the way users recognize, incorporate, and rebut the arguments of others as well as justifying their own arguments (Kiousis, 2002). Mor e specifically, the researcher wa s interested in the responsiveness and the homogeneity of discussion in Chinese new media about the tainted formula scandals answering questions such as: D id stories of particular story foci frames, or tone generate more
85 discussion sreceiving more reader comments in blogs or hav ing more p osts/threads in forums ? Did Chinese people really participate in online discussions? Or d id most of the m just read blogs/forums while not contributing ? For those online discussion participants, how active were they? In the discussions, were there many opposing views? Or were most opinions about the issue homogeneous? To find answer s, following research q uestions were asked: RQ5 : What were the patterns of participation in new media? RQ5a : What stories ( story foci frames, tones) generated more discussion? RQ5b : What was the homogeneity of discussion in online forums and blogs? In new media, m ultimedia and hyperlinks, if used appropriately, can attract more readers. The employment of multimedia and hyperlinks is another indicator of Chinese Internet users familiarity with new technologies. It is interesting to look at how multimedia such as pictures, audio, and video were embedded in blogs and forum posts; and how internal hyperlinks and external hyperlinks to various sources (i.e., newspapers, TV, news agencies, other blogs, online forums, government/health agency sites, etc.) and to different types of content (i.e. texts and multimedia content) were included in blog posts and forum threads. So, the following research questions were proposed : RQ6 : How were multimedia and hyperlinks employed in blogs and online forums? RQ6a : How many pictures were embedded in each blog post/forum thread? RQ6b : How many audio or video clips were embedded in each blog post/forum thread?
86 RQ6c : How many hyperlinks were included in each blog post/forum thread? Last but not least, c haracteristics such as the gender and background of authors may influence the content (i.e. story focus frame, tone, and source). Because participants in online discussion forums are anonymous, only the authors of newspaper articles and blogs will be analyzed. Yet, could the characteristics and background of authors influence the coverage of the tainted formula scandals in Chinese newspapers and blogs? D id different types of bloggers (i.e. parents health experts, journalists, average citizens) play different roles in the agendasetting process? For example, were health expert blogs more likely to be agenda setters? Were parent bloggers more likely to be agenda follower s? To find answers, the follo wing research questions were suggested: RQ7 : How did authors influence the coverage of the formula scandals in newspapers and blogs? RQ7a : Did the gender of authors influence the content of blog posts and newspapers? RQ7b : Did the type of blogger (health expert, parent, journalist, citizen) influence the content of blog posts?
87 CHAPTER 3 METHODS Content Analysis This study was aimed at examining the role of new media in China by analyzing the story focus frames, tones, sources, discussion patterns, hyperlinks and multimedia employed by newspapers, public forums, and citizen blogs when covering the tainted milk formula scandals Quantitative content analysis was used as the research method. Content analysis is defined as: the systematic and replicable ex amination of symbols of communication, which have been assigned numeric values according to valid measurement rules and the analysis of relationships involving those values using statistical methods, to describe the communication, draw inferences about its meaning, or infer from the communication to its context, both of production and consumption (Riffe, Lacy, & Fico, 2005, p. 25). Neuendorf (2002) pointed out that content analysis is a summarizing, quantitative analysis of messageswhich is not limited t o the types of variables that may be measured or the context in which the messages are created or presented (p. 10). To examine the information being disseminated by newspapers, public forums, and blogs during a public health crisis in this study, content analysis w as chosen as an appropriate research method because it is particularly well suited to the study of communications and to answer the classic question of communication research: Who says what, to whom and how (Babbie, 2007, p 320 ). Lomb a rd, Snyder Duch and Bracken (2002) stated that content analysis, as a method specifically for the study of messages, is fundamental to mass communication research. The strength of content analysis lies in its unobtrusiveness, although, as a research method it is basically
88 descriptive and limited by the availability of material (Riffe, Lacy, & Fico, 2005, p. 25). Nevertheless, ample related studies and abundant first hand data for analysis may help to overcome the weakness of content analysis, and better building theoretical assumptions and predictive power for this study. Sampling Newspapers, online forums, and blogs were the major media platforms chosen for this study. They are all very popular forms of communication where news and information is disseminated. Newspapers represent traditional mass media, while online forums and blogs represent new media as alternative ways for Chinese citizens to gather information and express opinions. Regardles s of Internet censorship, it is undeniable that Chinese citizens have been enabled by online media such as the online forums and blogs to express their opinions more freely. The time frame for the study was from May 2008, when the problem of the milk formula was first revealed by parents online, to June 2011. The thr eeyear research frame allows for a comprehensive observation of the milk formula scandals coverage in the time sequence from the exposure of the problem to the time that the topic died down. Newspaper Sampling Eight newspapers were selected: 1) the Peopl es Daily 2) the Beijing Youth News ; 3) the Southern Metropolitan News 4) the Oriental Morning Post 5) the Xinmin Evening News 6) the Changjiang Daily 7) the Chutian Metropolis Daily and 8) the Hebei Daily ( Table A 1) All of them are in Chinese. Wit h 2,350,000 in circulation, the Peoples Daily is a national daily newspaper that provides comprehensive daily coverage of domestic and
89 international news and events (Willings Press Guide, 2009). It is the most authoritative newspaper a nd the CCPs party throat and tongue in China, which can be understood as the voice of the government, as its story selection, editorial choice, and photo placement reflect the politics of the central government ( China Media Guide, 2008). The Peoples Daily s Chineselangua ge website covers an extensive range of topics and lists news feeds from news or ganizations across the country ( China Media Guide, 2008). Published since 1951 and led by the Chinese Communist Youth League committee, the Beijing Youth News is one of the most influential newspapers among younger Chinese audiences aged 21 to 48 years old ( Press Re ference, 2010; Willings Press Guide, 2009). W ith a circulation of 680, 000, it provides indepth and comprehensive news and feature coverage of domestic and international events (Willings Press Guide, 2009). In the late 1990s, the Beijing Youth News changed from a daily broadsheet with ei ght pages to 46 pages (Press Re ference, 2010). It is famous for its Freezing Point (bingdian in Chinese: ), a commentary based editorial column, which often pushes the boundaries of whats allowed by the authorities (China Media Guide, 2008). The Southern Metropolitan News is a daily ne wspaper known for its investigative journalism and provocative commentary (China Media Guide, 2008). It was the leading newspaper in reporting on SARS and the Zhigang Sun beating case, which led to the repeal of the forced repatriation law in the end (Chin a Media Guide, 2008). With a circulation of 1,580,000, the Southern Metropolitan News is among the metro daily newspapers of highest circulations
90 (Willings Press Guide, 2009). The Beijing Youth News and the Southern Metropolitan Daily although one is from the north and the other from the s outh, were chosen for this study because both of them are considered the most aggressive newspapers in China ( China Media Guide, 2008). The Changjiang Daily with a circulation of 400,000, focuses on local (Hubei Provinc e), domestic and international news covering a wide range of topics (Willings Press Guide, 2009). Established in 2003, the Oriental Morning Post is a comprehensive daily covering 15 cities in the Yangtze River delta with a targeted audience of well educated people aged between 30 to 35 years and with a fairly good income (East Day, 2003). The Changjiang Daily and the Oriental Morning Post were chosen because they were among the earliest Chinese newspapers to report on the tainted formula scandals with the latter being the first newspaper to lin k the tainted formula to Sanlu C ompany. First published in 1929 and with a circulation of 1,100,000 now, the Xinmin Evening News is a regional daily newspaper that reports on international and national news, includin g economic, business, culture and entertainment developments (Willings Press Guide, 2009). Its content is softer than party newspapers but harder than metro tabloids (Huang, 2001). The Chutian Metropolis Daily is a metro tabloid based in Hubei Province wit h a circulation of 1,300,000 that provides general news (Willings Press Guide, 2009). Its tenets are consider what people consider; worry about what people worry about; give what people need; solve the problems people have. The Xinmin Evening News and
91 th e Chutian Metropolis Daily were chosen to represent different types of metro news papers and newspapers from the east and the w est. With a circulation of 410,000, the Hebei Daily is a party newspaper reporting on domestic and international news and events with emphasis on provincial news coverage. The Hebei Daily was selected because Hebei is the province where the Sanlu Company is located. When the milk formula scandal was first exposed online, the Sanlu Company decided to cover it up in order to protect t he brand. The local government, the Shijiangzhuang city government and the Hebei provincial government, were also informed about the problems very early. But local authorities helped to cover up the matter because Sanlu is an important company to provide e mployment and tax income for the city (Zhao & Lim, 2008). The Hebei Daily was used to show a local perspective on the issue. All of the newspapers have archives of news stories from 2005. With the story archives available on newspaper websites, the keywords tainted formula ( wenti naifen in Chinese: ) and poisoned formula ( du naifen in Chinese: ) were used for searches of formulascandal related articles because they were the most widely accepted labels that Chinese newspapers designated for the f ormula incident s. Some other words were tested as search terms in the pre test of the study. For example, infant formula was tried as keywords. But the search with infant formula generated too many stories that were unrelated with the tainted formula scandals Sanlu, the company name of the biggest tainted formula producer in the scandals was also tested. However,
92 Sanlu was, although the biggest, not the only dairy company involved in the scandals The nationwide investigations of infant formula carr ied out by the government in later 2008 revealed that many dairy companies in China had illegally added melamine to their products. Consequently, newspaper coverage of the scandals after the investigation was not only about Sanlu. Using Sanlu as the keyw ord for search m ight have result ed in an incomprehensive search of qualified formulascandal related stories. All the articles identified through searche s on the selected newspapers online archives from May 1, 2008, to June 1, 2011, were indexed as the p opulation of the study. However, stories with less than 400 words, story compilations, duplicates, and Q&As were excluded from the final population for sampling, because such stories were not long enough to examine the standing of the newspapers on the iss ue or would cause sampling problems. Consequently, there were 77 articles about tainted formula scandals from the Peoples Daily in the final population, 72 from the Beijing Youth 83 from the Southern Metropolitan News 102 from the Oriental Morning Post 68 from the Xinmin Evening News 76 from the Changjiang Daily 64 from the Chutian Metropolis Daily and 65 from the Hebei Daily Then, 50 articles were randomly selected from the population of each newspaper. The final sample for newspaper analysis was 4 00 articles from eight major Chinese newspapers, which allow ed the researcher to build a comprehensive outlook on newspaper coverage on the milk formula scandals The unit of analysis was each news story.
93 Forum Sampling Five online public forums were selected for this study : Tianya Forum ( http://www.tianya.cn/) Sin a Forum ( http://bbs.sina.com.cn ) S ohu Forum ( http://club.sohu.com ) Qiangguo Forum ( Qiangguo is literally strengthening the nation in Chinese) ( http://bbs1.people.com.cn/ ), and Xinhua Forum ( http://ww w.xinhuanet.com/forum/index.htm ) (Table A 2). Launched in March 1999, Tianya had more than 30 million registered users by early 2010 and ranked 18th in the countrys most visited websites (from Alexa.com). Attracting more than 1.5 million pages views per day, Tianya is among the most popular online public forums in China, although it is always filled with random talks. Tianya forum has generated many controversial and influential online discussions, attracting huge public attention and media exposure. Moreover, Tianya was the first online media platform on which the Sanlu formula scandal was disclosed, causing a surge of online discussions over the scandals The popularity and its swift reaction to the incident s made it an ideal subject for this study. Sin a and Sohu are among the first and the most popular market oriented commercial Web portals, similar to Yahoo in the United States (Harwit & Clark, 2001). They ranked 4th and 8th on the list of most visited websites in China (from Alexa.com), having features such as email, news, and public forums. All three online public forums mentioned above are independent, while the last two are attached to news organizations. Established by the Peoples Daily C hina C ommunist P arty s party newspaper, one day after NATO s bombing of the Chinese Embassy in 1999, Qiang g uo Forum is the most famous and probably the most popular political discussion forum (Yang, 2003). Its loosely controlled
94 forum atmosphere, the endorsement it has from the Peoples Daily and its focus on cur rent domestic and foreign events have made the forum a hotbed for Chinese Internet users (Li, Xuan & Kluver, 2003; Harwit & Clark, 2001). Hosted by Xinhua News Agency, the major news agency in China, Xinhua Forum is another popular online public forum for Chinese. Using the search terms tainted formula ( wenti naifen, in Chinese : ) and poisoned formula ( du naifen, in Chinese: ), threads in online discussion forums from May 1, 2008 to June 1, 2011, that had a theme of the tainted formula incident s were used as the population of forum posts sample. Off topic threads, duplic ates, and threads with fewer than 10 posts were excluded from the sample to allow for indepth analysis. Consequent ly, 7797 threads from Tianya forum, 9 84 threads from Sina forum 8 63 threads from Sohu forum, 353 threads from Qiangguo forum, and 179 thread s from Xinhua forum were included in the population for sample selection. The final sample for forum study was 200 threads from the five discussion forums, with 40 threads randomly selected from each one. The unit of analysis was each thread with at least 10 posts. Blog Sampling The final sample of blogs consists of 80 blogs, with 20 blogs maintained by health experts 20 by parents 20 from journalists a nd 20 from citizens. Generally speaking, health expert blogs are blogs that deliver health information and comment s on health issues (i.e. the tainted formula incident s) from a professional health perspective, and are written by health experts such as doctors, pediatricians, etc. Parent blogs refer to those blogs maintained by individuals
95 who self identify as parents and commonly made their children or parenting the focus of their blogs. They wrote about the tainted formula on the basis of their parent status. Journalist blogs are online diaries or journals maintained by professional reporters to post news or commentary about current events (i.e. the tainted formula scandals ) on either newspaper websites or blog portals. Citizen blogs are blogs published independently by citizens for self presentation, self expression, and conversations with others. For this study, e ach qualified blog had to have start ed prior to May 1, 2008, when the first formula scandal was revealed online by parents. Also, each blog had to include more than five blog posts with a theme of the tainted milk formula incident s f rom May 1, 200 8 to June 1, 2011. Only blog posts with more than 50 words were used to allow for indepth and inclusive analysis. For the selection of health expert blogs, both the popularity of the blogs and the credibility of health expert bloggers were considered. Wh en choosing blogs from parents, journalists and citizens, both the popularity of the blogs and the depth of blog coverage on the tainted formula scandals were taken into consideration. There is no directory of blogs based on popularity or area available. Given the huge number of Chinese blogs maintained by health experts, parents, journalists, and citizens, as well as the aim of this research to study the tainted milk formula coverage in the most popular blogs or at least in blogs with a big audience base, th e search of qualified blogs was conducted in several ways. First, there were several major blog portals in China including Sina Blog, Sohu Blog, NetEase Blog, Hexun Blog, etc. Some of them such as Sina Blog and
96 NetEase Blog have lists of the most popular/most r ead blogs on their websites. So the 300 most popular blogs were pooled from the two blog portals individually. Among those most read and highly rated Chinese blogs, the keywords tainted formula ( wenti naifen in Chinese : ) and poisoned formula ( du naifen, in Chinese: ) were used to search within each blog to decide whether it would be a qualified blog. In addition, the blog roll a list of blogs that the blogger frequently read or link toon those most popular blog s were also examined to find suitable blogs for the study. Then, for the other two big blog portals, Hexun Blog and Sohu Blog, although there were no rankings of popular blogs on their websites, they recommended blogs for readers and named top bloggers for different sections such as politics, economics, health, lifestyle, entertainment, etc. For example, Sohu listed many blogs maintained by famous doctors on the websites health section and promoted them on the front page. After a keyword search using tainted formula ( wenti naifen, in Chinese: ) and poisoned formula ( du naifen, in Chinese: ) within each recommended health blog, parent blog, journalist blog or citizen blog on the two sites, qualified blogs were pooled into the population. Moreover, f or parent blogs, some parenting websites were examined to find more qualified parent blogs such as blogs ( http://blog.ci123.com ) on Yuer (literally parenting in Chinese: ), a website specifically providing parenting information. Journalist blogs are sometimes hard to recognize on blog portals unless the j bloggers identify themselves as journalists To make the study more comprehensive, the researcher examined journalist blogs from the websites of
97 the eight newspapers selected for this study (if they had any on the sites), using the keywords tainted formula ( wenti naifen in Chinese: ) and poisoned formula ( du naifen in Chinese: ) to search for qualified journalist blogs. Admittedly, there was no way to tell which journalist bl ogs had the largest following. Thus, those journalist blogs that started earlier with a higher number of posts in the ir history or with a higher number of posts concerning the tainted milk formula scandals were given the priority during the selection. Through the methods mentioned above, 68 health expert blogs, 77 parent blogs, 72 journalist blogs, and 84 citizen blogs were chosen to serve as the population for th e study. Twenty blogs were randomly selected from the population of each blog type, resulting in 80 blogs for analysis. Then, five posts were randomly chose n from each of the 80 blogs. So the f inal sample for the blog study was 400 blog posts. The unit of analysis was each blog post. Overall, the final sample of 400 newspaper articles, 200 forum threads each with at least 10 posts, and 400 blog posts allow ed the researcher to build an extensive understanding of the discourse s surrounding the milk formula incident s in China Variables Based on the literature review and an initial assessment of the Chinese newspaper and blog coverage as well as the public discussions in Chinese online forums concerning the tainted formula scandals the following variables and coding schemes were developed.
98 Story focu s: B ased on the previous literature (Nucci, Cuite, & Hallman, 2009) and a pilot examination of formulascandal rel ated content in Chinese media, the st ory focus variable was comprised of infected symptoms, information on spread of the incident s or infection (infected/deaths/confirmed cases), dairy products identified as contaminated, formula recall, guidelines for handling the tainted formula, guidelines for seeking medication/hospitalization, compensation to victims, n ew regulations and food/milk monitoring system the economic influence of the incident s/recall, health or medical issues related to food safety /formula scandals the politic al influence of the incident s/recall possible causes for the incident s, problems of the food industry or government agencies, potential responsibility takers, actions by government or health agencies, actions taken by dairy companies, actions by individuals, actions taken by hospital, and others (examples for each story focus in Appendix C Appendix D, and Appendix E ) Any of the above foci that appeared in each story/post was coded for presence/absence. A factor analysis was later employed to categorize the 19 story foc i into five constructs (T his proces s will be discussed in Chapter 4 in detail), including 1) a des cription of the formula scandals 2) solutions, 3) influence of the incident s (i nternationally and nationally), 4) investigation of the incident s, 5) action. From the five categories, one was chosen as the major story focus for each story Frame : Based on previous framing research (Dill & Wu, 2009; Luther & Zhou, 2005; Nisbet, Brossard, & Kroepsch, 2003; Shih, Wijaya, & Brossard, 2008) and a pilot examination o f formula scandal related content in Chinese
99 media, frames for this analysis included: 1) an attribution of responsibility frame, which attempts to blame the issue on the government, the dairy companies, or other stakeholders; 2) a severity frame, which reports on the spread of the tainted formula contamination, for example, the infected/deaths/confirmed cases; 3) a conflict frame emphasiz ing conflicts between individuals, companies, victimized groups, and the government; 4) an economic consequence frame, w hich talks about the possible economic losses for the individuals, companies, the dairy industry or the country due to the formula scandals ; 5) a reassurance frame, which tells the public they do not need to worry or should be less worried about the formul a scandals or food safety issues in China; 6) an action frame, which emphasizes any action taken by stakeholders such as the dairy companies, the government or individual s, to deal with the formula scandals including formula recall s, compensation, punishment, hospitalization of sickened children; 7) an uncertainty frame that describes uncertainties in any aspect of the issue, including the cause, the cure, the possible spread; 8) others (examples for coding in Appendix C, Appendix D, and Appendix E) Beca use it is possible that there was no frame in a story, especially in a blog post, 9) none was added to the coding categories. All the frames that appeared in the story/post as well as the most dominant frame in the story/post w ere coded for their presence. Tone : Tone was determined based on whether subjective judgment on the issue appeared in the article and what that major judgment was. For newspapers and blogs tone was categorized as 1) toward government: negative, neutral/mixed, positive, or no tone (vo id of judgment, only statements of facts ) ; 2)
100 toward the diary company: negative, neutral/mixed, positive, or no tone (Sha nnon, 2006) In each category, only one option was allowed. For online forum threads, the number of posts (original entry and responses) in each thread that qualify for each tone category favorable to government, unfavorable to government, favorable to dairy companies, unfavorable to dairy companies, neutral/mixed opinion(s) no tone presented was counted. Source : Sources were classified into 11 types: 1) state/local government/agency or its officials; 2) health agencies or their officials such as the Chinese Center for Disease Control and Prevention (CCDC), the State Food and Drug Administration of China (SFDA), the General Administrat ion of Quality Supervision, Inspection and Quarantine (AQSIQ), the Ministry of Health, or other food or health related institutions; 3) nongovernment organizations (NGOs), o ther independent organizations, or their officials; 4) health experts such as doct ors/nurses/pediatrician, professors/researchers, food/nutrition experts, etc. that were not associated with the three types of agencies mentioned above; 5) hospitals or their officials; 6) dairy companies, their spokespersons or lawyers, etc.; 7) victims parents, families, lawyers, etc.; 8) other citizens, witnesses, etc.; 9) newspapers, TV stations, magazines, or news agencies such as the Xinhua News Agency, and their websites; 10) Blogs, online forums, independent online news sites, or other citizen medi a; 11) others. Coders count ed the number of sources from each category in each newspaper article, blog post, and forum thread. If a source, such as a pediatrician, appeared several times in a story, it would be counted only once.
101 Responsiveness of discussi on: For blogs, the responsiveness of discussion was measured by the interactivity between bloggers and readers M ore specifically, it was measured by the number of reader comments and bloggers responses to each blog post as well as the ratio of reader com ments to bloggers responses For online forum threads, the responsiveness of discussion was measured by the total number of replies in a thread. A related concept with responsiveness of discussion was the unresponsiveness of discussion. P eriphery particip ation ( Wu, 2008) was used to show unresponsiveness of discussion, by calculating the ratio of hits and replies for each thread. Number of hits and replies can be found in each thread on all the Chinese online forums selected for this study. Homogeneity of discussion : For blogs, homogeneity of discussion was measured by the number of reader comments in each of the following categories, including 1) comments supportive or similar to viewpoints in the blog post; 2) comments that mention or propose different or opposite idea(s); 3) neutral comments that neither support nor oppose, or mixed comments that both support and oppose; 4) off topic comments such as ads; 5) others. For online public forums, based on previous studies (Zhou, Chan, & Peng, 2008) homogeneit y of discussion was about how the followingup posts replied to the initial entry in each thread, even if it was not a direct reply to the initial entry (i.e. it replied to the second or third entry). So, the homogeneity of discussion in Chinese online for ums was measured by the number of follow up posts or replies in each of the following categories : 1) pos ing any argument(s) to support the viewpoints in the
102 initial entry; 2) propos ing any different or opposite idea(s) to the initial entry; 3) mention neut ral or mixed points of view; 4) bring ing in new information/topic that is related to whats in the initial entry; 5) off topic replies such as ad; 6) others. Multimedia : In online forums and blogs, the use of multimedia was measured by the number of embedded videos, embedded audio, and embedded photos in each thread and each blog post. Hyperlink : In online forums and blogs, the number of hyperlinks and the type of hyperlinks were measured. Based on their destinations, hyperlinks were identified as : 1) inter nal hyperlinks that was to the site hosting the blog post/thread in analysis ; 2) external hyperlinks to legacy news sites (newspaper TV, news agency); 3) external hyperlinks to online news sites (Sina, Sohu, etc.) and citizen journalism sites (blogs, onli ne forums, etc.) ; 4) external hyperlinks to government/health agency sites; 5) external hyperli nks to others (please specify). In addition, hyperlinks were categorized based on the form of hyperlinked content, including 1) hyperlinks to only multimedia content (i.e. photos, videos, audio); 2) hyperlinks to only texts; 3) hyp erlinks to multimedia and texts together. The number of hyperlinks in each type was counted for every blog post and forum thread. Gender : Gender of the writer s may have an influence on t he stories of tainted milk formula. In most cases, gender of news authors can be judged based on their names. For bloggers, sometimes they reveal their gender in the profiles, pictures, or blog content. However, in some cases there is no clue to the gender of the reporter or the blogger. So, in addition to male and female, one category
103 unknown (cannot discern) or no author listed was added to analysis. For newspapers, sometimes their stories are from news agencies (i.e. the Xinhua New Agency ) Thus, another category news agency was listed for newspaper analysis. In addition, there could be several authors in a newspaper articl e. In that case, only the gender of the first author was coded. For discussants in online forums, it is very difficult to tell their g ender. So, online forums were excluded from coding on this variable. Intercoder Reliability In spring 2011, two graduate students in a southern university who are fluent in Chinese coded 20% of newspaper articles, forum threads, and blog posts in the final sample that were randomly selected to develop intercoder reliability. One of the coders was the researcher. A pretest of the codebook, a training session and a trial coding were carried out first. D isagreements in the trial coding were then discussed betw een the coders and the codebook was improved. After that two coders each coded 80 newspaper stories, 40 forum threads, and 80 blog posts. The intercoder reliability was evaluated using Krippendorffs alpha, for which a coefficient of .80 or greater is ac ceptable in most cases ( Neuendorf, 2002). The reliability for each variable was reported in Table B 1.
104 CHAPTER 4 RESULTS In this chapter, the r esults of data analysis are reported so as to answer all the research questions. Descriptive statistics, chi squ are s, T test, and one way ANOVA analyses were performed and explained. In total, 400 newspaper articles, 400 blog posts, and 200 forum threads were coded. Most news paper stories (79.8%) were news reporting articles; 5% of them were letters from readers; 8% were opinion columns by staff journalists or other people; 4% were editorial s; and 3% were feature articles. There were several periods of time in which the number of newspaper articles about tainted formula was higher. For example, from August 2008 to J anuary 2009 right after the break out of the 2008 Sanlu tainted formula scandal Or from late 2009 to August 2010 when the illegal reuse of material from the 2008 adulterations as well as the dairy products polluted by a carcinogenic protein derived from leather were revealed. It seems that news paper coverage was fairly event driven. Among the 400 blog posts about tainted formula scandals 100 were from health experts, 100 from parents, 100 from journalists, and 100 from citizens. For the 200 forum threads coded, most of them (44.0%) were from the talk/conversation/gossip section in the forums; 13.5% were from the news section; 13.0% were from the politics/public affair s/law section; 8.0% from the health section, 7.0% from the parenting/education section, 7 .0% from the economics section; few of them were from the region section (3.0%) or food/lifestyle section (1.0%). All the forum threads were selected from the five popular Chinese online forums, among which Tianya was the one with the most
105 active discussio n about the tainted formula scandals while Xinhua the online forums attached to the news agency was the one with the least active discussions. Moreover, in blogs and discussion forums, it was found that, similar to newspaper coverage, when the tainted for mula scandals were first revealed, the discussion in blogs and forums were the most vigorous. Story Focus The first set of research questions probed for the story foci and the major focus of tainted formula coverage in newspapers, blogs and online discuss ion forums. More specifically, these research questions looked at what attributes of the issue t he tainted milk formula scandals were covered as story foci in newspapers, discussion forums, and blogs ; and what story focus was emphasized as the dominant the me or the major focus in each story. T o classify the story foci in the coverage of the tainted formula incidents, a principal axis factor analysis with varimax rotation on the 19 items was conducted As a result, the initial items were first integrated into six summative constructs. The rotated component matrix of constructs with respective items for story foci i s listed in Table 3 1. Cronbachs alpha and corrected item total correlations were calculated to determine the internal consistency of each constr uct emerging from the factor analysis. For description, Cronbachs alpha was .77; for solutions, Cronbachs alpha was .70; for influence, Cronbachs alpha was .77; for investigation, Cronbachs alpha was .95, thus these four constructs were o ver the standard of .70 for acceptable reliability (Babbie, 2007). For the remaining items, the corrected item total correlation was .423 for actions by government/health agencies and actions by dairy companies and .415 for
106 actions by individuals and actions by hospital, exceeding the satisfactory level of .40 (Babbie, 2007). However, most actions mentioned in the tainted formula stories were initialized by the government, health agencies, or dairy companies while only few stories had actions by individuals an d actions by hospital as story foci So the last two items were dropped for later analysis. Consequently five story foci were retained, including a description of the tainted formula incidents, solutions, influence of the incidents, investigation of the incidents, and actions, explaining a total of 59.16% of the variance (Table 31). As shown in Table 32, the most frequent focus was actions (81.8%), followed by incident description (48.3%), solutions (46.0%) and investigation (41.3%). The influence of t he tainted formula incidents was the least covered in newspapers (35.0%). In blogs, actions appeared most frequently (69.8%), followed by description (57.8%), influence (47.5%), investigation (42.5%), and solutions (33.0%). Similar to blogs, online discuss ion forums also had actions as the most presented story focus (73.0%), followed by description (64.0%), investigation (52.5%) and influence (32.0%). Solutions (28.0%) was the least likely to appear in forum discussion about tainted formula incidents. A chi square analysis revealed that there were significant differences across newspaper s, blog s, and forum s for the presence of each story focus (Table 32). Although action was the most frequently appearing story focus among media, it seems that newspapers in China were more likely to talk about it than other media. Newspapers were also more likely to mention solutions in their coverage of the tainted formula, but least likely to describe the tainted formula
107 incidents, compar ed to blogs and online forums. In co ntrast, incident description and investigation were more likely to be included in forum discussions. Blogs were found to talk more about the influence of the scandals. Table 31. R otated component matrix for f actor analysis of story foci in the coverage of the tainted formula incident s Story F ocus Component 1 2 3 4 5 6 A description of the incidents Infected symptoms .635 Spread of the incident s .754 Dairy products contaminated .757 Solutions Formula recall .775 Guidelines for handling the tainted formula .774 Guidelines for seeking medication/hospitalization .407 Compensation .588 New regulation or system .560 Influence Economics of the incident s .727 Health issues of the incident s .7 51 Politics of the incident s .666 Investigation Possible causes of the incident s .922 Problems of the food industry/government .895 Potential responsibility taker(s) .925 Actions Actions by government/health agenci es .805 Actions by dairy companies .644 Actions by individuals 188 Actions by hospital 5 73 Variance Explained 8.77% 17.17% 8.99% 17.20% 7.03% 4 .73% Eigenvalues 1.53 1.88 1.62 3.10 1. 27 1.03 Extraction method: principal axis factoring. Rotation method: varimax with Kaiser Normalization. Rotation converged in 6 iterations. Table 32. Story focus in the coverage of the tainted formula incidents Focus Newspaper (N=400) Blog (N=400) Forum (N=200) (df=2) p Presence Percent Presence Percent Presence Percent Actions 327 81.8% 279 69.8% 146 73.0% 16.09 .000 Description 193 48.3% 231 57.8% 128 64.0% 15.13 .001 Solutions 184 46.0% 132 33.0% 56 28.0% 23.53 .000 Investigation 165 41.3% 170 42.5 % 105 52.5% 7.46 .024 Influence 140 35.0% 190 47.5% 64 32.0% 18.82 .000 E ach focus that appeared in each story was identified for presence/absence and analyzed. It was possible for one story to have all five story foci However,
108 with the above analys is only, it is hard to tell which story focus is the most important one or the theme of the article/post. To provide more indepth analysis, the major focus in each story as the theme of the article/post was identified and analyzed. Table 33 shows that, i n newspapers, action (33.8%) appeared most frequently as the major focus, followed by investigation (23.1%), solutions (16.8%), influence (13.3%), and description (13.3%). More blogs adopted investigation (28.2%), actions (20.6%), and influence (20.5%) as the major focus, while fewer blogs mainly focused on solutions (16.0%) and description (14. 8%). More forum threads emphasized investigation (36.0%) and actions (25.0%) as the major focus. However, forum discussions were less likely to mainly focus on incid ent description (16.5%), influence of the tainted formula (11.5%), or solutions (10.5%) (Table 33). Table 33 M ajo r s tory focus in the coverage of the tainted formula incident s Major Focus Newspaper Blog Forum p Frequency Percent Frequency Percent Frequency Percent Actions 135 33.8% 82 20.6% 50 25.0% 18.31 .000 Investigation 92 23.1% 113 28.2% 72 36.0% 11.35 .003 Solutions 67 16.8% 64 16.0% 21 10.5% 4.37 .112 Influence 53 13.3% 82 20.5% 23 11.5% 11.38 .003 Description 53 13.3% 59 14.8% 33 16.5% 1.52 .469 Total 400 100.0% 400 100.0% 200 100 .0 % 36.21 .000 Comparing the major focus in tainted formula coverage among newspapers, blogs, and forums, it was revealed that there were significant differences for actions ( =18.31, p <.001, df=2), investigation ( =11.35, p <.01, df=2), and influence ( =11.38, p <.01, df=2). Table 33 shows that newspapers were more likely to emphasize actions (33.8%) as the major focus than blogs (20.6%) and forums (25.0%); online forums wer e more likely to focus on investigation (36.0%) than blogs (28.2%) and newspapers (23.1%); blogs were
109 more likely to concentrate on influence (20.5%) when compar ed to newspapers (13.3%) and forums (11.5%). Frame The second set of research questions examined the frames and the dominant frames employed in the tainted formula scandals comparing newspapers and blogs (T here w ere no frame s in online discussion forums ) Frequency and chi square analyses were performed to answer the questions. As shown in Table 34 the most frequently presented frame s in newspapers were the action frame (87.5%), severity frame (58.3%), conflict frame (52.0%), and attribution of responsibility frame (42.3%), while the economic influence frame (18.8%), reassurance frame (17.3%), an d uncertainty frame (10.3%) were less likely to appear in newspaper coverage of the tainted formula scandals. In blogs, the most frequently appearing frames were the conflict frame (74.8%), severity frame (72.5%), action frame (71.5%), and attribution of r esponsibility frame (52.0%), followed by the uncertainty frame (16.3%), economic influence frame (10.0%), and reassurance frame (9.3%) (Table 34). Table 34. Frames in the coverage of the tainted formula incidents Frame Newspaper (N=400) Blog (N=400) ( df=1) p Presence Percent Presence Percent Action frame 350 87.5% 286 71.5% 31.42 .000 Severity Frame 233 58.3% 290 72.5% 17.94 .000 Conflict frame 208 52.0% 299 74.8% 44.60 .000 Responsibility Frame 169 42.3% 208 52.0% 7.63 .006 Economic frame 75 18.8% 40 10.0% 12.44 .000 Reassurance frame 69 17.3% 37 9.3% 11.1 4 .001 Uncertainty frame 41 10.3% 65 16.3% 6.26 .012 Others 5 1.2% 2 0.5% --Note: If any cell is less than 5, the chi square analysis is not performed for the variable.
110 Significant differences were revealed between newspapers and blogs in the frames they adopted to cover the tainted formula scandals Newspapers were significantly more likely to employ an action frame, economic influence frame, and reassurance frame. Blogs were more likely to use the severity frame, conflict frame, attribution of responsibility frame, and uncertainty frame. To better understand the coverage of tainted formula scandals the most dominant frame in each newspaper article or blog post was discerned. It was found that action was not only the most frequently appear ing frame but also the most dominant frame in newspapers (30.5%), followed by the attribution of responsibility frame (20.5%), severity frame (13.8%), reassurance frame ( 13.8%), conflict frame (13.8%), economic influence frame (4.3%), and uncertainty frame (3.8%) (Table 35). For blogs, the most dominant f rame was the conflict frame (26.8%), followed by attribution of responsibility frame (23.3%), and severity frame (20.5% ). A ction frame (15.8%), reassurance frame (5.5%), economic influence frame (4.3%), and uncertainty frame (4.08%) were less likely to be the most dominant frame in blogs (Table 35). Table 35. Dominant frame in the coverage of the tainted formula incidents Dominant Frame Newspaper Blog p Frequency Percent Frequency Percent Action frame 122 30.5% 63 15.8% 24.48 .000 Responsibility Frame 82 20.5% 93 23.3% .89 .347 Severity Frame 55 13.8% 82 20.5% 6.42 .011 Conflict frame 53 13.3% 107 26.8% 22.78 .000 Reassurance frame 55 13.8% 22 5.5% 15.65 .000 Economic frame 17 4.3% 17 4.3% .00 999 Uncertainty frame 15 3.8% 16 4.0% .03 .855 Others 1 0.3% 0 0 --Total 400 100.0% 400 100.0% 58.23 .000 To examine the framing differ ences between newspapers and blogs, a chi square analysis was conduct. It was found that there were significant differences
111 for the action frame ( = 24.28 p <.001, df= 1 ), severity frame ( = 6.42 p <.0 5 df= 1 ), reassurance frame ( = 15.65, p <.001, df= 1 ), and conflict frame ( = 22.78 p <.001, df= 1). Table 3 5 shows that action frame and reassurance frame were more likely to appear as the most dominant frame in newspapers, while blogs were significantly more likely to include severity frame and conflict frame as the dominant frame. Tone The thi rd s et of research questions looked at the tones of the tainted formula coverage in newspapers, blogs and onli ne forums. Frequency and chi square analyses were conducted for newspaper articles and blog posts. Mean value analysis was used for forum analysis. Table 36 shows that most newspapers articles did not appear to have a tone toward the government ( 78.0%) o r toward dairy companies (71.8%) when covering the tainted formula scandals Similarly, most blog posts did not appear to have a tone toward the government (56.3%) or toward dairy companies (48.5%). In total, 61.0% of newspaper stories and 36.1% of blog po sts did not appear to have a tone toward both the government and diary companies. Most of the remaining newspaper articles or blog posts were negative toward government or dai ry companies (Table 36). There were a limited number of newspaper articles and blog posts with neutral/mixed tone to either government or dai ry companies such as Sanlu, Shengyuan, etc. Virtually no stories or blogs posts were coded as including positive tones to ward the government or dai ry companies.
112 Similarly, in online forums, more than half of the posts (53.82%) in 200 threads did not appear to convey a tone toward either government or dairy companies. The remaining forum sample was mostly split with 24.35% negative toward government and 20.13% negative in tone toward dairy compani es. T he mean number of posts in each thread that had no tone at all was the highest ( M =37.24, SD =89.27), followed by the posts with negative tone toward the government ( M =16.85, SD =48.53), and negative tones toward dairy companies ( M =13.93, SD =40.11). The mean number of posts in each thread that had a positive tone toward the government ( M =.55, SD =3.73) or a positive tone toward dairy companies ( M =.05, SD =.46) w ere the lowest (Table 37). Table 36 Tones in the newspaper and blog coverage of the tainted f ormula incident s Tone Newspaper (N=400) Blog (N=400) p Frequency Percent Frequency Percent Tone toward government 47.78 .000 Negative 70 17.5% 156 39.0% 45.61 .000 Positive 4 1.0% 2 0.5% --Neutral/mixed 14 3.5% 17 4.3% .30 .583 No tone 312 78.0% 225 56.3% 42.87 .000 Tone toward dairy companies 47.58 .000 Negative 103 25.8% 195 48.8% 45.26 .000 Positive 1 0.3% 0 0 --Neutral/mixed 9 2.3% 11 2.8% .21 .651 No tone 287 71.8% 194 48. 5% 45.09 .000 Tone toward both the government and diary companies No tone 244 61.0% 144 36.0% 50.05 .000 Table 37 Tones in the coverage of the tainted formula incident s in online forums Tone (N=200) Sum* Sum in proportion M SD Negative toward gove rnment 3370 24.35% 16.85 48.53 Positive toward government 110 .79% .55 3.73 Negative toward dairy companies 2786 20.13% 13.93 40.11 Positive toward dairy companies 10 .07% .05 .46 Neutral/mixed 116 .84% .58 2.90 No tone 7448 53.82% 37.24 89.47 *Sum is the total number of posts in 200 forum threads
113 For newspapers and blogs, tone was coded by identifying the presence of tone in each story, while for forum threads, tone was coded by the number of posts in each thread that present ed different tones, so there is no way to compare forums to the other two media types Table 36 shows the tone differences between newspapers and blogs. It was found that newspapers were significantly more likely to present no tone toward the government (78%) and dairy companies (71.8%) than blogs. Compared to newspapers, blogs were significantly more likely to show negative tones toward the government (39.0%) and dairy companies (48.8%). However, due to the small number of positive tone sto ries and neutral tone stories, no statistical analys e s could be conducted on these variables without violating analytical assumptions. Source The fourth set of research questions examined the sources used in the tainted formula coverage in newspapers, bl ogs and online forums. Mean value was calculated to find out what sources were used frequently. Oneway ANOVA analysis was conducted for comparison of sourcing differences among media. Table 38. One way ANOVA e xamining sources used in the coverage of the tainted formula incident s Source Newspaper (N=400) Blog (N=400) Forum (N=200) F (df=2) p Sum M SD Sum M SD Sum M SD Government 324 .81 ab 1.00 110 .28 a .74 72 .36 b .78 42.30 .000 Health agency 239 .60 a .78 121 .30 ab .67 99 .50 b 1.06 13.71 .000 NGO 32 .08 .29 27 .07 .32 12 .06 .31 .33 .722 Expert 57 .14 .48 71 .18 .67 37 .18 .69 .47 .625 Hospital 15 .04 .30 26 .06 .40 16 .08 .32 1.16 .313 Dairy company 89 .22 .66 65 .16 .51 38 .19 .61 1.03 .359 Victim 49 .12 .54 51 .13 .51 27 .13 .51 .04 .962 Citizen 59 .15 a .53 108 .27 .86 104 .52 a 2.71 4.95 .007 Traditional media 72 .18 a .45 213 .53 b .91 171 .85 ab 1.27 43.71 .000 Online media 22 .06 ab .57 100 .25 a .87 50 .25 b .69 8.66 .000 Others 53 .13 .53 31 .08 .33 13 .07 46 2.20 .111 ab Sources with the same superscript are significantly different at the p < .05 level.
114 Table 38 shows that, for newspapers, government sources ( M =.81, SD =1.00) and health agency sources ( M =.60, SD =.78) were most frequently cited. For blogs, traditional media ( M =.53, SD =.91), health agency ( M =.30, SD =.67), government ( M =.28, SD =.74), citizens ( M =.27, SD =.86), and online media ( M =.25, SD =.87) were heavily cited as sources. For forums, traditional media ( M =.85, SD =1.27), citizens ( M =.52, SD =2.7 1), health agencies ( M =.50, SD =1.06), government ( M =.36, SD =.78), and online media ( M =.25, SD =.69) were often cited as sources (Table 38). One way ANOVA analysis revealed that there were significant differences across newspapers, blogs, and forums in thei r use of government sources ( F= 42.30, p <.001, df= 2 ) health agency sources ( F= 13.71, p <.001, df= 2 ) citizen source s ( F= 4.95 p <.01, df= 2 ) traditional media ( F= 43.710, p <.001, df= 2 ) and online media ( F= 8.66 p <.001, df= 2 ) Post hoc test s u sing Scheffes show ed that newspapers ( M =.81, SD =1.00) were significantly more likely to use g overnment sources than blogs ( M =.28, SD =.74) and online forums ( M =.36, SD =.78). Blogs ( M =.30, SD =.67) were significantly less likely to use health agency sources than newspapers ( M =.60, SD =.78) and online forums ( M =.50, SD =1.06). Online forums ( M =.52, SD =2.17) were significantly more likely to use citizens as sources than were newspapers ( M =.15, SD =.53). Newspapers ( M =.06, SD =.57) were less likely to use online media as sources th an blogs ( M =.25, SD =.87) and forums ( M =.25, SD =.69). In contrast, online forums ( M =.85, SD =1.27) and blogs ( M =.53, SD =.91) both cited traditional media more frequently as a source ( T able 3 8).
115 Further, oneway ANOVA analysis revealed that there were signi ficant sourcing differences among four different forums in terms of their use of traditional media ( F=3.70, p <.05, df=4 ) Tianya was the forum that used the highest number of traditional media sources ( M =1.02, SD =1.87) and the highest number of hyperlinks to multimedia content ( M =.51, SD =.76) among the four. Pattern of Participation The fifth set of research questions probed into the pattern of participation in new media, namely, blogs and online discussion forums. In the blog sample, seven posts were found forbidding reader comments. After excluding these posts from analysis, Table 315 shows that the average number of time s a blog post about tainted formula scandals was read was 3 413.04, while the average number of reader comments to each post was 26.54. In forums, the average time that a thread about tainted formula incident s was hit was 9 875.86, while the average number of replies was 67.74. To answer RQ5a about what stories generated more discussion, an independent sample T test was conducted to find out the influence of story focus and frame on peoples participation in online discussions In other words, whether stories with a specific story focus or frame would generate a higher number of times that readers read the article / post, a higher number of reader commen ts or a higher number of supportive/ opposite/neutral/mixed/ new information/ off topic comments were explored through T test analyses. For story focus, the results showed that the appearance of incident influence in the blog post elicited sig nificantly more neutral/mixed comments from readers ( M =.72, SD =3.26); when action appeared as the story focus in blogs,
116 there were significantly fewer opposing comments ( M =.38, SD =1.83) but more neutral comments ( M =1.08, SD =4.05) and more blogger responses ( M =.39, SD =1.44) (Table 39). In addition when investigation of the tainted formula scandals was the major theme of the blog posts, there were significantly more reader comments ( M =46.40, SD =177.20), more supportive comments ( M =31.90, SD =125.90), and mor e comments from readers providing new information ( M =10.59, SD =48.38) (Table 310). For frame, Table 311 shows that blog post s with an action frame would be less likely to have oppos ing comments ( M =.40, SD =1.82), neutral/mixed comments ( M =.28, SD =1.33), o r blogger responses ( M =.41, SD =1.66). Interestingly, blogs with a reassurance frame included more blogger responses ( M =1.67, SD =5.00) (Table 311). For dominant frame, no significant influence w as found except that blogs with reassurance as the dominant fr ame had more blogger responses ( M =2.29, SD =6.33) (Table 312). For the forum sample, the independent sample T test revealed that forum post s describing solutions or influence of the tainted formula scandals had more neutral comments from readers (Table 31 3). When investigation of the tainted formula incident s was discussed in the forums, users left significantly more replies ( M =96.58, SD =197.83) especially more supportive replies that showed agreement or at least similar viewpoints to the original post ( M =33.57, SD =51.29) (Table 314).
117 Table 39. Independent sample T test of story focus and reader participation in blogs Description Solution Influence Appeared (N= 225 ) Not (N= 168 ) Appeared (N= 131 ) Not (N= 262 ) Appeared (N= 185 ) Not (N= 208 ) M SD M SD M SD M SD M SD M SD Read times 2949.07 9858.36 4034.43 18983.09 4198.13 17994.92 3020.50 12355.78 3735.83 16976.06 3125.94 11836.28 Reader comment 23.99 56.46 29.96 146.98 24.37 72.16 27.63 118.22 25.44 73.77 27.52 126.71 Supportive 15.89 35.82 21.61 109.64 16.66 57.79 19.18 84.49 16.79 53.26 19.71 92.61 Opposite .94 6.12 .39 1.74 .63 2.30 .75 5.62 1.01 6.54 .44 2.22 Neutral/mixed .51 2.89 .35 1.41 .52 1.75 .40 2.63 .72 a 3.26 .19 a 1.03 Off topic 1.58 3.13 1.98 4.58 2.31 5.20 1.47 2.86 1.82 3.81 1.69 3.83 New info 4.74 15.80 5.63 36.79 5.82 32.03 3.72 10.30 4.68 16.03 5.51 33.67 Blogger response .63 2.25 .57 2.70 .68 2.80 .57 2.26 .58 2.34 .63 2.55 a. t = 2.21, df=391, p < .05 Table 39. Continued Investigation Action Appeared (N= 166 ) Not (N= 227 ) A ppeared (N= 275 ) Not (N= 118 ) M SD M SD M SD M SD Read times 3465.39 13492.60 3374.76 15172.15 3077.63 15134.93 4194.72 12808.86 Reader comment 35.14 146.10 20.25 58.83 25.84 117.75 28.18 67.04 Supportive 24.18 103.80 14.07 47.53 18.94 88.08 16.93 38.31 Opposite .90 6.88 .57 2.22 .38 b 1.83 1.47 b 8.22 Neutral/mixed .63 3.33 .30 1.27 1.08 c 4.05 .16 c .89 Off topic 1.55 2.68 1.89 4.47 1.51 3.39 2.32 4.63 New info 7.84 39.84 3.13 8.91 4.87 29.32 5.71 19.91 Blogger response .84 2.60 .43 2.32 .39 d 1.44 1.1 1 d 3.86 b. t = 2.07, df=391, p < .05; c.t = 3.58, df=391, p < .001; d. t = 2.70, df=391, p < .01
118 Table 310. Independent sample T test of major focus and reader participation in blogs Description Solution Influence Major focus (N= 55 ) Not (N= 338 ) Major focus (N= 63 ) Not (N= 330 ) Major focus (N= 82 ) Not (N= 311 ) M SD M SD M SD M SD M SD M SD Read times 1176.22 3229.10 3777.02 15520.63 5674.79 24071.90 2981.25 11792.65 2296.72 9287.56 3707.38 15548.77 Reader comment 11.80 29.21 28.94 112.50 27.13 92.31 2 6.43 107.43 16.54 26.93 29.18 117.17 Supportive 7.85 21.98 20.04 81.98 18.38 76.12 18.33 76.76 10.37 18.07 20.44 85.48 Opposite .44 2.35 .75 5.06 .51 1.98 .75 5.14 .76 2.04 .69 5.26 Neutral/mixed .11 .57 .49 2.54 .43 1.15 .44 2.54 .49 2.05 .43 2.45 Off topic .87 1.96 1.89 4.02 2.98 a 6.32 1.52 a 3.08 1.55 3.40 1.80 3.92 New info 2.60 7.17 5.53 28.76 3.81 10.64 5.37 28.90 3.24 8.53 5.61 29.82 Blogger response .36 .87 .64 2.62 .89 3.79 .55 2.10 .40 1.88 .66 2.58 a. t = 2.82, df=391, p < .01 Table 310. Continued Investigation Action Major focus (N= 111 ) Not (N= 282 ) Major focus (N= 82 ) Not (N= 311 ) M SD M SD M SD M SD Read times 4557.71 16338.17 2962.48 13668.33 2742.49 10148.14 3589.84 15416.23 Reader comment 46.40 b 177.20 18.73 b 53.71 19.11 4 5.79 28.50 115.70 Supportive 31.90 c 125.90 13.00 c 43.26 14.95 36.8 9 19.23 83.99 Opposite 1.23 8.39 .50 2.02 .28 1.78 .82 5.28 Neutral/mixed .79 3.90 .30 1.35 .15 .92 .52 2.62 Off topic 1.81 2.93 1.73 4.12 1.51 3.51 1.81 3.90 New info 10.59 d 48.38 2 .97 d 8.42 2.29 7.14 5.86 29.89 Blogger response .88 2.74 .50 2.33 .38 1.88 .67 2.58 b. t = 2.37, df=391, p < .05; c. t = 2.21, df=391, p < .05; d. t = 2.55, df=391, p < .05
119 Table 311. Independent sample T test of frame and reader participation in blogs Responsibility Severity Appeared (N= 206 ) Not (N= 187 ) Appeared (N= 283 ) Not (N= 110 ) M SD M SD M SD M SD Read times 3136.55 12283.64 3717.62 16574.67 16139.59 959.40 2767.86 8852.65 Reader comment 32.87 132.11 19.57 62.67 30.43 122.08 16.55 31.34 Supportive 22.57 93.92 13.68 50.90 21.37 89.27 10.55 19.42 Opposite .82 6.19 .59 2.38 .74 5.43 .62 2.36 Neutral/mixed .63 3.09 .24 1.12 .43 2.57 .45 1.76 Off topic 1.58 2.71 1.94 4.75 1.67 3.56 1.96 4.42 New info 7.25 36.11 2.78 8.30 5.93 30.92 3.04 10.36 Blogger response .59 2.07 .63 2.82 .53 2.06 .80 3.25 Table 311. Continued Action Conflict Appeared (N= 282 ) Not (N= 111 ) Appeared (N= 98 ) Not (N= 295 ) M SD M SD M SD M SD Read times 3591.73 15721.13 2959.07 10694.46 3555.67 15824.43 2983.70 9315.56 Reader comment 27.75 117.15 23.48 65.17 29.56 119.62 17.45 33.28 Supportive 20.10 87.49 13.86 36.16 21.17 87.39 9.82 21.06 Opposite .40 a 1.82 1.49 a 8.47 .74 5.31 .62 2.58 Neutral/mixed .28 b 1.33 .84 b 3.91 .50 2.65 .26 1.21 Off topic 1.64 3 .66 2.04 4.20 1.64 3.46 2.09 4.75 New info 5.33 29.19 4.59 19.65 5.48 30.43 4.04 9.95 Blogger response .41 c 1.66 1.09 c 3.75 .55 2.03 .79 3.43 a. t = 2.04, df=391, p < .05; b. t = 2.09, df=391, p < .05; c. t = 2.48, df=391, p < .05
120 Table 311. C ontinued Economic influence Reassurance Uncertainty Appeared (N= 40 ) Not (N= 353 ) Appeared (N= 36 ) Not (N= 357 ) Appeared (N= 64 ) Not (N= 329 ) M SD M SD M SD M SD M SD M SD Read times 4336.75 11813.49 3308.37 14750.02 2186.67 3719.78 3536.71 15132.29 2072. 58 5259.80 3673.80 15635.20 Reader comment 31.03 52.51 26.03 109.45 11.03 13.90 28.11 110.02 18.92 35.65 28.02 113.71 Supportive 20.18 36.87 18.13 79.86 5.31 8.28 19.65 80.18 11.38 19.90 19.69 83.18 Opposite .55 1.36 .73 5.01 .67 2.64 .71 4.94 .61 2.9 3 .73 5.06 Neutral/mixed .53 2.25 .43 2.39 .53 1.86 .43 2.42 .45 1.84 .44 2.46 Off topic 2.78 4.76 1.63 3.69 2.56 6.12 1.67 3.51 1.85 4.06 1.25 2.18 New info 7.00 14.37 4.91 27.89 1.86 4.54 5.45 28.09 5.17 13.58 5.11 28.71 Blogger response .25 1.01 .6 5 2.56 1.67 d 5.00 .50 d 2.01 .50 1.25 .63 2.62 d. t = 2.75, df=391, p < .01
121 Table 312. Independent sample T test of dominant frame and reader participation in blogs Responsibility Severity Dominant (N= 92 ) Not (N= 301 ) Dominant (N= 79 ) Not (N= 314 ) M SD M SD M SD M SD Read time 1462.13 2257.96 4009.33 16446.06 1997.49 5250.13 3769.18 15958.69 Reader comment 20.25 35.36 28.47 118.42 18.67 38.011 28.52 115.92 Supportive 15.34 28.10 19.26 86.12 11.11 25.6 9 20.16 84.62 Opposite .38 1.62 .81 5.38 .63 2.55 .73 5.19 Neutral/mixed .67 3.84 .37 1.69 .18 80 .51 2.62 Off topic 1.35 2.31 1.87 4.17 1.65 3.08 1.78 3.99 New info 2.52 6.50 5.91 30.41 4.33 11.09 5.32 29.50 Blogger response .66 2.72 .59 2.37 .41 1.24 .66 2.67 Table 312. Continued Actio n Conflict Dominant (N= 63 ) Not (N= 330 ) Dominant (N= 98 ) Not (N= 295 ) M SD M SD M SD M SD Read time 6053.25 20801.56 2909.00 12895.68 4148.10 20770.58 3145.05 11372.41 Reader comment 45.76 222.80 22.87 60.89 28.00 90.92 26.01 109.88 Supportive 32.83 1 63.01 15.57 43.92 20.08 66.59 17.70 79.99 Opposite .41 2.08 .76 5.13 1.29 8.53 .50 2.14 Neutral/mixed .37 1.51 .45 2.51 .53 2.27 .41 2.41 Off topic 1.86 4.03 1.73 3.78 1.47 3.70 1.85 3.86 New info 10.33 58.95 4.12 14.00 4.51 18.91 5.34 29.20 Blogger response .83 2.86 .56 2.37 .40 1.27 .68 2.76
122 Table 312. Continued Economic influence Reassurance Uncertainty Dominant (N= 40 ) Not (N= 353 ) Dominant (N= 21 ) Not (N= 372 ) Dominant (N= 16 ) Not (N= 377 ) M SD M SD M SD M SD M SD M SD Read time 7395.59 17 574.14 3232.98 14314.52 2670.57 4316.89 3454.95 14837.67 3143.44 8767.18 3424.48 14669.1 2 Reader comment 38.41 63.20 26.01 106.57 11.24 13.85 27.41 107.86 23.81 60.05 26.66 106.58 Supportive 25.29 41.59 18.02 77.79 5.43 9.36 19.07 78.61 12.38 29.55 18.5 9 77.94 Opposite .65 1.69 .71 4.87 .19 .68 .74 4.90 1.06 3.99 .69 4.81 Neutral/mixed .47 1.94 .44 2.39 .10 .30 .46 2.44 .50 2.00 .44 2.39 Off topic 3.59 6.06 1.67 3.68 4.00 7.71 1.62 3.45 1.13 2.34 1.78 3.87 New info 8.41 14.84 4.97 27.24 1.52 4.40 5 .32 27.54 8.63 24.41 4.97 26.94 Blogger response .06 .24 .63 2.50 2.29 a 6.33 .51 a 2.00 .13 .50 .63 2.50 a.t = 3.269, df=391, p = .001
123 Table 313. Independent sample T test of story focus and reader participation in forums Description Solution Influen ce Appeared (N= 128 ) Not (N= 72 ) Appeared (N= 56 ) Not (N= 144 ) Appeared (N= 64 ) Not (N= 136 ) M SD M SD M SD M SD M SD M SD Hits 10574.15 22390.90 8634.47 18616.78 11902.41 22321.97 9087.76 20607.61 11445.56 18147.32 9137.18 22356.24 Replies 72.77 146.85 5 8.79 124.97 77.73 134.57 63.85 141.25 70.64 122.74 66.38 146.74 Supportive 26.50 39.88 23.15 51.93 27.36 45.70 24.49 44.15 27.53 40.47 24.24 46.37 Opposite 2.87 8.32 5.65 13.62 3.95 10.65 3.84 10.60 3.08 7.98 4.24 11.62 Neutral/mixed 1.37 5.92 .63 1. 81 2.27 a 8.55 .65 a 2.03 2.30 b 8.05 .54 b 1.92 Off topic 38.65 107.65 27.10 68.25 41.00 85.49 31.96 99.07 33.75 77.96 34.84 102.77 New info 3.34 5.83 2.26 3.95 3.00 4.60 2.93 5.50 3.86 6.72 2.52 4.35 a.t = 2.14, df=198, p < .05 ; b. t = 2.07, df=198, p < .05 Table 313. Continued Investigation Action Appeared (N= 105 ) Not (N= 95 ) Appeared (N= 146 ) Not (N= 54 ) M SD M SD M SD M SD Hits 11356.35 23585.29 8239.54 17899.38 9975.98 19320.80 9605.19 25446.83 Replies 82.13 168.22 51.83 96.09 75.83 150.25 45 .87 101.69 Supportive 30.05 45.93 20.04 42.47 28.29 47.33 17.1 9 34.84 Opposite 4.05 11.51 3.67 9.53 4.34 11.26 2.61 8.48 Neutral/mixed 1.55 6.12 .60 2.87 1.22 5.50 .78 2.46 Off topic 43.0 8 121.30 25.00 52.80 38.81 105.63 22.81 58.39 New info 3.35 4 .88 2.51 5.62 3.14 5.60 2.44 4.17
124 Table 314. Independent sample T test of major focus and reader participation in forums Description Solution Influence Major focus (N= 34 ) Not (N= 166 ) Major focus (N= 21 ) Not (N= 179 ) Major focus (N= 23 ) Not (N= 177 ) M S D M SD M SD M SD M SD M SD Hits 9084.44 22678.04 10037.96 20810.15 7907.90 11781.64 10106.74 21927.56 8850.78 17081.58 10009.07 21585.74 Replies 46.85 60.52 72.02 150.11 48.43 49.04 70.01 146.11 45.17 65.66 70.67 145.94 Supportive 15.79 18.59 27.24 47 .9 2 15.67 26.57 26.42 46.06 17.96 27.03 26.25 46.25 Opposite 2.41 7.09 4.17 11.16 2.86 6.48 3.99 10.98 .78 1.17 4.27 11.19 Neutral/mixed .44 1.89 1.23 5.27 1.14 3.10 1.10 5.04 1.65 5.56 1.03 4.78 Off topic 24.53 39.02 36.53 103.11 26.90 28.93 35.38 10 0.30 21.09 34.39 36.23 100.51 New info 3.68 5.68 2.80 5.16 1.86 2.57 3.08 5.47 3.43 5.35 2.89 5.25 Table 314. Continued Investigation Action Major focus (N= 72 ) Not (N= 128) Major focus (N= 50 ) Not (N= 150) M SD M SD M SD M SD Hits 11961.88 23535. 63 8702.48 19569.04 8708.26 21404. 0 0 10265.07 21032.59 Replies 96.58 a 197.83 51.52 a 88.04 58.90 120.89 70.69 145.06 Supportive 33.57 b 51.29 20.64 b 39.63 27.26 56.10 24.64 40.09 Opposite 4.97 13.39 3.25 8.63 5.12 11.59 3.45 10.24 Neutral/mixed 1.81 7 .10 .70 2.92 .26 1.08 1.38 5.56 Off topic 53.18 144.63 23.98 47.31 23.70 62.42 38.09 103.94 New info 3.01 4.63 2.91 5.58 2.60 6.51 3.07 4.78 a.t = 2.22, df=198, p < .05; b. t = 1.99, df=391, p < .05
125 To answer RQ5b about the responsiveness of discus sion in blogs and online forums, it was found that among the 393 blog posts that allowed comments only 59 had blogger responses Th e average ratio of reader comments to blogger responses in blogs wa s 43.5. In forums, the average ratio of hits and replies was 1 45.79, with an average reply rate at .69 % (Table 315). To answer RQ5c about the homogeneity of discussion in online forums and blogs Table 316 shows that the number of supportive reader comments ( M =18.34, SD =76.56) in blogs was the highest among all the categories, followed by comments providing new information ( M =5.12, SD =26.82). In forums, more replies were off topic comments ( M =34.49, SD =95.34) and supportive comments ( M =25.30, SD =44.49), followed by oppos ing comments ( M =3.87, SD =10.59) (Table 316). Table 315. Responsiveness of discussion in blogs and online forums Min. Max. Sum M SD Blog (N=393) Times being read 30 188 653 1 341 325 3 413.04 14 468.74 Blogger response 0 29 238 .61 2.45 Reader co mments 0 1755 10 431 26.54 105.04 Forum (N=200) Number of hits 68 125 010 1 975 173 9 875.86 21 082.82 Number of replies 5 1 261 13 548 67.74 139.21 Table 3 16. Homogeneity of discussion in blogs and online f orums Blog (N=393) Forum (N=200) Max. Sum M SD Max. Sum M SD Comments/replies 1 755 10 431 26.54 105.04 1261 13 548 67.74 139.21 Supportive 1 278 7 207 18.34 76.56 358 5 059 25.30 44.49 New information 468 2 012 5.12 26.82 44 590 2.95 5.25 Opposite 86 278 .71 4.77 68 774 3.87 10.59 Neutral/mixed 34 173 .44 2.37 57 220 1.10 4.87 Off topic 33 688 1.75 3.82 1032 6 898 34.49 95.34 Others 6 12 .03 .34 3 8 .04 .33 Multimedia and Hyperlinks The sixth set of research questions w as about multimedia and hyperlink s in blogs and online forums when covering the tainted formula scandals Table 317 shows that,
126 for embedded multimedia features, pictures were the most frequently used in either blogs ( M =.43, SD =1.22) or forums ( M =.70, SD =1.90), while audio and video were barely employed in coverage. For hyperlinks, on averagely the highest number of hyperlinks in blogs ( M =.73, SD =2.07) and forums ( M =1.62, SD =5.42) linked to texts rather than multimedia only or multimedia and text combined pages. In blogs, there were more external hyperlinks to online media ( M =.47, SD =1.92) and internal hyperlinks to other posts on the blog portal ( M =.36, SD =1.71). In forums, the number of internal hyperlinks to other posts in the forum ( M =1.22, SD =5.02) was the highest, followed by external hyperlinks to online media ( M =.40, SD =3.04) and external hyperlinks to legacy news sites ( M =.20, SD =1.16) (Table 317). Table 3 17. Multimedia and hyperlinks in blogs and online forums Blog (N=400) Forum (N=200) Max. Sum M SD Max. Sum M SD Embedded multimedia Pictures 10 174 .43 1.22 16 139 .70 1.90 Audio 3 4 .01 .16 0 0 .00 .00 Videos 1 2 .01 .07 0 0 .00 .00 Type of hyperlinks Internal 16 142 .36 1.71 50 243 1.22 5.02 External to online media 30 187 .47 1.92 42 80 40 3.04 External to legacy sites 6 61 .15 .56 11 41 .20 1.16 External to others 3 26 .06 .39 7 26 .13 .77 External to government sites 2 5 .01 .13 1 3 .02 .12 Hyperlinks to Texts 23 291 .73 2.07 42 323 1.62 5.42 Multimedia and texts 17 81 .20 1.07 8 42 .21 .90 Multimedia content 5 10 .02 .27 5 19 .10 .56 Given the low number of multimedia embedded or hyperlinked in blogs and forums, it is difficult to show how significant the influence multimedia features were on reader participation throug h statistical analyses. However, it was found that on average, blog posts with multimedia included were read 6712.51 times and commented on 45.64 times, while blog posts with only text were read 2617.45 times and
127 commented on 43.02 times. In forums, thos e threads with multimedia features were read 18243.22 times and got 112.49 replies on average, while those threads without multimedia included were read 7011.87 times and got 52.42 replies on average. Influence of Author The last set of research questions explored the influence of authors, namely newspaper article writers and bloggers, on the coverage of formula scandals in newspapers and blogs. Generally speaking, male writers were dominant in both newspaper and blog reporting of tainted formula scandal s47.75% of newspaper stories were written by male journalists while only 20% were by female journalists; 61.25% of blog post were from men while only 38.75% were from women. To answer RQ7a about the gender influence on content, chi square analysis was con ducted. Table 318 shows that, in newspapers, female reporters were significantly more likely to emphasize action (43.8%), while male reporters were more likely to use a conflict frame (18.8%). Table 319 shows that, in blogs, female bloggers were signific antly more likely to emphasize solutions (25.8%) and use an action frame as the dominant frame, while male bloggers were more likely to focus on the impact of the tainted formula scandals (24.9%) and investigation of the scandals (32.7%). To test the influence of gender on sources and patter ns of participation, oneway ANOVA analysis was conducted. There was no significant finding of gender influence on sourcing in newspapers. But for blogs, male bloggers were more likely to use government sources ( F=2.31, p <.05, df=1 ) and NGO sources ( F=4.40, p <.05, df=1). In addition, female bloggers were more likely to attract reader comments ( F=6.30, p <.05, df=1 ) and leave responses for reader s ( F=4.48, p <.05, df=1 ).
128 Table 3 18. Gender differences of authors in the news paper coverage of tainted formula incident s Gender (N=400) Male (N=191) Female (N=80) Unknown (N=48) News Agency (N=81) p Major Focus 27.01 .008 Actions 54 (28.3%) 35 (43.8%) 10 (20.8%) 36 (44.4%) 13.87 .003 Investigation 44 (23.0%) 12 (15.0%) 14 (29.2%) 22 (27.2%) 4.71 .194 Solutions 40 (20.9%) 12 (15.0%) 8 (16.7%) 7 (8.6%) 6.40 .094 Influence 33 (17.3%) 8 (10.0%) 7 (14.6%) 5 (6.2%) 7.03 .071 Incident description 20 (10.5%) 13 (16.3%) 9 (18.8%) 11 (13.6%) 3.18 .365 Dominant Frame 36.73 .018 Action frame 51 (26.7%) 32 (40.0%) 13 (27.1%) 26 (32.1%) 5.07 .167 Responsibility frame 45 (23.6% ) 10 (12.5%) 10 (20.8%) 17 (21.0%) 4.15 .235 Conflict frame 36 (18.8%) 6 (7.5%) 6 (12.5%) 5 (6.2%) 11.06 .011 Reassurance frame 20 (10.5%) 17 (21.3%) 7 (14.6%) 11 (13.6%) 5.56 .135 Severity frame 20 (10.5%) 11 (13.8%) 8 (16.7%) 16 (19.8%) 4.54 .209 Economic frame 10 (5.2%) 1 (1.3%) 4 (8.3%) 2 (2.5%) --Uncertainty frame 9 (4.7%) 2 (2.5%) 0 4 (4.9%) --Others 0 1 (1.3%) 0 0 --Table 3 19. Gender differences in the blog coverage of tainted formula inc ident s Gender (N=400) Male (N=245) Female (N=155) p Major Focus 26.76 .000 Investigation 80 (32.7%) 33 (21.3%) 6.05 .014 Influence 61 (24. 9%) 21 (13.5%) 7.50 .006 Actions 46 (18.8%) 36 (23.2%) 1.15 .283 Incident description 34 (13.9%) 25 (16.1%) .38 .536 Solutions 24 (9.8%) 40 (25.8%) 18.11 000 Dominant Frame 18.87 .004 Conflict frame 71 (29.0%) 36 (23.2%) 1.60 .205 Responsibility Frame 64 (26.1%) 29 (18.7%) 2.92 .087 Severity Frame 48 (19.6%) 34 (21.9%) .32 .572 Action frame 28 (11.4%) 35 (22.6%) 8.90 .003 Economic frame 15 (6.1%) 2 (1.3%) --Reassurance frame 10 (4.1%) 12 (7.7%) 2.45 .118 Uncertainty frame 9 (3.7%) 7 (4.5%) .18 .675 To answer RQ7b about the influence of blogger type on the content, chi square analysis was used. The data showed that different types of bloggers health expert blogger s, parent blogger s, journalist blogger s, and citizen blogger swere significantly different in the major focus and dominant frames in taintedformula scandal related blog posts. Table 320 shows that health expert blogger s were most lik ely to focus on
129 solutions (31.0%) and incident descriptions (23%); parent bloggers were most likely to focus on actions (33.0%) and to use an action frame as the dominant frame (28.0%); journalist bloggers were more likely to emphasize the impact of the sc andals (33.0%) and the investigation (35.0%) of the incidents while citizen bloggers were the most interested in the investigation of the incident s (39.0%) and using the attribution of responsibility frame (36.0%) as the dominant frame in their blogs. Tab le 3 20. Influence of blogger type in the coverage of tainted formula incident s Type of Blogger (N=400) Health Expert (N=100) Parent (N=100) Journalist (N=100) Citizen (N=100) p Major Focus 80.58 .000 Solutions 31 19 7 7 29.46 .000 Incident description 23 21 6 9 17.24 .001 Investigation 21 18 35 39 15.73 .001 Influence 16 9 33 24 19.70 .000 Actions 9 33 19 21 17.86 .000 Dominant frame 86.69 .000 S everity Frame 26 21 17 18 3.01 .391 Conflict frame 20 26 33 28 4.43 .219 Action frame 17 28 8 10 18.45 .000 Reassurance frame 16 4 0 2 --Responsibility f rame 13 16 28 36 19.21 .000 Uncertainty frame 7 5 3 1 --Economic frame 1 0 11 5 --In addition, using oneway ANOVA analysis, a significant influence of blogger type w as found on the number of reader comment s ( F=3.87, p <.05 df=3 ) in that citizen bloggers received 76.53 more comments than p arent bloggers on averag e Health expert bloggers were more likely to use picture s than j ournalist bloggers and citizen bloggers ( F=6.20, p <.001 df=3 ).
130 CHAPTER 5 DISCUSSION AND CONCLUSIONS Given the severity of Chinese food safety problems in recent years, people are paying mor e attention to food safety issues in China. One of the most prominent food safety issues has concerned the milk formula s for babies. Considering the low full breastfeeding rate and the fact that the mean duration of any breastfeeding in most areas in China was between seven and nine months (Xu, Qiu, Binnes, & Liu, 2009), milk formula has become the main food for babies. Moreover, since there is a one child policy in China (Xu, Qiu, Binnes, & Liu, 2009), the social impact of the tainted formula incidents is magnified. For such a severe public health crisis in the society, rapid and reliable communication is critical. It is necessary to examine h ow and what information about the scandals was revealed to people thr ough traditional media, how the scandals were p erceived by the public, how and what public discourse on the issue was framed through public discussions, as each of these aspects contributes to the understanding of both public health and public communication in China. This study sought to examine the di scourse on formula scandals in China from 2008 to 2011, when the tainted formula incidents broke out intensively, to enable understanding of the messages disseminated by the government, traditional media, and the public. In doing so, this study looked at t he story focus, frames, tone, and sources in the coverage of the tainted formula scandals across newspapers, blogs, and online forums, testing author/gender influence on the content. B y comparing the government discourse represented by traditional media a nd the public discourse expressed through Internet based media such as blogs and public forums, significant differences were revealed. Through analysis of these differences,
131 this study provided a baseline for understanding the reasons and implications for differences and linkages between old and new media in China. Specifically, this investigation extensively analyzed the characteristics of citizen journalism and the role of new media in China from an agendasetting p erspective ( i.e., McCombs & Shaw, 1972), predicting its possible development in the future Such inquiry not only helps provide insights into the particular genres of new media, but also adds an important theoretical dimension of the underlying mechanism regarding gatekeeping, agendasetting, an d agendabuilding in new media as well as inter media agendasetting between old and new media. As a n integral part of the research, this part of the investigation aimed to further understanding of the growing impact of new media in China and how that may change the media landscape, information flow, and power structure in Chinese society The pattern of participation in new media, including the quantity and characteristics of comments/replies communication among readers and bloggers as well as communicati on among forum discussants, especially the responsiveness and homogeneity in their discussions, was examined. Special attention was paid to those blog/forum posts with more replies, revealing possible factors contributing to active participation. Further, multimedia and hyperlinks employed in forums and blogs coverage of tainted formula scandals were examined. Coverage of the Tainted Formula Scandals Story Focus and Frame Analyzing the coverage of the tainted formula scandals action and incident descripti ons were presented in at least half or more stories in newspapers, blogs, and discussion forums, and were apparently the most prevalent story foci It is probably because those issues were more pertinent to readers, as people were interested in the
132 development of the tainted formula incident s and the actions taken to deal with the incident s such as taking care of sick children recalling/destroying the tainted formula, or guaranteeing the quality of dairy products and protecting public health. The researcher found that while several story foci and frames appeared in the articles and posts, often a single major story focus or frame w as dominant While many story foci or frames may be present and manifest in ones reporting, one particular story focus or fram e category may dominate the writers thinking at the time, showing the primary concern, which is why it i s investigated here (Rogan, 2006). For at least half of all newspaper articles, blogs, and forum posts, actions and investigation were the major story focus. However, although action was the main theme for the highest number (33.8%) of newspaper articles, this study showed that investigation became the most prevalent major story focus in blogs and forums. This finding is important because it gives a bett er idea of the way the public perceived the issue and conveyed its concerns. Due to state ownership, Chinese newspapers were under the close direction of the government. As a result, actions, especially actions by the government, w ere emphasized in the overall newspaper coverage of the formula incident s, so as to stress the ability and determination of the government in protecting the public health In contrast, grass root s new media such as blogs and forums were more like alternative platforms for citizens wanting to speak out. People angry about the tainted formula were keen on knowing the facts, possible causes, and persons responsib le for the incident s Interestingly, incident descri ption, the second most frequent story focus across media, was on the bott om of the list for the presence of major story focus in newspaper s and blog s. It seems that the d escription of the incident s was more likely to be mentioned
133 in most stories of tainted formula scandals but less likely to be the main theme of the article s/p ost s Some other differences for story focus were also revealed among the three platforms For example, blogs, rather than newspapers or forums, were the most l ikely to mention and emphasize the social impact of the incident s Forums were the least likely to suggest solutions for the tainted formula incident s. B ut each devot ed a majority of the space to the discussion of investigation and actions. Based on framing theory (Entman, 1993) and secondlevel agendasetting theory (McCombs, 2005), frames appear in g in each story/post and the most dominant frame in each story w ere analyzed. Frames can help readers to better process the messages received and understand the issues being covered (Entman, 1993) T he results of f requency analysis revealed that, consisten t with the findings for action as a story focus, the act ion frame appeared most frequently in newspaper stories (87.5%) and was also the most frequently the dominant frame in newspaper coverage (30.5%) It seems that newspaper journalists, whether under th e direction of the government or through self constraint s, constantly emphasi zed actions especially actions by the government and health agencies This may reflect the routines of news reporting in China in that when any accident or incident with negative social impact has happens, action by the government always takes precedence over all other issues in news coverage, so as to help establish the authority and positive image of the government However, for blogs, the conflict frame was the most fr equently appear ing frame (74.8%) and the most dominant frame (26.8%). T o the general public, the tainted formula scandals first and foremost represented a conflict milk formula that was
134 supposed to nourish babies ironically poison ed them, causing serious concerns a bout public health in the society. The s everity frame was the second most frequent frame and the third most dominant frame appear ing in both newspapers and blogs as authors contin ually mentioned the number of cases the number of deaths and the spread of the tainted formula contamination. The a ttribution of responsibility frame which identified those responsible for the tainted formula scandals, was the second most dominant frame in newspapers (20.5%) and blogs (23.3%) showing that both traditional media and new media were interested in why such a public health crisis happened and who should take the responsibility Although in some cases the coverage was similar among media, framing differences between newspapers and blogs often dwarf ed the similarities As mentioned before, newspapers were more likely to use an action frame. When using the action frame to show the efforts that the government health agencies, and individuals had made to stop or correct the tainted formula scandals newspapers were also more likely than blogs to use the reassurance frame, so as to show readers that the tainted formula incident s were under control and the public need not to worry or should be less worried about it. On the contrary bloggers seemed to adopt an alternative approach by accentuating the s everity frame and conflict frame showing an obvious discrepancy between traditional media and new media in China. From the agendasetting perspective, it seems that the statecontrolled media were trying to get people to think about how the government and its agencies are looking after their own welfare. However, the bloggers and forum discussants did not
135 follow the agenda set by the newspapers. Instead, they were trying to get people to think about why the tainted formula scandals happened and who or what may be responsible. Obviously, the government and traditional media were concerned about their own interests, while the independent citizen media were concerned about the welfare of the people. Tone and Source As tones diffe r in news reporting or online discussion s, the attitude or the understanding of the issues by authors or discussants may be revealed. Ov erall, about 50% or more of the sample including newspaper articles, blog and forum posts, were basic statements of fac ts without subjective judgment, convey ing no tone to ward either the government or da i ry companies However, significant differences were noted between traditional and new media in that more than 60% of the newspaper articles had no tone while only about 3 6 % of blog post and 54% of forum posts showed no tone toward the government and dairy companies T his makes sense considering that stories in blogs and forums are more opinionbased, while objective reporting is valued in newspapers. Further it was found that, among those articles/posts with tones, the majority w ere negative toward the government or dairy companies Barely any item had positive tones. Given the damages and adverse social i mpact caused by the tainted formula scandals and the malfunction of the government/ dairy companies in the scandal s, it is easy to understand why criticism was common throughout the coverage. In addition the percentage of tainted formula stories with negative tones in newspapers was significantly lower than that in blogs. The reason might be that objectivity is a professional value of journalists. In the effort to appear objective, Chinese newspaper reporters likely worked hard to avoid
136 sounding negative or positive about the story. It is also possible that Chinese newspapers represent the interests of the government hence being less likely to criticize the government and da i ry companies. In contrast, citizen media like blogs and forums are platforms for regular citizens to express opinions straightforwardly, regardless o f the intentions of the governmen t. It indicates that, with blogs and forums, Chinese people not only are able to investigate and discuss social issues previously kept out of public discussion, but also to express competing opinions on issues with a negati ve social impac t perhaps criticizing the government that was more likely to be presented in a positive way by traditional media. It also reveals that citizens who are not professional journalists are not bound by industry norms in journalism t hey can feel free to show their opinions. Nevertheless, w hether they r ealiz ed it or not, Chinese Internet users have begun to use crowdsourcing with collective intelligenc e or p articipating in news reporting in new media such as blogs and forums. Certainly future rese arch could more fully delve into this idea. Source is another important issue to look at because sources have the power to shape news reports and influence media agendas ( Bryant & Oliver, 2008; McCombs, 2005). In the coverage of the tainted formula inciden ts, government and health agency sources were cited most frequently in newspapers. This makes sense because organizations such as the government are seen as having more political or economic power than individuals in either Eastern or Western society and thus are more likely to influence the news. Health agencies are the most professional and authoritative in providing health information. Given the nature of the tainted formula incidents as a public health crisis, the voices of health agencies should not b e ignored. Chinese
137 journalists in traditional media rely heavily on routines in reporting, citing more organizational sources, especially government sources, which are convenient for them to access For blogs and forums, traditional media such as news agencies, newspapers or TV were the most frequently used sources, along with government and health agencies. This is probably because, f or Chinese citizens who posted on blogs and forums, most of their information about the tainted formula scandals came from m edia especially traditional media. So it is not surprising to find that traditional media ranked high est in sources for blogs and forums, showing the strong agendasetting power of traditional news sources in the society. Some s ignificant differences were shown to exist among media in their use of sources. Compar ed to newspapers, new media were significantly less likely to cite government and health agencies, while being more likely to use traditional media and citizen sources. This finding is consistent w ith previous research showing that online citizen media featured fewer official and organizational sources ; online citizen journalists were less likely to be influenced by professional routines where sources representing organizations were regularly visite d (Carpenter, 2008) Sources are often selected based on their availability and suitability (Gans, 1979). It would have been difficult for Chinese people who were post ing in blogs and forums to get information directly from organizations or elites affiliat ed with those organizations. They were more likely to refer to newspapers, TV, and online news sites and spreading what they kne w online. So there were more traditional/ online media sources in blogs and forums. As observed in coding, some bloggers and for um discussants simply copy andpasted what journalists have already published, while many of them not only repeated what the professionals
138 have said but also posted his or her own opinion regarding the issues covered to support or question the original reporting. However, s ince this is not part of coding for this study, there is no statistical evidence. Future research could further explore this issue. Concerning the higher number of citizen sources in forums and blogs i t may show that new media such as bl ogs and forums adopted a more grassroots perspective in reporting. Without denying the huge agendasetting power of mass media, new media allow regular citizens more opportunities to serve as sources and reporters in the agendasetting process. Influence of Gender and Blogger Type Consistent with previous studies revealing that there were gender and occupational differences in the health blogosphere ( Miller, Pole, & Bateman 2011), this study found that characteristics and background of the authors influen ce d the coverage of the tainted formula scandals in Chinese newspapers and blogs Concerning gender, the data showed that male authors were dominant in both newspaper and blog reporting of the tainted formula scandals which reflects the dominant role of m en in traditional journalism as well as the gendered structure of Internet/blogging in China With a total population of 1 .34 billion in 2010, 51.27% of Chinese citizens were men, while 48.73% were women ( National Bureau of Statistics of China, 2011). In t he media industry, there are more male journalists. For example, about 84% of investigative journalists in China are men (Bandurski & Hala, 2010). For the online world, there were nearly 10% more male Internet users than female Internet users in China (CNN IC, 2010). Following the uneven male and female journalists/netizens ratio in China, the researcher found there were at least 20% more newspaper or blog stories written by men than women in this study.
139 M oreover, the analysis shows that c overage of the tai nted formula incidents in articles/blogs written by men differ ed greatly from th ose written by women, which is consistent with previous gender research in social psychology and communication showing that men and women have different ways of thinking diffe rent language styles and interests (Tannen, 1986; Rodgers & Thorson, 2003). In this study, male authors were found to be more likely to use a conflict frame, and to focus on the investigation and social impact of the incidents. Although there is no systematic evidence showing that males were generally more interested in conflict, some previous studies revealed that m en tend to be more assertive and aggressive, highlighting investigations in a conflict style, while women were more likely to employ cooperativ e orientations in conflicts (Cupach & Canary, 1997; Rogan, 2006) Also, male bloggers were found to use more government and NGO sources, showing a preference for authoritative organizations over individual s. In contrast, female authors were more likely to emphasize action and solutions when covering the tainted formula scandals Additionally, female bloggers were comparatively more active in communication with readers by attracting more reader comments and responding more to those comments. This makes sense given that women tend to be more relationally focused in their communication with readers, employing cooperative orientations (Rogan, 2006; Tannen, 1986). Gender socialization may offer a theoretical explanation for all these differences. The traditional gender model contends that men and women socialize differently in everyday life bec ause they have different values, priorities and social roles (Rodgers & Thorson, 2003). Men are presumed to value independent, assertive, and goal directed
1 40 behaviors, while women are presumed to value interdependence and relationship with others (Rodgers & Thorson, 2003), which may explain the gender differences in this study that male authors were more likely to focus on the investigation and social impact of the incidents, while female bloggers were more active in twoway communication with readers. Similar to gender, blogger type can also contribute to differences in the coverage of the tainted formula scandals For example, health expert blogger s were most likely to focus on solutions (31.0%) and incident descriptions (23%), probably because their professional background makes it possible for them to provide medical information on diagnosis (i.e., symptoms of affected children) and solutions (i.e., possible treatment for a ffected children) (Miller, Pole, & Bateman, 2011) Parent bloggers were most likely to focus on actions (33.0%) and to use the action frame as the dominant frame (28.0%), because most parents talked about their own experiences and worries (i.e. what they did to diagnose, treat, or prevent kidney stones resulting from tainted formula for their children, etc.) and paid more attention to governmental actions in dealing with t he incidents. B logs by professional journalists were more likely to focus on the investigation (3 5.0%) and the social impact (33.0%). It seems that Chinese journalists perceive blogging as an extension of investigative journalism, continuing the ir gate keeping function and setting the agenda for public debate (Yu, 2011). Citizen bloggers w ere reve a led to be more interested in investigation and identifying who should take responsibility in their coverage, and they receiv ed more comments from readers. F or citizen bloggers, when they were writing about the tainted formula incidents, it seems t hat most of them were very concerned about the issue, not only reporting on what they
141 knew but also writing their own thoughts on blogs to contribute to the so cial deliberation, and thereby attracting comments from those who were also very interested in the same issue for discussion. Compared to journalist bloggers who did not attract a lot of comments by continuing their same professional detachment in their blogs exactly as they do in the newspapers citizen bloggers, by showing their human side and expressing points of view, stimulated the readers to join in and share their own thoughts. Th e difference between journalist bloggers and citizen bloggers in engaging audience reveals their different understanding s of the participatory nature of new media as we ll as the significance of audience engagement N o significant sourcing difference was revealed among different types of blogs T he researcher noticed that there was one thing in common for the four types of bloggers traditional media were the most frequently cited sources. Apparently, no matter how professional or attentive the blogger was, they were more likely to be agenda followers compared to established media and organizations in the society. Also, when covering the tainted formula scandals it seems that bloggers also frequently cited or hyperlinked to the same types of blogs. For example, health expert blogs often provided links to other health experts blogs; more journalist blogs included content or hyperlinks from other journalists. This may indic ate the agendasetting power of bloggers in their own specialized blogosphere, showing an i nsular cycle of agendasetting process. It seems same type bloggers were feeding each other and building upon each other. Another finding concerning the effect of b logger type on the content is that h ealth expert bloggers used more pictures than other bloggers. This is probably because some
142 health expert bloggers, especially two pediatricians who engaged intensively in the treatment of tainted formula victims posted many photos of kidney stones The Growth of New Media The past decade has witnessed the growth of new media around the world. China is one of the countries that has experienc ed high speed development of the Internet and online media in the past ten years. While traditional media are limited by their routinized and institutional forms of news production in China, Internet based media such as blogs, forums, and other social media provide alternative platforms for citizens to participate in information produc tion and dissemination, changing the roles of regular people from passive recipients into active creators in communication, hence liberalizing journalism and mass communication in Chinese society (Moyo, 2009). The key in this revolution is the participatio n of people and their efforts in using new media technologies for spreading news and expressing opinions. Participation The pattern of participation in new media concerning the tainted formula scandals is one of the key issues this study explored Internet users participat e in online activities including information seeking, news publication, interactive discussions, etc. Their techniques, references, and beliefs based on past experiences and socialization contribut e to their understanding of media ( Bourdieu, 1993) influencing their online practice s and discourse created through such practice. Generally speaking, when examining patterns of participation, this study focused on three dimensions : size/frequency of participation, level of interactivity/responsi veness and homogeneity /heterogeneity in discussion
143 This study showed, on average, each blog post was read 3 413 times and each forum thread was read 9 ,875 times indicating a significant level of attention from Chinese netizens to the tainted formula inc idents. Further, it was shown that, on average, each blog post received 26.54 reader comments and each forum thread included 67.74 replies. Chinese Internet users in this case did participate in the new media sphere. However, comparing the number of times that a blog post or a forum thread was read with the number of comments or replies from readers may suggest t hat more Chinese online users were lurkers who were more likely to engage in periphery participation: on average, every 129 reads of a blog post generated one comment and every 146 reads of a thread generated one reply This may be because the participatory atmosphere in Chinese new media was still frail or more online users, although interested in the topic, felt it would be difficult or were reluc tant to say any t hing. In addition, it was revealed that the average ratio of blogger responses to reader comments was 1:43.5. T h e se finding s are consistent with previous research showing that there is a 909 1 rule in most online communities 90% of users are lurkers who never contribute, 9% of users contribute a little, and 1% of users account for almost all the participation (Nielsen, 2006). It seems that, although reader contribution, feedback, or crowdsourcing can contribute to newsgathering and informat ion sharing, it was not yet a feature of online journalism in China (Yu, 2011). Regarding the homogeneity/heterogeneity in discussion, it was found that the majority of reader comments in blogs w ere supportive of or supplement ed the original posts while only a limited number of comments offered opposi ng or neutral/mixed views. In other words, discussion in blogs was quite homogen e ous. This could be a
144 result of the fact that people selectively choose to read blogs in accordance with their own points of view, hence making them more likely to agree with what the bloggers say. It is also possible that the topic of this study tainted formula scandal, may have had a specific impact on the results. T he tainted formula scandal is a health issue, in which many innocent babies were hurt, so it is possible that people were more likely to demonstrate outrage about the egregiousness of this issue. The topic may have influenced the level of agreement with each other upon causes, solutions, actions, etc. W hen another iss ue such as economic or political relationship between countries is discussed, there would be a higher level of disagreement because it is a conflict loaded topic. In forums, the mean number of replies in each thread was higher than that in blog s. This mak es sense considering that forums are designed for public discussions while blogs are well suited for self expression. I n forums as in blogs except for comments that were off the topic, most replies were supportive of the original post. However, the propor tion of supportive replies in forums was lower than that in blogs, and the proportion of opposing or neutral/mixed replies was higher, showing more diversity in conversations and a higher level of deliberation in public discourses Through the back and for th of disagreement and discussion, a range of opinions can be debated and public consensus can be reached. To further explore th e factors that may influence the vitality and diversity of online discussion, it was revealed that when talking about the i mpact of the tainted formula incidents in blogs and forums there were more neutral/mixed replies from readers, showing the uncertainty about the tainted formulas social impact among the public.
145 When the blog post covered actions or used an action frame, fewer opposing comments appeared indicating there was a public consensus for taking actions to deal with the tainted formula incidents. In addition, it seems that the public was highly interested in the investigation of the tainted formula incidents; when inve stigation was the dominant frame in the blog posts and forum threads there were significantly more replies especially more supportive replies. Not yet satisfied by the information that had been released, the public wanted to know not only how the problem was being fixed but also the causes of the problem. Relating this finding to the nature of blogs and forums as citizen media a t a time when traditional media emphasized actions taken by the government to resolve the tainted formula problem trying to est ablish a positive image of the government and somewhat concealing its responsibility for the scandals citizen s showed enthusiasm for continuing investigation on the issue through crowdsourcing, reflecting the autonomy of public di scourses. Through writing and commenting on the unknown or unresolved issues in the tainted formula scandals, the public was demanding accountability. As the public discussions online would not allow the scandals to die down and disappear, the issue of responsibility, which could possibly fade away and never be resolved in mainstream media, was kept alive by the public. One good case in point for crowdsourcing by the Chinese netizens is that they studied the website of AQSIQ, Chinas major food quality control bureau, after the bre akout of the Sanlu tainted formula in 2008, finding out that the AQSIQ had ignored several complaints from parents about the formula before the breakout of the scandal (Change, 2008). The revelation of this finding online increased the severity of criticis m
146 of the government in the public, leading to the resignation of the head of A QSIQ later (Chang, 2008; Ma, 2008). There is an Internet saying recently in China, public attention and scrutiny is power, surrounding and watching can change China ( guanzhu ji ushi liliang, weiguan gaibian zhongguo in Chinese: ), meaning that when citizens are not able to alter the performance or actions of the government authorit ies through other means paying attention and expressing opinions online for the issues of concern may also contribute to t he solving of social problems. In fact, citizens themselves become watchdogs, taking on the role that the press has in a democracy. The impact of a multiplicity of citizens watching and responding is greater than the sum of the individual s (Shirky, 2009). As the online public discourse gets stronger, the government and government authorities are less able to ignore public opinion. Instead, they have to respond to the public scrutiny as happened in the tainted milk formula scandals (Chang, 2008; Yang, 2003) In this way, the power of the government is better watched and supervised, which is a good start toward more interactivity and balance between the government discourse and public discourse (Chang, 2008). When more inquiries from citizens are invoked and more social problems are revealed online, the transparency of the society can be improved and citizens could be empowered in the long run (Zheng & Wu, 2005). Generally speaking, online citizen media have revived and magnified the public sphere in China. Ac cording to Habermas (1989), public s phere is a realm of social life in which people come together to discuss public issues and form public opinion. The findings of this study show that now there is a functional public sphere online in China,
147 in which peopl e exchange ideas and form public opinion. Although the statecontrolled media were trying to get people to think about how the government and its agencies are looking after their welfare, the bloggers and forum discussants set up their own agenda to get pe ople focusing on the causes and the investigation of the public health crisis through civic participation. Multimedia and Hyperlinks Multimedia and hyperlinks are important features in new media. The application o f hyperlinks and multimedia in new media is not only an issue of technology and skills but also an issue of understanding and developing a new culture of information sharing ( Deuze, 2003). This study revealed that picture s were the most frequently embedded multimedia feature in both blogs and forum s, while audio and video were barely employed in the coverage of the tainted formula scandals online. Similarly, when examining the content that was hyperlinked, it was revealed that a majority of hyperlinks linked to text content; only a very small number of hyperlinks connected to multimedia content or multimediaandtext combined content. Whether and how Internet users employ multimedia features indicates their attitudes and ability in using new media technologies. The limited use of multimedia in Chines e blogs and forums could result from users inadequate knowledge of technology or their narrow understanding of its function to attract audiences and spread information. Or it might be topic centric, as there might not be much opportunity for relevant audi o or video materials available online for tainted formula scandals. In other words, this news did not lend itself to audio or video reports. Overall, it seems that multimedia is still an untapped source However, when multimedia were present in a blog/foru m post here, it was more likely for the post to attract more readers and get more replies. So perhaps Chinese Internet users should
148 adopt more multimedia features in online discussions in the future, so as to draw attention from the audience. Hyperlink s ar e an easy and convenient way for Internet users to connect a story to other archives and resources ( Deuze, 2003). In fact, compared to multimedia features, hyperlinks were more prevalent in Chinese new media when covering the tainted formula incidents. Mor e specifically, most hyperlinks in blogs linked to independent online news sites, citizen media, or other posts i n the same blog portal rather than government/health agency sites or legacy news sites maintained by newspapers, TV stations or news agencies Similarly, the majority of hyperlinks in online forums connected to other posts in the same forum independent news sites or personal blogs. It seems bloggers or forum discussants had heavy reliance on online media such as independent online news sites, personal blogs, etc. It is possible that, g iven their practice of blogging/forum posting and their liberalized way s of information seeking, Chinese online users have developed confidence in online media rather than traditional media ( i.e., newspaper/TV websi tes) and other organizational sites ( i.e., government or health agency sites). Or maybe Chinese Internet users were not satisfied with the information released from the government and traditional media sites for the tainted formula scandals, hence referring to sites outside government control that served better to the welfare of the public. Implications Theoretical Implications This study makes a variety of important contributions to the development of agendasetting theory, new media research, and Chinese media research.
149 T he results of this content analysis added to the understanding of agendasetting theory, expanding the knowledge of the theory, especially within the new media sphere and gover nment controlled media systems. Through the examination of the tainted formula scandal coverage in mainstream newspapers, online forums, and blogs from an agendasetting perspective, this study revealed how traditional media and new media were involved in the agendasetting process, showing the influence of the m edia agenda on the public agenda. A vital criterion in assessing the usefulness of a social scienc e theory is its theoretical scope of generality (Shoem aker, Tankard, & Lasorsa, 2004). With the development of the Internet and online media, there have been a lot of controversies about whether and how agendasetting theory fits in the new media environment. By analyzing first and secondlevel agenda setting in the coverage of the tainted formula incidents in new media and comparing it with what appeared in tradit ional mainstream media, the value of the agendasetting theory in explaining the influence of new media ( i.e., establishing issue salience mirroring public discourses, etc.) was confirmed. This study not only extends agendasetting theory to new media pl atforms but also tests the theory in a different media system set Chinas traditional media is government controlled, while Chinas online media is a somewhat semi government controlled media system. This study analyzed agendas about the same public heal th issue presented in both traditional media and online media in China, providing an opportunity for the researcher to understand the agenda differences between the two systems. Further, media, however powerful they may seem to be, are entities in a social setting under the influence of political systems, cultures, etc. By examining the agenda-
150 setting in the tainted formula incident coverage in Chinese media, this study extend ed the agendasetting theory to a culturally, politically, and economically differ ent social setting. Chinas unique political system dual ownership of traditional media, and censorship in both old and new media posit s challenges to the traditional Western agenda setting research. Through a case analysis of news coverage in both tradit ional media and new media in China, this study may help to expand t he research on agendasetting theory in a culturally and politically different domain. Th is study also provides an exploratory inquiry into an emerging area of scholarship where few researc hers have venturedChinese new media research. A cademic scholarship about new media in China is still in its infancy with a limited body of knowledge. This research allows scholars to understand Chinese new media in comparison with traditional media in Chi na. T h e issue specific quantitative study provides empirical data to support the claim that there is growing influence of new media in China, revealing the interplay among the Internet, media and social structures in Chinese society today, shedding light o n journalism and media research. Hopefully, this work would begin a line of research in understanding the emergi ng new media phenomena in China and its influence on the traditional media as well as the emergence of a genuine public sphere. Practical Implic ations There has been monumental development in the digital landscape in the past decade, alongside huge political, economic, cultural, and technological changes in China. T raditional media (newspapers and TV) were the only or major information sources an d agendasetters previously The emergence of new media (Internet) not only creates new venues of information for people but also provides bottom up
151 approaches for agenda setting and public opinion formation in the society It is possible that there is a t ransformation of public sphere in China today as what was in Europe in the 18th century. According to Habermas (1989), prior to the 18th century, European culture had been dominated by a representational culture, in which one party sought to represent itself in the society by overwhelming its agendas. However, the growth of newspapers, coffeehouses, reading clubs, etc., gave rise to the public sphere that was outside the state control and allowed individuals to exchange ideas, hence contributing to the development of the culture of public sphere in the 18th century in Europe. Similarly in todays China, the development of the Internet and citizen online media makes it possible for the public to speak out and openly discuss public issues, therefore changing t he previous situation in the society that only the government has the right to say. Consequently, a transformation of public sphere is under way in China now. By examining the story focus, fram es tone, sources, participation patterns, and use of multimedi a and hyperlinks in formula scandal coverage among mainstream newspapers, online forums and blogs, this study offers an overview of communication strategies and tactics used by media practitioners and Internet users. It seems that traditional media are us ed to their established ways of news reporting on storytelling, framing, and sourcing. Probably due to government ownership, they were more likely to report positively on the government. In contrast, new media adopted a quite different approach to the coverage, reflecting a public discourse that was more negative to the government and dairy companies in the tainted formula incidents, and providing a strong alternative to the government discourse presented in traditional mainstream media.
152 Considering the power of setting agendas, i t seems impossible to change the situation that mainstream media dominate in information pr oduction and dissemination in the short run. However, online media including blogs and forums have been catching up on becoming alternativ e information sources and establishing themselves as important news outlets for the public. In the case of the tainted formula coverage, the public has set a different agenda from what the traditional media were setting the traditional media wanted to reas sure everyone and to show the actions of the government to correct the scandal, while the public had an additional agenda of finding out why the problem occurred and who was to blame. This is of great significance to the journalism and media landscape in C hina, warning traditional media and communication professionals that they face the risk of being challenged. In a media world in which audiences can decide what news they want to get and how they want to get it, the future belong s to those who understand the public s changing behavior and can target content to audience (Project for Excellence in Journalism, 2011). Although new forms of journalism such as journalist bl og s, have been developed by traditional media in China to catch the trend toward online journalism, more forms of crowdsourcing and participatory journalism are worth exploration in the future, because it is likely that an increasing inline public w ould use crowdsourcing and other forms of participation to demand accountability from both medi a and government. T he interactions between professional journalists and nonjournalists such as bloggers or forum discussants were noticed in the study As new media was one of the sources for mainstream media, an emerging symbiosis for news reporting betw een communication
153 professionals and amateurs is emerging, providing new possibility for media development in China. As to the new media and Internet users who practice participatory journalism and crowdsourcing in the burgeoning domain it seems they are s till i n an early stage of development, although certain influence has been exerted. New media technologies such as comments, multimedia and hyperlinks have not been fully explored. Gender differences and occupational differences may show us that the online sphere in China is dominated by a small number of the population who are the hard core of Internet users in China. Nonetheless, no matter whether people participate intentionally in citizen journalism online to practice news reporting, or they randomly involve themselves in the crowdsourcing and participatory journalism practice to contribute their knowledge and help news organizations with news making and news dissemination, Chines e Internet users have started to utiliz e the new media tools to promote the public agenda, which is the foundation of a public sphere that barely exist ed in China before. With such efforts by citizens online, the Chinese government is under more pressure ( i.e., citizens studie d the website of AQSIQ to learn about the dysfunction of the agency), and is more likely to yield to the united public voice in policy making ( i.e., the resignation of the head of AQSIQ). Therefore, the advent of new media and citizen journalism online not only has changed the media landscape but is also tran sform ing the social and political landscape in China. Limitations This study suffers from several limitations that should be taken into account when evaluating the results and weighing the conclusions.
154 One limitation of this research is the sample for anal ysis. Eight major Chinese newspapers were selected to represent the traditional media agenda for the study based on their availability and representativeness However, there are a huge number of other print media outlets and broadcasting media outlet s, whi ch work together to set the media agenda. For online media, only 80 blogs were chosen from the countless Chinese blogs online. B ecause there was no complete directory or ranking of Chinese blogs available, although the researcher tried every possible way t o collect blogs from popular blog portals, newspaper/TV websites, and parenting websites, determining their popularity based on the number of time they were read and recommendation by the hosting websites for sampling, they may still not be the ideal sampl e for research. In addition, given the censorship of the Internet in China, some blog posts or forum threads that were very critical to ward the government on the tainted formula scandals may already have been removed before the researcher started sam pling. All these sampling limitations may have an influence on the findings. This wa s an issue specific quantitative study So o nly the content related to the tainted formula incidents were sampled and analyzed, hence limiting generalizability of the results. Fo r example, i t is possible that the findings on the framing or sourcing patterns on newspapers or fi ndings about the interactivity i n Chinese new media are specific to the coverage of the tainted formula incidents. Because they may not be representative of the overall media sphere in China, the results of this study should be understood with caution and the conclusions should be drawn carefully when referring to the broad contextual meaning s and implications.
155 On a related note, because this is a quantitative study using content analysis that is basically descriptive considering this l imitation of th e research method, it is difficult to explore further reasons for the phenomena and draw conclusions about causalities. Another potential limitation of this inve stigation is that the process of integrating initial different story focus categories into summative groups may have negated some of the subtle qualities of each story s focus Through factor analysis, the original 19 categories were collapsed into five si ngle story foci Although such a reduced categorization scheme may ignore some of the subtle differences among the original story foci, it does provide a more integrated and parsimonious structure that facilitate d crossmedium analysis and comparison. Future Research This study extensively examined the coverage of the tainted formula incidents in Chinese newspapers blogs and forums The tainted formula incidents as a public health crisis provided a meaningful context for the analysis of government/media di scourse and public discourse in China as well as the comparison between them Future research may examine discourse on other issues such as political events, economic/business issues, natural disasters, or lifestyle subjects in traditional and new media, so as to draw a full picture about how journalism i s practiced in China today More media could be used for investigation. For example, more newspapers, blog posts and forum threads could be included in the sample. Moreover, e merging n ew media such as micro blogging which is currently gaining more and more popularity among Chinese netizens could also be used for future study. Looking beyond agendasetting, there are a variety of other theories or theoretical frameworks that could provide future guidance to understanding the role of new media
156 in Chinese media landscape. For example, future researchers could draw on the gatekeeping theory to examine the process of information flow and the gatekeeping function of different practitioners ( i.e., professional journalists, bloggers, forum posters, micro bloggers, etc.) in communication, asking questions such as whether the gatekeeping process or the role of gatekeepers is changing For future research about new media in China, several possibilities could be taken i nto consideration. Building on the findings about new media in this study, future researchers could us e surveys interviews, or focus groups to analyze the motivations of online users in activities such as news reading, blog ging, online discussion, linking etc., exploring the pressure and barriers they face in online practice, as well as their attitudes to ward new media technologies. Also, differences between those hardcore participants in crowdsourcing or participatory journalism and those who are less interested in participation may be explored. Moreover, Chinese journalists can be interviewed to find out whether they are influenced by the competitions from independent media online and what their possible reactions are. Interactivity is a key characterist ic of new media. More dimensions could be included in its measurement Moreover, future research should attempt to measure the effects and effectiveness of interactivity, asking questions like whet her and how readers perceptions of content on new media is influenced by the interactiv e and multimedia features etc. in China. In addition, it will be interesting to find out more about gender differences and medium differences in the new media sphere in China. Finally, future research should pay more attention to the de velopment of new media in China. W ith its traditional ly controlled media environment, Chinese online
157 media provide an interesting case for communication researchers to explore. The general developing trends, structure of digital sphere, and the political and social implications of new media in China are worth further examination. The development of newspapers and coffeehouses was recognized by Habermas (1989) as the most important contributor to the transformation of public sphere in Europe in the 18th century The findings of this study reveal that the Internet and online citizen media such as blogs and discussion forums may play a similar role in todays China, as they are giving birth to an emerging online public sphere that challeng es the repre sentational culture by the government and giv es people opportunities to speak out. The development of citizen participation due to the new media technologies promotes the sharing of information and views with less control from the government, hence bringin g in liberalization to journalism and mass communication. Public conversations facilitated by the Internet make it possible for more civic deliberations, which are the cornerstones of a democratic society where the public has more say With such transformation of public sphere, new media and new China are on the horizon. The results of this study have tapped into that transformation, providing a new framework from which other studies can further investigate this topic.
158 APPENDIX A NEWSPAPERS AND ONLINE FOR UMS FOR STUDY Table A 1. Chinese newspapers selected for analysis Newspapers Circulation Influence Party/Metro Peoples Daily 2,350,000 National Party Beijing Youth News 680,000 Regional with national influence Led by Chinese Youth League Southern Me tropolitan News 1,580,000 Regional with national influence Metro Oriental Morning Post -Regional Metro Xinmin Evening News 1,100,000 Regional Metro Changjiang Daily 400,000 Regional Party Chutian Metropolis Daily 1,300,000 Regional Metro Hebei Dail y 410,000 Regional Party Sources: Willings Press Guide, 2009; China Media Guide, 2008. Table A 2. Chinese online public forums selected for analysis Forum Web address Affiliation Tianya http://www.tianya.cn/ Indepe ndent (commercial portal site) Sina http://bbs.sina.com.cn/ Independent (commercial portal site) Sohu http://club.sohu.com/ Independent (commercial portal site) Qiangguo http://bbs1.people.com.cn/ Affiliated with the Peoples Daily ( traditional media site) Xinhua http://www.xinhuanet.com/foru m/index.htm Affiliated with the Xinhu a News Agency ( traditional media site)
159 APPENDIX B RESULTS OF INTERCODE R RELIABILITY Table B 1. Intercoder reliability for each variable (Krippendorff's Alpha) Variables Newspaper Forum Blog Newspaper/Forum/Blog 1 1 1 Date 1 1 1 Story origin/Forum s ection .93 .91 -Type of blogger --1 Story f ocus Infected symptoms .90 1 .97 Information on spread .98 .87 .90 Dairy products identified .94 .87 .91 Formula recall .93 .92 1 Guidelines for handling the tainted formula .93 .83 .85 Guidelines for seeking medication .95 1 .95 Compensation to victims 1 .88 1 New regulations & food/milk monitoring system 1 .93 .96 Economics of the incident s .85 1 .96 Health/medical issues related to food safety .92 .88 1 Politics of the incident s .87 .81 .90 Possible causes for the incident s .89 .90 1 Problems of the fo od industry, health agencies, and government .94 .90 .97 Potential responsibility takers .85 .89 .95 Actions by government/health agencies .86 .89 .92 Ac tions by dairy companies .86 1 .89 Actions by individuals .87 .91 .95 Actions by hospitals .89 .91 1 Major focus in the story .84 .86 .92 Frames appeared Attribution of responsibility frame .92 -.91 Severity frame .84 -.90 Action frame .9 4 -.90 Conflict frame .87 -82 Economic consequence frame .89 -1 Reassurance frame .94 -.90 Uncertainty frame .85 -1 Other frame .85 -1 The most dominant frame .84 -.91 Tone Tone to the government and officials .97 -.96 Tone to d airy companies and officials .87 -.93 Favorable to the government and officials -1 -Unfavorable to the government and officials -.92 -
160 Table B 1 Continued Variables Newspaper Forum Blog Tone Favorable to d a i ry companies -1 -Unfavorable to d a i ry companies -.92 -Neutral/mixed tone -.95 -No tone presented -.89 -Number of sources cited in each categories Government and their officials .87 .84 .97 Health agencies and their officials .88 .88 .97 NGOs, independ ent orgs, their officials 1 1 .92 Health experts 1 1 1 Hospitals and their officials 1 1 1 Dairy companies, spokespersons, lawyers, etc. 1 1 .95 Victims parents, family, lawyers, etc. 1 1 1 Other individuals 1 1 1 Traditional media source .96 .92 .9 6 New media source 1 .87 .88 Others 1 1 .88 Gender of the author/blogger .94 -.92 Hits (times of being read) -.97 1 Replies in thread/Comments from blog reader s -1 .97 Responses from the blogger to comments -.93 .94 Homogeneity of discussio n Pose argument(s) to support -.87 .89 Propose different/opposite ideas -.79 .90 Mention neutral or mixed points of view -.83 .83 Bring in new related information/topic -.87 .98 Off topic -.97 .94 Number of hyperlinks Internal hype rlinks -1 1 External hyperlinks to legacy news sites -1 .93 External hyperlinks to online news sources -1 .83 External hyperlinks to government/health agency -1 1 External hyperlinks to others -1 1 Hyperlinks to only multimedia content -1 1 Hyperlinks to only texts -1 .92 Hyperlinks to multimedia and text combined -1 1 Number of multimedia Embedded audio -1 1 Embedded video(s) -1 1 Embedded photo(s)/picture(s) -.94 1
161 APPENDIX C CODEBOOK FOR NEWSPAP ERS Story qualificati ons: Include stories mainly covering the tainted formula incident s (> 400 words) No story compilation, duplicate, or Q&A 1. Coder ID 2. Case number: (Start with 001) 3. Newspaper (1) Peoples Daily ( http://paper.people.com. cn/ ) (2) Beijing Youth News ( http://bjyouth.ynet.com/ ) (3) Southern Metropolitan News ( http://epaper.oeeee.com ) (4) Oriental Morning Post ( http://w ww.dfdaily.com/ ) (5) Xinmin Evening News ( http://www.xinmin.cn/) (6) Changjiang Daily ( http://news.cnxianzai.com/ ) (7) Chutian Metropolis Daily ( http://ctdsb.cnhubei.com/) (8) Hebei Daily ( http://epaper.hebnews.cn/ ) 4. Story date ( i.e. October 1, 2008 = 20081001) 5. What is the story origin? (1) News article ( i.e., regular news reporting) (2) Editorial (opinion piece writ ten by the editorial staffs or publisher) (3) Opinion column by staff journalists (4) Opinion column by others ( i.e., health experts, economist, critics, etc.) (5) Feature (6) Letter from a reader (7) Other 6. Focus of the story (choose ALL that appeared in the story) (1) Infected s ymptoms ( i.e., kidney stones, etc.) (2) Information on spread of the incident s ( i.e., find ing more tainted milk formula) or the spread of the infection ( i.e., infected/deaths/confirmed cases) (3) Dairy products identified as contaminated ( i.e., Sanlu, Shengyuan, etc.) (4) Formula recall (5) Guidelines for handling the tainted formula/finding untainted formula (i.e., where to return the tainted formula, etc.) (6) Guidelines for seeking medication/hospitalization ( i.e., possible treatment for affected children, etc.) (7) Compensation to victims (i.e., how the compensation system works, etc.) (8) New regulations or monitoring system (i.e., new regulations concerning food safety, new standards for milk test, etc.) (9) Economics of the incident s/recall ( i.e., bankruptcy of dairy companies, etc. ) (10) Health or medical influence of the incident s
162 (11) The politics o f the incident s (i.e., legal restrictions on Chinese milk formula importation in foreign countries, etc.) (12) Possible causes for the incident s (i.e., outdated standards for formula safety test, etc .) (13) Problems of the food industry or government agencies (i.e. government corruptcy, dairy companies slow reactions, etc.) (14) Potential responsibility takers (i.e., government should be blamed, etc.) (15) Actions by government or health agency (i.e., free treatmen t to tainted formula victims, etc.) (16) Actions by dairy company (i.e., raise fund for victim compensation, etc.) (17) Actions by individuals (i.e., go to hospital for body check, stop drinking milk formula, etc.) (18) Actions by hospital (i.e. provide treatment to vict ims, etc.) (19) Others (please specify ) 7. Major focus (choose ONE major focus that is the dominant theme of the story) (1) A description of the incident s (2) Solutions (3) Influence of the incident s (internationally and nationally) (4) Investigation of the incident s (5) Actions 8. Fram es appeared (choose ALL the frames appeared in the story) (1) Attribution of responsibility frame: to blame the issue on the government, the dairy companies, or other stakeholders. Keywords: responsible ( zeren), problem ( wenti ), etc. (2) Severity frame: the spread of the tainted formula contamination, for example, the infected/deaths/confirmed cases. Keywords: severe ( yanzhong), infected ( yingxiang), panic ( konghuang), etc. (3) Action frame: any action taken by stakeholders such as the dairy company, the government or the individual, to deal with the formula scandals such as formula recall, compensation, punishment, hospitalization of sickened kids, etc. Keywords: action ( xingdong), recall ( zhaohui ), hospitalization/medication ( yizhi ), etc. (4) Conflict frame: conflicts bet ween individuals, companies, victimized groups, the government, etc. Keywords: conflict ( chongtu, maodun), etc. (5) Economic consequence frame: the possible economic losses for the individuals, companies, the dairy industry or the country due to the formula scandals Keywords: loss ( sunshi ), pay ( chuqian), etc. (6) Reassurance frame: tells the public do not need to worry or should be less worried about the formula scandals or food safety issues in China. Keywords: reassured ( fangxin), safe ( anquan ), etc. (7) Uncertaint y frame: uncertainties in any aspect of the issue, including the cause, the cure, the possible spread, etc. Keywords: uncertain ( buqueding), unclear ( buming), etc.
163 (8) Others (please specify) 9. The most dominant frame (choose ONE dominant frame in the story from above) 10. What is the tone of the story? (1) To government: N egative ( hostile or aggressive language) = 0 P ositive (optimistic, stressing logical and sensitive demeanor) = 1 N eutral/mixed (between or a mix of negative and positive language) = 2 N o tone appeared (void of judgment, only statements of facts) = 3 (2) To the dairy company: Negative ( hostile or aggressive language) = 0 Positive (optimistic, stressing logical and sensitive demeanor) = 1 Neutral/mixed (between or a mix of negative and positive language) = 2 No tone appeared (void of judgment, only statements of facts) = 3 11. What sources are cited in the story? Count the number of sources from each category. If a source, for instance a pediatrician, appeared several times in a story, it is counted only once. (1) St ate/local government/agency or their officials (2) Health agencies or their officials ( i.e., the Chinese Center for Disease Control and Prevention (CCDC), the State Food and Drug Administration of China (SFDA), the General Administration of Quality Supervision, Inspection and Quarantine (AQSIQ), the Ministry of Health, or other food or health related institutions) (3) Non government organizations (NGOs), other independent organizations, and their officials (4) Health experts ( i.e., doctor/nurse/pediatrician, professor/ researcher, food/nutrition expert, etc. that were not associated with the previous three types of organizations) (5) Hospitals and their officials (6) Dairy companies, their spokespersons, lawyers, etc. (7) Victims parents, family, lawyers, etc. (8) Other citizens, witne sses, etc. (9) Newspapers, TV stations, magazines, or news agencies such as the Xinhua News Agency, and their websites (10) Blogs, online forums, independent online news sites, or other citizen media (11) Others (please specify) 12. Gender of the author (If there are sever al authors, code the gender of the first author) (1) Male (2) Female (3) Unknown (cannot discern) or no author listed (4) News agency
164 APPENDIX D CODEBOOK FOR ONLINE FORUM POSTS Post qualifications: Include threads in online forums mainly talking about the tainted milk formula incident s The thread selected for analysis should include no less than 10 posts 1. Coder ID 2. Case number: (Start with 401) 3. Online forum (1) Tianya ( http://www.tianya.cn/) (2) Sina ( http://bbs.sina.com.cn/ ) (3) Sohu ( http://club.sohu.com/ ) (4) Qiang g uo ( http://bbs1.people.com.cn/ ) (5) Xinhua ( http://www.xinhuan et.com/forum/index.htm ) 4. Date of the original post in each thread ( i.e., October 1, 2008 = 20081001) 5. Section of the forum (the specific subject/section that the post belongs to in the forums) (1) Lifestyle/food (2) Health (3) Parenting/Education (4) Politics/Public affair s/Laws (5) Economics (6) Entertainment (7) Talk/conversation/gossip (8) News (9) Region (10) Others (please specify) 6. Focus of the story (choose ALL that appeared in the thread) (1) Infected symptoms (i.e., kidney stones, etc.) (2) Information on spread of the incidents (i.e., finding more tainted milk formula) or the spread of the infection (i.e., infected/deaths/confirmed cases) (3) Dairy products identified as contaminated (i.e., Sanlu, Shengyuan, etc.) (4) Formula recall (5) Guidelines for handling the tainted formula/finding untainted formula (i.e ., where to return the tainted formula, etc.) (6) Guidelines for seeking medication/hospitalization (i.e., possible treatment for affected children, etc.) (7) Compensation to victims (i.e., how the compensation system works, suggestions for victims to seek compens ations, etc.) (8) New regulations or monitoring system (i.e., new regulations concerning food safety, new standards for milk test, etc.) (9) Economics of the incidents/recall (i.e., bankruptcy of dairy companies, etc.)
165 (10) Health or medical influence of the incidents (11) The politics of the incidents (i.e., legal restrictions on Chinese milk formula importation in foreign countries, etc.) (12) Possible causes for the incidents (i.e., outdated standards for formula safety test, etc.) (13) Problems of the food industry or government agencies (i.e. government corruptcy, dairy companies slow reactions, etc.) (14) Potential responsibility takers (i.e., government should be blamed, etc.) (15) Actions by government or health agency (i.e., free treatment to tainted formula victims, etc.) (16) Actions by dairy company (i.e., raise fund for victim compensation, etc.) (17) Actions by individuals (i.e., go to hospital for body check, stop drinking milk formula, etc.) (18) Actions by hospital (i.e. provide treatment to victims, etc.) (19) Others (please specify) 7. Major focus (choose ONE major focus that is the dominant theme of the thread) (1) A description of the incident s (2) Solutions (3) Influence of the incident s (internationally and nationally) (4) Investigation of the incident s (5) Actions Please follow the directions in Figure D 1 for question 8 10: Figure D 1. A n example thread in online discussion forums in China.
166 8. Popularity of the thread: Record the number of hits for each thread 9. Responsiveness of discussion: Record the number of replies in each thread 10. Homogeneity of discussion: how the post replied to the initial entry in each thread, even if it was not a direct reply to the initial entry ( i.e., it replied to the second or third entry). Count the number of replies in each thread that: (each post can only be counted ONCE) (1) Pose any ar gument(s) to support the viewpoints in the initial entry (2) Propose any different or opposite idea(s) to the initial entry (3) Mention neutral or mixed points of view (4) Bring in new information/topic that is related to whats in the initial entry (5) Off topic replies such as ads (6) Others (please specify) 11. Tone of the thread: Count the number of posts (including the initial entry and replies) in each thread that: (a post can be counted twice. i.e., it includes information unfavorable to both government and dairy companies .). (1) Include any opinion(s) favorable to government or supportive of policy/officials (2) Include any opinion(s) unfavorable to government or any criticism of policy/officials (3) Include any opinion(s) favorable to dairy companies (4) Include any opinion(s) unfavorabl e to dairy companies (5) Include any neutral/mixed opinion(s) concerning the government, dairy companies, or victims (6) No tone presented 12. What sources are cited? For each thread, count the number of sources in all the posts in each following category. If a source, for instance a pediatrician, appeared several times in a story, it is counted only once. (1) State/local government/agency or their officials (2) Health agencies or their officials ( i.e., the Chinese Center for Disease Control and Prevention (CCDC), the State Fo od and Drug Administration of China (SFDA), the General Administration of Quality Supervision, Inspection and Quarantine (AQSIQ), the Ministry of Health, or other food or health related institutions) (3) NGOs, other independent organizations, and their officials (4) Health experts ( i.e., doctor/nurse/pediatrician, professor/researcher, food/nutrition expert, etc. that were not associated with the previous three types of organizations) (5) Hospitals and their officials (6) Dairy companies, their spokespersons, lawyers, etc. (7) Victims parents, family, lawyers, etc. (8) Other citizens, witnesses, etc. (9) Newspapers, TV stations, magazines, or news agencies (10) Blogs, online forums, independent online news sites, or other citizen media (11) Others (please specify)
167 13. Hyperlinks (count the number of hyperlinks in each category for each thread) Based on the destination hyperlinked to: (1) Internal hyperlinks (2) External hyperlinks to legacy news sites (newspaper, TV, news agency) (3) External hyperlinks to online news sites (Sina, Sohu, etc.) and citizen journalism sites (blogs, online forums, etc.) (4) External hyperlinks to government/health agency site (5) External hyperlinks to others (please specify) Based on the content hyperlinked to: (6) Hyperlinks to only multimedia content ( i.e., photos, videos, audio) (7) Hyperlin ks to only texts (8) Hyperlinks to multimedia and text together 14. Multimedia (count the number of different multimedia that appeared in the thread) (1) Embedded audio (2) Embedded videos (3) Embedded pictures/photos
168 APPENDIX E CODEBOOK FOR BLOG POSTS Blog post qualific ations: Include blog posts mainly focus on the tainted milk formula incident s Include posts with more than 100 words 1. Coder ID 2. Case number: (Start with 601) 3. Type of the blogger (Self identification of the blogger in blogs) (choose the major ONE only) (1) Heal th expert (2) Parent (3) Journalist (4) Citizen 4. Date of the blog post ( i.e., October 1, 2008 = 20081001) 5. Focus of the story (choose ALL that appeared in the blog post) (1) Infected symptoms (i.e., kidney stones, etc.) (2) Information on spread of the incidents (i.e., finding more tainted milk formula) or the spread of the infection (i.e., infected/deaths/confirmed cases) (3) Dairy products identified as contaminated (i.e., Sanlu, Shengyuan, etc.) (4) Formula recall (5) Guidelines for handling the tainted formula/finding untainted formula (i.e., where to return the tainted formula, etc.) (6) Guidelines for seeking medication/hospitalization (i.e., possible treatment for affected children, etc.) (7) Compensation to victims (i.e., how the compensation system works, suggestions to victims seeking comp ensations, etc.) (8) New regulations or monitoring system (i.e., new regulations concerning food safety, new standards for milk test, etc.) (9) Economics of the incidents/recall (i.e., bankruptcy of dairy companies, etc.) (10) Health or medical influence of the inciden ts (11) The politics of the incidents (i.e., legal restrictions on Chinese milk formula importation in foreign countries, etc.) (12) Possible causes for the incidents (i.e., outdated standards for formula safety test, etc.) (13) Problems of the food industry or government agencies (i.e. government corruptcy, dairy companies slow reactions, etc.) (14) Potential responsibility takers (i.e., government should be blamed, etc.) (15) Actions by government or health agency (i.e., free treatment to tainted formula victims, etc.) (16) Actions by dairy company (i.e., raise fund for victim compensation, etc.)
169 (17) Actions by individuals (i.e., go to hospital for body check, stop drinking milk formula, etc.) (18) Actions by hospital (i.e. provide treatment to victims, etc.) (19) Others (please specify) 6. Major foc us (choose ONE major focus that is the dominant theme of the post) (1) A description of the incident (2) Solutions (3) Influence of the incident (internationally and nationally) (4) Investigation of the incident (5) Actions 7. Frames appeared (choose ALL the frames appeared in t he post) (1) Attribution of responsibility frame: to blame the issue on the government, the dairy companies, or other stakeholders. Keywords: responsible ( zeren), problem ( wenti ), etc. (2) Severity frame: the spread of the tainted formula contamination, for exampl e, the infected/deaths/confirmed cases. Keywords: severe ( yanzhong), infected ( yingxiang), panic ( konghuang), etc. (3) Action frame: any action taken by stakeholders such as the dairy company, the government or the individual, to deal with the formula scandals such as formula recall, compensation, punishment, hospitalization of sickened kids, etc. Keywords: action ( xingdong), recall ( zhaohui ), hospitalization/medication ( yizhi ), etc. (4) Conflict frame: conflicts between individuals, companies, victimized groups, t he government, etc. Keywords: conflict ( chongtu, maodun), etc. (5) Economic consequence frame: the possible economic losses for the individuals, companies, the dairy industry or the country due to the formula scandals Keywords: loss ( sunshi ), pay ( chuqian), e tc. (6) Reassurance frame: tells the public do not need to worry or should be less worried about the formula scandals or food safety issues in China. Keywords: reassured ( fangxin), safe ( anquan ), etc. (7) Uncertainty frame: uncertainties in any aspect of the issue, including the cause, the cure, the possible spread, etc. Keywords: uncertain (buqueding), unclear ( buming), etc. (8) Others (please specify) 8. The most dominant frame (choose ONE dominant frame in the post from above) 9. What is the tone of the blog post? (1) To gove rnment: Negative ( hostile or aggressive language) = 0 Positive (optimistic, stressing logical and sensitive demeanor) = 1 Neutral/mixed (between or a mix of negative and positive language) = 2 No tone appeared (void of judgment, only statements of facts) = 3 (2) To the dairy company: Negative ( hostile or aggressive language) = 0 Positive (optimistic, stressing logical and sensitive demeanor) = 1 Neutral/mixed (between or a mix of negative and positive language) = 2
170 No tone appeared (void of judgment, only stat ements of facts) = 3 10. What sources are cited in the story? Count the number of sources in each blog post for each category. If a source, for instance a pediatrician, appeared several times in a story, it is counted only once. (1) State/local government/agency or their officials (2) Health agencies or their officials ( i.e., the Chinese Center for Disease Control and Prevention (CCDC), the State Food and Drug Administration of China (SFDA), the General Administration of Quality Supervision, Inspection and Quarantine ( AQSIQ), the Ministry of Health, or other food or health related institutions) (3) Non government organizations (NGOs), other independent organizations, and their officials (4) Health experts ( i.e., doctor/nurse/pediatrician, professor/researcher, food/nutrition ex pert, etc. that were not associated with the previous three types of organizations) (5) Hospitals and their officials (6) Dairy companies, their spokespersons, lawyers, etc. (7) Victims parents, family, lawyers, etc. (8) Other citizens, witnesses, etc. (9) Newspapers, TV stations, magazines, or news agencies such as the Xinhua News Agency (10) Blogs, online forums, independent online news sites, or other citizen media (11) Others (please specify) 11. Gender of the blogger: (1) Male (2) Female (3) Unknown (cannot discern) or no author listed 12. Times bei ng read (count the number of times being read for each blog post.) 13. Reader comments (count the number of reader comments for each blog post. If comment was forbidden by the blogger, code 999.) 14. Bloggers responses (count the number of comments left by the bl ogger. If comment was forbidden by the blogger, code 999.) 15. Type of reader comments (count the number of reader comments in each category) (1) The comment supports the viewpoints in the blog post or proposes similar idea(s) (2) The comment mentions or proposes diff erent or opposite idea(s) (3) The comment is neutral or mixed (neither supportive or opposite, or both supportive and opposite) (4) The comment is off topic ( i.e., ads) (5) The comment bring s in new information that is related to whats in the blog (6) Others (please spec ify) 16. Hyperlinks (count the number of hyperlinks in the content posted by the blogger) Based on the destination hyperlinked to: (1) Internal hyperlinks to posts by this blogger
171 (2) External hyperlinks to legacy news sites (newspaper, TV, news agency) (3) External hyper links to online media sites ( i.e., independent online news site like Sina, citizen journalism sites like blogs, etc.) (4) External hyperlinks to government/health agency site (5) External hyperlinks to others (please specify) Based on the content hyperlinked to: (6) H yperlink to only multimedia content ( i.e., photos, videos, audio) (7) Hyperlink to only texts (8) Hyperlinks to multimedia and text together 17. Multimedia (count the number of different multimedia in th e content posted by the blogger ) (1) Embedded audio (2) Embedded videos (3) Embedded pictures/photos
172 LIST OF REFERENCES Antaki, C., Ardvol, E., Nez, F., & Vayreda, A. (2005). For she who knows who she is : Managing accountability in online forum messages. Journal of Computer Mediated Communication, 11(1), 114 132. Atton, C. (2003). What is alternative journalism? Journalism, 4(3), 267 272. Babbie, E. R. (2007). The practice of social research. Belmont, CA: Wadsworth Publishing. Barak, A., & Gluck Ofri, O. (2007). Degree and reciprocity of self disclosure in online forums. CyberPsychology & Behavior, 10(3), 407 417. Baran, S. J., & Davis, D. K. (2009). Mass communication theory: Foundations, ferment, and future. Boston: Wadsworth Cengage Learning. Bandurski, D. & Hala, M. (Eds.) (2010). Investigative journalism in China. S eattle, WA: University of Washington Press. Beckett, C. (2008). SuperMedia: Saving journalism so it can save the world. Malden, MA: Wiley. Berkowitz, D. (1987). TV news sources and channels: A study in agenda setting. Journalism Quarterly, 64(2 3), 508 513. Berkowitz, D., & Adams, D. B. (1990). Information subsidy and agendabuilding in local television news. Journalism Quarterly, 67(4), 723 731. Berry, T. D., Wharf Higgins, J. & Naylor, P. J. (2007). SARS wars: An examination of the quantity and const ruction of health information in the news media. Health Communication, 21(1), 35 44. Bourdieu, P. (1993). Sociology in question. London: Sage. Bowman, S., & Willis, C. (2003). We media: How audiences are shaping the future of news and information. The me dia center: At the American Press I nstitute. Retrieved October 23, 2010, from http://www.hypergene.net/wemedia/download/we_media.pdf Brants, K. (2005). Guest editors introduction: The Internet and the public sphere. Political Communication, 22(2), 143 146. Bryant, J. & Oliver, M. B. (Eds.) (2008). Media effects: Advances in theory and research. New York, NY: Routledge.
173 Burger, G. (2009). How the Internet impacts on international news : Exploring paradoxes of the mo st global medium in a time of hyperlocalism. International Communication Gazette, 71(5), 355 371. Burgoon, M., Miller, M. D., Cohen, M., & Montgomery, C. L. (1978). An empirical test of a model of resistance to persuasion. Human Communication Research, 5(1), 27 39. Canel, M. J., Llamas, J. P., & Rey, F. (1996). El primer nivel del efecto agenda setting en la informacion local: Los problems mas importante de la ciudad de Pamplona [The first level agendasetting effect on local information: The most important problems of the city of Pamplona]. Communicacion y Sociedad, 9(1 2) 17 38. Carpenter, S. (2010). A study of content diversity in online citizen journalism and online newspaper articles. New Media & Society, 12(7), 1064 1084. Carpenter, S. (2008). How online citizen journalism publications and online newspapers utilizing the objectivity standard and rely on external sources. Journalism and Mass Communication Quarterly, 85(3), 531 548. Chaffee S. H., & Metzger M J. ( 2001). The end of mass communication? Mass Communication & Society, 4(4), 365 379. Chan, Z. C. Y., & Lai, W. (2009). Revisiting the melamine contamination event in China: Implications for ethics in food technology. Trends in Food Science & Technology, 20(8), 366 373. Chang, R. (2008). Tainted. Global Journalist, 14(4), 14 15. Chen, Z. (2010, July 22). China issues food safety warning. China Daily Retrieved October 21, 2010, from http://www.chinadaily.com.cn/china/201007/22/content_11037611.htm Chen, G., & Chiu, M. (2008). Online discussion processes: Effects of earlier messages evaluations, knowledge content, social cues and personal information on later messages. Computers & Education, 50(3), 678 692. China Internet Network Information Center (CNNIC) (2006). 2006 survey report of blog development in China. Retrieved Dec ember 11, 2010, from ht tp://www.cnnic.cn/uploadfiles/pdf/2006/9/28/182836.pdf (In Chinese). China Internet Network Information Center (CNNIC) (2010). The 26th statistical report on Internet development in China Retrieved October 19, 2010, from http://www.cnnic.cn/uploadfiles/pdf/2010/8/24/93145.pdf
174 China Media Guide. (2008). China Media Guide. Retrieved December 18, 2010, from http://www.danwei.o rg/media_guide/newspapers Co r nfield, M., Carson, J., Kalis, A., & Simon, E. (2005). Buzz, blogs, and beyond: The Internet and the national discourse in the fall of 2004 Pew Internet & American Life Project. Retrieved October 16, 2010, from http://www.pewtrusts.org/uploadedFiles/wwwpewtrustsorg/News/Press_Releases /Society_and_the_Internet/PIP_Blogs_051605.pdf Cohen, B. (1963). The press and foreign policy Princeton: Princeton University Press. Craft, S., & Wanta, W. (2004). U.S. public concerns in the aftermath of 911: A test of second level agendasetting. International Journal of Public Opinion Research, 16(4), 456 463. Cupach, W. R., & Canary, D. J. (1997). Competence in interpersonal conflict Prospect, IL: Waveland Press. Dahlberg, L. (2001). Democracy via cyberspace: Mapping the rhetoric and practices of three prominent camps. New Media and Society, 3(2), 157 1 77. Dailey, L., Demo, L., & Spillman, M. (2008). Newspaper political blogs generate little interaction. Newspaper Research Journal, 29(4), 53 65. Dearing, J. W., & Rogers, E. M. (1996). Agenda setting Thousand Oaks, CA: Sage. Delwiche, A. (2005). Agenda setting, opinion leadership, and the world of Web logs. First Monday, 10(12). Retrieved from March 1, 2011, from http://firstmonday.org/htbin/cgiwrap/bin/ojs/ index.php/fm/article/view/1300/1220 Deuze, M. (2003). The Web and its journalisms: Considering the consequences of different types of news media online. New Media & Society, 5(2), 203 230. Dewey, P. (1987). Essential guide to bulletin board systems Wes tport, CT: Mecklermedia. Dill, R. K., & Wu, H. D. (2009). Coverage of Katrina in local, regional, national newspapers. Newspaper Research Journal, 30(1), 6 20. Domingo, D., Quandt, T., Heinone, A., Paulussen, S., Singer, J. B., & Vujnovic, M. (2008). Par ticipatory journalism practices in the media and beyond: An international comparative study of initiatives in online newspapers. Journalism Practice, 2 (3), 326 342.
175 Dong, D., Chang, T., & Chen, D. (2008). Reporting AIDS and the invisible victims in China: Official knowledge as news in the Peoples Daily, 1986 2002. Journal of Health Communication, 13(4), 357 374. Dubrovsky, V., Kiesler, S., & Sethna, B. (1991). The equalization phenomenon: Status effects in computer mediated and faceto face decisionmak ing groups. HumanComputer Interaction, 6(2), 119 146. Dylko, I. B., & Kosicki, G. M. (2006). Sociology of news and new media: How the blogosphere transforms journalism and changes news Paper presented at the 2006 annual conference of the Association for Education in Journalism and Mass Communication (AEJMC), August 4 7, San Francisco, CA. East Day (2003). New daily ready to debut in Shanghai Retrieved December 18, 2010, from http://www.ch ina.org.cn/english/culture/68586.htm Edwards, G. C. & Wood, B. D. (1999). Who influences whom? The President, the media, and the public agenda. American Political Science Review, 93(2), 327 344. Entman, R. (1993). Framing: Toward clarification of a frac tured paradigm. Journal of Communication, 43(4), 51 58. Fallows, D. (2008). Most Chinese say they approve of government Internet control Pew Internet & American Life Project. Retrieved October 10, 2010, from http://www.pewinternet.org/~/media//Files/Reports/2008/PIP_China_Internet_200 8.pdf.pdf Farrell, H. & Drezner, D.W. (2008). The power and the politics of blogs. Public Choice 134(1 2), 15 30. Feng, J. & Wu, D. D. (2009). Changing ideologies and advertising discourses in China: A case study of Nanfang Daily. Journal of Asian Pacific Communication, 19(2), 218 328. Fleming, K., Thorson, E., & Zhang, Y. (2006). Going beyond exposure to local news media: An information processing examination of public perceptions of food safety. Journal of Health Communication, 11(8), 789 806. Frewer, Lynn. (1999). Risk perception, social trust, and public participation into strategic decision making Implications for em erging technologies. Ambio, 28(6), 569 574.
176 Friedland, L. A., & Nichols, S. (2002). Measuring civic journalisms progress: A report across a decade of activity Pew Center for Civic Journalism. Retrieved December 2, 2010, from http://www.pewcenter.org/doingcj/research/r_measuringcj.html Funkhouser, G. R. (1973). The issues of the sixties: An exploratory study in the dynamics of public opinion. Public Opinion Quarterly, 37(4), 6 2 75. Gamson, W. S. & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology, 95(1), 1 37. Gandy, O. H., Jr. (1982). Beyond agenda setting: Information subsidies and public policy Norwood, NJ: Ablex. Gans, H. (1979). Deciding whats news: A study of CBS Evening News, NBC Nightly News, Newsweek and Time. New York: Vintage Books. Garfield, B. (2006). YouTube vs. boob tube. Wired, 14(12), 222 227. Ghanem, S. (1997). Filling in the tapestry: The second level of agenda setting. In M. McComb s, D. L. Shaw, & D. Weaver M. (E ds.), Communication and democracy: Exploring the intellectual frontiers in agendasetting theory (pp. 3 14). Mahwah, NJ: Lawrence Erlbaum Associates. Giddens, A (1984). The constitution of society Cambridge: Polity Press. Gillmor, D. (2004). We the media: Grassroots journalism by the people. Sebastopol, CA: OReilly. Gillmor, D. (2010). MediActive Retrieved December 22, 2010, from http://mediactive.com/wpcontent/uploads/2010/12/mediactive_gillmor.pdf Glaser, M. (2006, September 27). Your guide to citizen journalism MediaShift. Retrieved December 2, 2010, from http://www.pbs.org/mediashift/2006/09/your guideto citizen journalism270.html Golan, G. & Wanta, W. (2001). Secondlevel agendasetting in the New Hampshire Primary: A comparison of coverage in three newspapers and public perceptions of candidates. Journalism and Mass Communication Quarterly 78(2), 247 259. Gold, D., & Simmons, J. (1965). News selection patterns among Iowa dailies. Public Opinion Quarterly 29 (3), 425 430.
177 Goldstein, J. & Rotich, J. (2008). Digitally networked technology in Kenyas 20072008 post election cris i s. Internet & Democracy Case Study Series. The Berkman Center for Internet & Society at Harvard University. Retrieved November 8, 2011, from http://unpan1.un.org/intradoc/groups/public/documents/undpadm/unpan042523.pdf Goode, L. (2009). Social news, citizen journalism and democracy. New Media & Socie ty, 11 (8), 1287 1305. Gwendolyn, B. (2009). Branding beef: Marketing, food safety, and the governance of risk. Canadian Journal of Communication, 34(2), 229 244. Habermas, J. (1989). The structural transformation of the public sphere: An inquiry into a c ategory of bourgeois society Cambridge, MA: MIT Press. Harwit, E. & Clark D. (2001). Shaping the Internet in China: Evolution of political control over network infrastructure and content. Asian Survey, 41(3), 377 408. Hellsten, L., & Nerlich, B. (2010). Bird flu hype: The spread of a disease outbreak through the media and Internet discussion groups. Journal of Language & Politics, 9 (3), 393 408. Heppler, K. (2010). How social media and mobile helps Haiti: Ushahidi, Samasource, CrowdFlower and FrontlineS MS envisionGood. Retrieved November 8, 2011, from http://envisiongood.com/interview ushahidi samasourcecrowdflower frontlinesms mission 4636helping haiti via tech mobile crowdsourcing social media/2010/02 Hester, J. B., & Gibson, R. (2003). The economy and second level agenda setting: A time series analysis of economic news and publi c opinion about the economy. Journalism & Mass Communication Quarterly, 80(1), 73 90. Huang, C. (2001). Chinas staterun tabloids: The rise of city newspapers. Gazette, 63(5), 435 450. Huang, Y. & Chen, J. (2009). Discursive democratization in mainlan d China: A diachronic study of the New Years editorials in the Peoples Daily. Journal of Asian Pacific Communication, 19(2), 199 217. Hurrell, A. C. (2005). Civility in online discussion: The case of the foreign policy dialogue. Canadian Journal of Comm unication, 30(4), 633 648. International Food Information Council (IFIC) (2010). 2010 food & health survey: Consumer attitudes toward food safety, nutrition, and health. Retrieved October 21, 2010, from http://www.foodinsight.org/Content/3651/2010FinalFullReport.pdf
178 Jacob, A. (2008, December 3). Chinese release increased numbers in tainted milk scandals The New York Times p. 1. James, M. L. & Wotring, C. E. (1995). An expl oratory study of the perceived benefits of electronic bulletin board use and their impact on other communication. Journal of Broadcasting & Electronic Media, 39(1), 30 50. Johnson, T. J., & Kaye, B. K. (2004). Wag the blog: How reliance on traditional media and the Internet influence perceptions of credibility of weblogs among blog users. Journalism & Mass Communication Quarterly 81(3), 622 642. Johnson, K. A., & Wiedenbeck, S. (2009). Enhancing perceived credibility of citizen journalism Web sites. Jour nalism and Mass Communication Quarterly, 86(2), 332 348. Kiousis, S. (2002). Interactivity: A concept explication. New Media & Society, 4(3), 355 384. Kiousis, S., Bantimaroudis, P., & Ban, H. (1999). Candidate image attributes: Experiments on the subst antive dimension of secondlevel agenda setting. Communication Research, 26(4), 414 428. Kiousis, S., & McCombs, M. (2004). Agendasetting effects and attitude strength: Political figures during the 1996 presidential election. Communication Research, 31(3 ) 36 57. Kiousis, S., Mitrook, M., Wu, S., & Seltzer, T. (2006). Firstand secondlevel agendabuilding and agendasetting effects: Exploring the linkages among candidate news releases, media coverage, and public opinion during the 2002 Florida Gubernat orial Election. Journal of Public Relations Research, 18(3), 265 285. Kosicki, G. (1993). Problems and opportunities in agendasetting research. Journal of Communication, 43(2), 100 127. Ku, G., Kaid, L. L., & Pfau, M. (2003). The impact of Web s ite c amp aigning on t raditional n ews media and public i nformation p rocessing. Journalism and Mass Communication Quarterly, 80(3), 528 547. Kwak, N., Williams, A., Wang, X., & Lee, H. (2005). Talking politics and engaging politics: An examination of the interactive relationship between structural features of political talk and discussion engagement. Communication Research, 32(1), 87 111.
179 Lacy, R. S., Riffe, D., Thorson, E., & Duffy, M. (2009). Examining the features, policies, and resources of citizen journalism: Citizen news sites and blogs. Web Journal of Mass Communication Research, 15(1). R etrieved September 30, 2010, from http://www.scripps.ohiou.edu/wjmcr/vol15/ Lacy, S., Duffy, M., Riffe, D., Thorso n, E., & Fleming, K. (2010). Citizen journalism Web sites complement newspapers. Newspaper Research Journal, 31(2), 34 46. Lane, R. (2010). Police departments restructure press pass policies to adapt to bloggers, freelancers. News Media & T he Law, 34(4), 30 32. Lee, B., Lancendorfer, K. M. & Lee, J. L. (2005). Agendasetting and the Internet: The intermedia influence of Internet bulletin boards on newspaper coverage of the 2000 general election in South Korea. Asian Journal of Communication, 15(1), 57 71. Lee, C. C. (Ed.). (2000). Power, money, and media: Communication patterns and bureaucratic control in cultural China. Evanston, IL: Northwestern University Press. Lee, J. (2006). Blog agenda: What did they blog about in the 2004 U.S. Presidential Election. Paper presented at the 2006 annual conference of the Association for Education in Journalism and Mass Communication (AEJMC), August 4 7, San Francisco, CA. Lenhart, A., & Fox, S. (2006). Bloggers: A portrait of the Internets new storytellers Pew Internet & American Life Project. Retrieved December 16, 2010, from http://www.pewinternet.org/Reports/2006/Bloggers.aspx Li, S. (2010). The online public sphere and popular ethos in China. Media, Culture & Society, 32(1), 63 83. Li, X, Xuan, Q., & Kluver, R. (2003). Who is setting the Chinese agenda? The impact of online chatrooms on party presses in China. In K. C. Ho, R. Kluver, and K. C. C. Yang (Eds.), Asia.com: Asia encounters the I nternet (pp. 143 158). London: RoutledgeCurzon. Light, A., & Rogers, Y. (1999). Conversation as publishing: The role of news forums on the Web. Journal of Computer Mediated Communication, 4(4). Retrieved December 20, 2010, from http://jcmc.indiana.edu/vol4/issue4/light.html Lim, J. (2006). A c ross lagged analysis of agenda setting among online media. Journalism and Mass Communication Quarterly 83 (2), 298 312. Lippmann, W. (1922). Public opin ion New York: MacMillan.
180 Liu, B. R. (2010). Chinese TV changes face: The rise of independents. Westminster Papers in Communication & Culture, 7(2), 73 90. Liu, Z. (2 010). Striving for hegemony in a public health crisis: Xinhua News Agencys framing of the Sanlu tainted formula scandals Paper presented at the 93rd annual conference of the Association for Education in Journalism and Mass Communication (AEJMC), August 4 7, Boston, MA. Lombard, M., Snyder Duch, J., & Bracken, C. C. (2002). Content analysis in mass communication: Assessment and reporting intercoder reliability. Human Communication Research, 28(4), 587 604. Lowrey, W., Evans, W., Gower, K. K., Robinson, J A., Ginter, P. M., McCormick, L. C., & Abdolrasulnia, M. (2007). Effective media communication of disasters: Pressing problems and recommendations. BMC Public Health, 7 (97). Retrieved October 21, 2010, from http://www.biomedcentral.com/content/pdf/147124587 97.pdf Luther, C. A., & Zhou, X. (2005). Within the b oundaries of politics: News framing of SARS in China and the United States. Journalism & Mass Communication Quarterly, 8 2 (4), 857 872. Ma, J. (2008, September 13). Complaints were received in June but nothing was done. South China Morning Post p. 4. MacKinnon, R. (2008). Flatter world and thicker walls? Blogs, censorship and civic discourse in China. Public Choice, 134(1 ), 47 65. Major, L. H. (2009). Break it to me harshly: The effects of intersecting news framing in lung cancer and obesity coverage. Journal of Health Communication, 14(2), 174 188. McCombs, M. (1994). The future agenda for agenda setting research. Journ al of Mass Communication Studies, 45(3), 181 217. McCombs, M. (1997). Building consensus: The news medias agenda setting roles. Political Communication 14 (4), 433 443. McCombs. M. E. (2004). Setting the agenda: The mass media and public opinion. Malden, MA: Polity. McCombs, M. E. (2005). A look at agenda setting: Past, present, and future. Journalism Studies 6 (4), 543 557.
181 McCombs, M. E., & Ghanem, S. I. (2001). The convergence of agenda setting and framing. In S. D. Reese, O. H. Gandy, & A. E. Gran t (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 67 82). Mahwah, NJ: Lawrence Erlbaum Associates. McCombs, M. E., Llamas, J. P., Lopez Escobar, E., & Rey, F. (1997). Candidate images in Spanish elections: Secondlevel agendasetting effects. Journalism and Mass Communication Quarterly 74 (4), 703 717. McCombs, M. E., Lopez Escobar, E., & Llamas, J. P. (2000). Setting the agenda of attributes in the 1996 Spanish election. Journal of Communication, 50(2), 7 7 92. McCombs, M. E., & Reynolds, A. (2002). News influence on our pictures of the world. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 1 18). Mahwah, NJ: Lawrence Erlbaum Associates. McCombs, M. E., & Shaw, D. L (1972). The agendasetting function of mass media. Public Opinion Quarterly, 36(2), 176 187. McCombs, M. E., & Shaw, D. L. (1993). The evolution of agendasetting research: Twenty five years in the marketplace of ideas. Journal of Communication, 43(2), 58 67. McCombs, M. E., Shaw, D. L., & Weaver, D. (Eds.). (1997). Communication and democracy: Exploring the intellectual frontiers in agendasetting theory Mahwah, NJ: Lawrence Erlbaum Associates. McLeod, J. M., Scheufele, D. A., & Moy, P. (1999). Comm unity, communication, and participation: The role of mass media and interpersonal discussion in local political participation. Political Communication, 16(3), 315 336 Meraz, S. (2009). Is there an elite hold? Traditional media to social media agenda sett ing influence in blog networks. Journal of Computer Mediated Communication, 14(3), 682 707. Messner, M., & Distaso, M. W. (2008). The source cycle, Journalism Studies, 9(3), 447 463. Messner, M. & Garrison, B. (2006). News for the w orld: A case study of English language newspapers in China. Paper presented at the 56th annual conference of the International Communication Association (ICA), June 19 23, Dresden, Germany. Miller, K. (2005). Communications theories: Perspectives, processes, and contexts New York: McGraw Hill.
182 Miller, E. A., Pole, A., & Bateman, C l. (2011). Variation in health blog features and elements by gender, occupation, and perspective. Journal of Health Communication, 16(2), 726 749. Mitchelstein, E. & Boczkowski, P. J. (2009). Be tween tradition and change: A review of recent research on online news production. Journalism, 10(5), 562 586. Morris, M., & Ogan, C. (1996). The Internet as mass medium. Journal of Communication, 46(4), 39 50. Moyo, D. (2009). Citizen journalism and the parallel market of information in Zimbabwes 2008 election. Journalism Studies, 10(4), 551 567. Muthukumaraswamy, K. (2010). When the media meet crowds of wisdom Journalism Practice, 4 (1), 48 65. National Bureau of Statistics of China. (2008). Press re lease on major figures of the 2010 national population census Retrieved September 28, 2011, from http://www.stats.gov.cn/was40/gjtjj_en_detail.jsp?channelid=1175&rec ord=59 Neuendorf, K. A. (2002). The content analysis guidebook London: Sage Publications. Nielsen, J. (2006). Participation inequality: Encouraging more users to contribute Retrieved November 8, 2011, from http://www.useit.com/alertbox/participation_inequality.html Nielsen China Forum. (2010). Go digital in China. Retrieved December 16, 2010, from http://blog.nielsen.com/nielsenwire/consumer/nielsenchina forum go digitalin china/ Nisbet, M .C., Brossard, D., & Kroepsch, A. (2003). Framing science: The stem cell controversy in an age of press/politics. Harvard International J ournal of Press/Politics, 8 (2), 36 70. Nucci, M. L., Cuite, C. L., & Hallman, W. K. (2009). When good food goes bad: Television network news and the spinach recall of 2006. Science Communication, 31(2), 238 265. Nucci, M. L., & Kubey, R. (2007). We begin tonight with fruits and vegetables: Genetically modified food on the evening news 19802003. Science Communication, 29(2), 147 176. Outing, S. (2005). The 11 layers of citizen journalism. The Poynter Institute. Retrieved December 2, 2010, from http://www.poynter.org/uncategorized/69328/the11layers of citizen journalism/
183 Polumbaum, J. (1990). The tribulations of Chinas journalist after a decade of reform. In C. C. Lee (Ed.). Voices of China: The interplay of politics and journalism (pp. 33 68). New York: Guilford Press. Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New York: Simon & Schuster. Press Reference (2010). Ch ina press, media, TV, radio, newspapers Retrieved December 18, 2010, from http://www.pressreference.com/BeCo/China.html Project for Excellence in Journalism. (2007). Digital journalism: A classification of news websites Pew Research Center. Retrieved December 2, 2010, from http://www.stateofthemedia.org/2007/narrative_overview_intro.asp?media=1 Project for Excellence in Journalism. (2011). The state of the news media 2011 Pew Research Center. Retrieved September 28, 2011, from http://stateofthemedia.org/ Qiang, X. (2004). The rising tide of Internet opinion i n China. Nieman Reports, 58(2), 103 104. Ramsey, S. (1999). A benchmark study of elaboration and sourcing in science stories for eight American newspapers. Journalism & Mass Communication Quarterly, 76(1), 87 98. Reese, S. D., & Danielian, L. H. (1989). Intermedia influence and the drug issue: Converging on cocaine. In P. J. Shoemaker (Ed.), Communication campaigns about drugs: Government, media, and the public (pp. 29 46). Hillsdale, NJ: Erlbaum. Reese, S. D., Rutigliano, L., Hyun, K., & Jeong, J. (2007). Mapping the blogosphere: Professional and citizenbased media in the global news arena. Journalism 8 (3), 235 261. Reich, Z. (2008). How citizens create news stories. Journalism Studies, 9(5), 739 758. Riffe, D., Lacy, S., & Fico, F. G. (2005). Analyz ing media messages: Using quantitative content analysis in research. Mahwah, NJ: Lawrence Erlbaum Associates. Roberts, M., & McCombs, M. (1994). Agenda setting and political advertising: Origins of the news agenda. Political Communication, 11(33), 249 262 Roberts, M., Wanta, W., & Dzwo, T. (2002). Agenda setting and issue salience online. Communication Research, 29(4), 452 465.
184 Robinson, S. (2009). Someones gotta be in control here : The institutionalization of online news and the creation of a shared journalistic authority. Journalism Practice, 1 (3), 305 321. Roche, J. P., & Muskavitch, M. A. T. (2003). Limited precision in print media communication of West Nile virus risks. Science Communication, 24 (3), 353 365. Rodgers, S. & Thorson, E. (2003). A socialization perspective on male and female reporting. Journal of Communication, 53(4), 658 675. Rogan, R. G. (2006). Conflict framing categories revisited. Communication Quarterly, 54(2), 157 173. Rogers, E. M. (1996). Upto date report. Journal of Health Communication, 1(1), 15 24. Salwen, M. B. (1995). News of Hurricane Andrew: The agenda of sources and the sources agendas. Journalism & Mass Communication Quarterly, 72 (4), 826 840. Sato, H. (2003). Agenda setting for smoking control in Japan, 1945 1990: Influence of the mass media on national health policy making. Journal of Health Communication, 8(1), 23 44. Schaffer, J. (2007). Citizen media: Fad or the future of news? Knight Community News Network. Retrieved October 23, 2010, from http://www.kcnn.org/research/citizen_media_report/ Schneider, S. (1996). Creating a democratic public sphere through a political discussion. Social Science Computer Review 14 (4), 373 393. Schudson, M (2003). The sociology of news New York: Norton. Schultz, T. (2000). Mass media and the concept of interactivity: An exploratory study of online forums and reader email. Media, Culture & Society, 22(2), 205 221. Schweitzer, S. J. (2005). Functionalities of online communication of practice. CALDOL Research Center. Retrieved December 19, 2010, from http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.116.8877 Seltzer, T., & Dittmore, S. W. (2009). Down, set, frame: Secondlevel agenda building and the NFL network carriage dispute. International Journal of Sport Communication, 2(3), 340 359.
185 Shah, D. V., Cho, J., Eveland, W. P. Jr., & Kwak, N. (2005). Information and expression in a digital age: Modeling Internet effects on civic participation. Communication Research, 32(5), 531 565. Shehata, A. (2010). Unemployment on the agenda: A panel study of agendasetting effects during the 2006 Swedish National Election. Journal of Communication, 60(1), 182 203. Shih, T., Wijaya, R., & Brossard, D. (2008). Media coverage of public health epidemics: Linking framing and issue attention cycle toward an integrated theory of print news coverage of epidemics. Mass Communication & Societ y, 11(2), 141 160. Shirky, C. (2009). Here comes everybody: The power of organizing without organizations Penguin Books. Shoemaker, P. J., Tankard, J. W., & Lasorsa, D. L. (2004). How to build social science theories Thousand Oaks, CA: Sage. Singer, J. B., Domingo, D., Heinonen, A., Hermida, A., & Paulussen, S. (Eds.) (2011). Participatory journalism: Guarding open gates at online newspapers New York: Wiley Blackwell. Singer, E., & Endreny, P. M. (1993). Reporting on risk: How the mass media portray accidents, diseases, disasters and other hazards New York: Russell Sage Foundation. Sparks, C. (2010). Chinas media in comparative perspective. International Journal of Communication, 4(1), 552 566. Takeshita, T., & Mikami, S. (1995). How did the mass media influence the voters choice in the 1993 General Election in Japan? A study of agenda setting. Keio Communication Review, 17(2), 27 41. Tannen, D. (1986). Thats not what I meant! How conversational style makes or breaks your relationships with others New York: William Morrow and Company, Inc. Thorsen, E. (2008). Journalistic objectivity redefined? Wikinews and the neutral point of view. New Media & Society, 10 (6), 935 954. Thurman, N. (2008). Forums for citizen journalists? Adoption of user generated content initiatives by online news media. New Media & Society, 10 (1), 139 157. Tremayne, M. (2004). The Web of context: Applying network theory to the use of hyperlinks in journalism on the Web. Journalism & Mass Communication Quarterly, 81(2), 237 253.
186 Tremayne, M. (2007). Blogging, citizenship and the future of media. New York: Routledge. Tremayne, M., Zheng, N., Lee, J., & Jeong, J. (2006). Issue publics on the web: Applying network theory to the war blogosphere. Journal of Computer Mediated Communication, 12(1), 355 371. Trumbo, C. (1995). Longitudi nal modeling of public issues: A n application of the agendasetting process to the issue of global warming. Journalism and Mass Communication Monographs 152 (2), 1 57. Voss, M. (2003). Why reporters and editors get health coverage wrong. Nieman Reports, 57(1), 46 50. Wallington, S. F., Blake, K., Taylor Clark, K., & Viswanath, K. (2010). Antecedents to agenda setting and framing in health news: An examination of prior ity, angle, source, and resource usage from a national survey of U.S. health reporters and editors Journal of Health Communication, 15 (1), 76 94. Wang, X. (2007). Online public spheres: How Internet discussion forums promote political participation in Ch ina Paper presented at the 93rd annual conference of the National Communication Association (NCA), November 15 18, Chicago, IL. Wanta, W., Golan, G., & Lee C. (2004). Agenda setting and international news: Media influence on public perceptions of foreig n nations. Journalism & Mass Communication Quarterly, 81(2), 364 377. Wanta, W., & Wu, Y. C. (1992). Interpersonal communication and the agenda setting process. Journalism Quarterly 69 (4), 847 849. Weaver, D., & Elliott, S. N. (1985). Who sets the agenda for the media? A study of local agenda building. Journalism Quarterly, 62(1), 87 94. Wei, R. (2009). The state of new media technology research in China. Asian Journal of Communication, 19(1), 116 127. White, D. M. (1949). The gate keeper: A case study in the selection of news. Journalism Quarterly, 27(3), 383 390. Willings Press Guide (2009). Willings press guide 2009: World (Volume 3). UK: Cision. Winter, J. P., & Eyal, C. H. (1981). Agenda setting for the civil rights issue. Public Opinion Quarter ly, 45 (3), 376 383. Wright, S., & Street, J. (2007). Democracy, deliberation and design: The case of online discussion forums. New Media & Society, 9(5), 849 869.
187 Wu, M. (2008). Measuring political debate on the Chinese Internet forum. Javnost/ The Public 15 (2), 93 110. Xenos, M., & Moy, P. (2007). Direct and differential effects of the Internet on political and civic engagement. Journal of Communication, 57(4), 704 718. Xin, X. (2010). The impact of citizen journalism on Chinese media and society. Jo urnalism Practice, 4(3), 333 344. Xu, F., Qiu, L., Binnes, C. W., & Liu, X. (2009). Br eastfeeding in China: A review. International Breastfeeding Journal, 4(6). Retrieved September 27, 2011, from http://www.internationalbreastfeedingjournal.com/content/pdf/174643584 6.pdf Xu, J. (2009). Examining the trust crisis for multi national corporations in China: A triplecase study of international public relations. J ournal of Development Communication, 20(2), 84 93. Yang, G (2003). The Internet and the rise of a transnational Chinese c ultur al s phere. Media, Culture & Society, 25(4), 469 490. Yang, J., & Stone, G. (2003). The powerful role of interpersonal communication in agenda setting. Mass Communication and Society, 6(1), 57 74. Yu, H. (2011). Beyond gatekeeping: J blogging in China. Journalism, 12(4), 379 393. Yu, J. & Aikat, D. (2006). News on the Web: Agenda setting of onli ne news in W eb sites of major newspaper, television, and online news services Paper presented at the 56th annual meeting of the International Communication Association (ICA), June 19 23, Dresden, Germany. Yu, K. (2002). Toward an incremental democracy and governance: Chinese theories and assessment criteria. New Political Science, 24 (2), 181 199. Zhang, L. (20 06). Behind the Great Firewall : Decoding Chinas Internet media policies from the inside. Convergence: The International Journal of Research into New Media Technologies, 12(3), 271 291. Zhao, L., & Lim, T. S. (2008). The tainted milk formula scandals : Another hard lesson for China. EAI Background Brief No. 406. Retrieved October 7, 2010, from http://www.eai.nus.edu.sg/BB406.pdf Z heng, Y., & Wu, G. (2005). Information technology, public space and collective action in China. Comparative Political Studies, 38(5), 507 536.
188 Zhou, X., Chan, Y., & Peng, Z. (2008). Deliberativeness of online political discussion: A content analysis of th e Guangzhou Daily website. Journalism Studies, 9(5), 759 770. Zhou, Y., & Moy, P. (2007). Parsing framing processes: The interplay between online public opinion and media coverage. Journal of Communication, 57(1), 79 98. Zoch, L. M., & Vanslyke Turk, J. (1998). Women making news: Gender as a variable in source selection and use. Journalism and Mass Communication Quarterly, 75(4), 762 775.
189 BIOGRAPHICAL SKETCH Fangfang Gao was born in Hangzhou, China. She received her B.A. in Journalism and Communicatio n as well as a bachelors degree in f inance, e conomics, from Zhejiang University, China, in 2005. During her college time, she was a reporter for the campus newspaper. Prior to her graduate study in the U.S., Gao worked in the Chinese government for one year. She graduated from University of Florida in August 2008 with the M.A.M.C degree and continued studying at UF to pursue a Doctor of Philosophy in mass communication. In 2011, she received her Ph.D. from the University of Florida. Gao s major research interests are journalism, new media and global communication. She has presented more than twenty refereed conference papers at international, national, and regional conferences During her graduate studies, Gao has received five conference paper awards and p ublished eight refereed journal articles.