臉書引言框架對科學新聞理解之影響 - 政大學術集成

全文

(1)國立政治大學國際傳播英語碩士學位學程 International Master’s Program in International Communication Studies College of Communication National Chengchi University. 碩士論文. Master’s Thesis. 臉書引言框架對科學新聞理解之影響. Framing Effect: The Influence of "Facebook Lead" on Science News Comprehension. Student: Pei-Ju Chen 陳姵如 Advisor: Tsung-Jen Shih 施琮仁教授中華民國 103 年 9 月 September 2014.

(2) 臉書引言框架對科學新聞理解之影響. Framing Effect: The Influence of "Facebook Lead" on Science News Comprehension. 研究生: 陳姵如 Student: Pei-Ju Chen 指導教授: 施琮仁 Advisor: Tsung-Jen Shih. 國立政治大學國際傳播英語碩士學位學程碩士論文. A Thesis Submitted to International Master’s Program in International Communication Studies National Chengchi University. In partial fulfillment of the Requirement For the degree of Master of Arts. 中華民國 103 年 9 月 September 2014.

(3) Acknowledgement Three years of study at IMICS has finally come to an end, I need to express my immense gratitude to a lot of people who have helped me and accompanied me on this long and fruitful way. Firstly, I would like to thank my thesis advisor, Professor Tsung- Jen Shih for his patient instruction throughout the whole process. With his professional expertise in the research field, he helped me narrow down the research scope and examined my work in every detail. Secondly, I would also like to thank my two angelic committees, Professor YiNing Chen and Professor Yu-Chan Chiu, who gave me valuable feedbacks to refine and improve my experiment. Besides, I also want to thank all my dear friends who helped me during my pretest. You gave me a lot of wonderful ideas and suggestions to solve my research design problems. Lastly, I would like to thank Yu-Ping Chen for helping me distribute the reward to my participants while I was doing my exchange in Mannheim, Germany..

(4) 1. Abstract As social media continue to gain popularity, scientific news agencies also adopt the trend to increase online presence and fuel exposure because of the growing need to strengthen the communication between scientists and the general public. This study investigated the potential power of the current most popular social media- Facebook in fostering public understanding of science by employing two one-factor betweensubject web-based experimental factorial designs. Experiment 1 used genetic modification news and experiment 2 used nanotechnology news. Each piece of science news had three different Facebook framed leads: technique, conflict and human interest. Research results found human interest frame only served to arouse reader’s situational interest to click on the news feed to access to the full story but didn’t affect how readers interpret the main theme of the science news articles. On the other hand, conflict frame not only enhanced reader’s intention to click on the news feed but also reinforced how they perceive the main theme of the news story. However, the two soft news frames have no effect on reader’s perceived news attention, understanding and relevance. This study aimed to contribute to the literature as one of the first studies on the use of the social media in the realm of the public understanding of science.. Key words: framing, human interest, conflict, comprehension, Facebook.

(5) 2. 中文摘要隨著社群媒體的熱潮持續延燒，許多科學新聞機構也開始透過臉書的粉絲專頁來發佈自己的新聞以增加新聞曝光率。過去有關於媒體粉絲專頁的研究多探討行銷行為，鮮少針對新聞機構的臉書專頁做研究，對於粉絲專頁社群編輯的引言所形成的框架效果研究也付之闕如。本研究包含兩個實驗法，將兩則不同主題的科學新聞（基因改造、奈米科技）各自配上三種不同框架的臉書引言-技術性框架、爭議性框架、和人情趣味框架，來探討臉書的引言對於科學新聞閱聽眾的影響。結果發現，人情趣味的臉書引言只會增進讀者想要點擊新聞全文的慾望，卻不會對新聞主文的詮釋造成影響。爭議性臉書引言除了增進讀者閱讀新聞全文的慾望之外，也會再強化讀者對新聞全文的認知。然而這兩種框架對於讀者的新聞注意度、理解度和議題相關性沒有影響。. 關鍵字：框架、人情趣味、爭議、理解、臉書.

(6) 3. Table of Contents Introduction…………………………………………………………………………..5 Literature Review…………………………………………………………………….8 Research on Facebook………………………………………………………………....8 Functions of headline……………………………………………………………….....9 Framing effect of headline and lead……………………………………………….....11 Science news frames………………………………………………………………....13 Method……………………………………………………………….........................16 Study1………………………………………………………………...........................16 Participants………………………………………………………………...................16 Design………………………………………………………………...........................16 Material………………………………………………………………........................17 Measures……………………………………………………………….......................18 Procedure………………………………………………………………......................21 Results……………………………………………………………….........................23 Study2…………………………………………………………...................................28 Participants…………………………………………………………...........................28 Design………………………………………………………………...........................28 Material………………………………………………………………........................28 Measures………………………………………………………………......................29 Procedure………………………………………………………………......................31 Results……………………………………………………………….........................32 Discussion……………………………………………………………........................36 Limitation and future studies………………………………………........................41 Conclusion…………………………………………………………….......................44 References…………………………………………………………..........................46 Appendices………………………………………………………..............................52 Questionnaire………………………………………………………..........................52 Text of Facebook framed leads……………………………………...........................65.

(7) 4. Lists of Tables Table 1. Descriptive analysis of independent variables in GM news..........................19 Table 2. Descriptive statistics of dependent variables in GM news experimental condition………………………………………………………………………….…..20 Table 3. ANOVA: Perceived frame of GM news readers…………………………....24 Table 4. ANOVA: Intention to click on GM news…………………………………..25 Table 5. Results of Chi-square Test and Descriptive Statistics for readers’ thematic thought on GM news by Facebook lead frame……………………………………...26 Table 6. Descriptive analysis of independent variables in Nanotech news…………..29 Table 7. Descriptive statistics of dependent variables in Nanotech news experimental condition……………………………………………………………………….……..31 Table 8. ANOVA: Perceived frame of Nanotech news readers…………………...…33 Table 9. ANOVA: Intention to click on Nanotech news………………………….…34.

(8) 5. Introduction Science and technology have long played an important role in modern lives. The growing amount of science and technology- related content in news is something that cannot be underestimated as science is highly intertwined to policy making policy making and is also a part of culture per se (Nelkin, 1995). In the past, the traditional dominant paradigm of science communication was based on deficit model, which assumed that when public were provided with more information, they would become more science literate (Bauer, Allum & Miller, 2007; Hart & Nisbet, 2012). As most science laypersons rely on media for new science information, how to make science news more applicable to the public has long been academia’s focus. Researches have studied the relationship of science communication and television, newspaper, magazine and Internet (Pellechia, 1997; Miller, Augenbraun, Schulhof &Kimmel, 2006; Huang & Southwell, 2009). However, over the courses of years, regardless of public’s accessibility to increasing amount of science and technology news via media and their interest in science news, public scientific literacy remains relatively low. In recent years, as a result of digital convergence and development of communicational technologies, scholars from various disciplines all turn to explore the power of social media, including science communication. Li (2012) indicated that new media have the characteristics such as abundant diverse users, sociality, instant dissemination, and two-way communication; thus, they would play an even more important role in science communication, compared to mass media. Besides, scientists also recognize the grave importance of incorporating social media into their work to improve their research efficiency, flexibility and productivity and how they share knowledge with scientists and general public. Among all the new media, Facebook has received most attention. For example, Brian Sidlauskas, an ichthyologist,.

(9) 6. successfully identified more than 5000 fish species by uploading photos and tagging friends on Facebook. Within 24 hours, about 90 percent of the posted fish are correctly identified by his Facebook friends (Woods, 2011). Another group of scientists even employed Facebook to test public’s health condition (Venkataramanan, 2011). In addition, Goldstein (2013) indicated that Facebook actually helps scientists reach a greater different audience than Twitter or blogs. Last but not least, there are also more and more Facebook Fan Pages which aim to advance public’s scientific knowledge and to bridge science with the public such as Huffpost Teck, Science | The New York Times, Guardian technology, BBC science news, National geography being set up as a result of growing scientific news audience. Oftentimes, these scientific news organizations form an online community by liking each other’s Fan Page. Consequently, Facebook users may be accidently exposed to a scientific content on Facebook that originated from online news outlets with a lead-in identifying the sources or from individual users even he does not subscribe to scientific Facebook Fan pages, as Facebook has become a place where information is far beyond what individual user seeks. What’s worthy of note is that when Facebook Fan Page editors release or share news via hyperlinks, they usually add a text to summarize the content of the piece of news so as to enhance audience’s situational interest to access to the original news stories. The texts, or “Facebook lead” that are added to explain the news contents are not necessary a part of the original news text, but they function as summary or attention grabber, which work similar as headline or lead. Even though these commentators/ lead producers do not participate in the real production of news coverage, how they frame the issue by adding a “Facebook lead” may influence audience’s intent to read and how they interpret news as well. On top of that, the top-.

(10) 7. down layout design of Facebook may suggest audience to see the “Facebook lead” first before clicking on the news hyperlink and linking back to original news websites. Combined these factors together, people may decide which piece of information they want to see solely based on “Facebook lead”. Thus, how Fan Page editors frame the issues to pique audience’s situational interest to click on the news, and how do Facebook lead influence reader’s representation of news are noteworthy. This mechanism is especially important for science news- the hard news that demands more cognitive efforts for comprehension (Carmines & Stimson, 1980; Snow, 1980; Chen, 2011). To enhance readers’ news interest and understanding in science news, past researches have studied the effect of news text and hyperlink structures (Yaros, 2006; 2011), as well as news values (León, 2008). As for the framing research, most of them studied the framing effect on headline, only a few on lead, but not Facebook lead. In an attempt to address the gap, this work would like to study the influence of Facebook lead on science news, drawing on the study of framing by applying the two soft news values. This study can help scientific Facebook Fan Page editors to know better about how to retain reader’s interest, and how issues are interpreted. This study chooses Facebook as the researched medium; for it is the worldwide fastest growing and also Taiwanese’s most frequently used social media..

(11) 8. Literature Review Research on Facebook Facebook, founded by Mark Zukerberg in 2004, is now the most popular social networking sites (Grove, 2013), and the primary social media currently being widely used in newsrooms (Lysak, et al., 2012). It has more than 751 million monthly active users as of March 31, 2013 all over the world and each user joins an average of 80 Fan Pages (Facebook, 2013). Facebook has received significant research attention from many disciplines. Wilson, Gosling and Graham (2012) conducted a most comprehensive study on Facebook researches and categorized Facebook studies as follows: descriptive analyses of users (24%), motivations for using Facebook (19%), identity presentation (12%), role of Facebook in social interactions (27%), privacy and information disclosure (18%). Up to now, despite its theoretical importance on news dissemination, only a few studies have been done, and little attention has been paid to its power in science communication. Pew research center (2011) found out that though Google is still the biggest traffic driver for the top 25 news organizations in the U.S., the power of Facebook cannot be underestimated. When browsing news feeds, 13 percent of news on Facebook is from news organizations’ Fan Page or journalists’ posts. Besides, Facebook accounts for around 6 to 8 percent of the total traffics of the highly revered news agencies such as New York Times and Huffington Post. Another yearly review also conducted by Pew Research center (2013) pointed out social media use results in deeper news consumption: 15 % of adults get news from their family and friends via social media, and nearly 77 percent of them follow the links to the original stories; while for the youngsters aged 18- 29, nearly a quarter of them relies on social media to get current news. Mobile communication continues to fuel the fashion (Sasseen, Olmstead &.

(12) 9. Mitchell, 2013). In Taiwan, social networking sites have been greatly integrated into Taiwanese people’s lives as well; 1/3 of the top 100 sites in Taiwan are social networking sites, and Facebook has been in the first place since 2011 (BusinessNext, 2013). Taiwanese Facebook users spent an average of 318 minutes and browsed 417 pages per month (comScore Media Metrix, 2014). The number of news channels that direct Taiwanese Facebook users to news organizations’ official Web sites grew from 29 to 45 from July 2010 to July 2011 (comScore Media Metrix, 2013). These data all suggest that Facebook has become an important player in information dissemination. To understand the strategies online editor use in grabbing Facebook users’ attention among information sea, Cheng (2011) interviewed several Taiwanese news workers and reached some consensuses. Firstly, Facebook is beneficial to membership fosterage and brand image. Secondly, to generate more responses and get more “like”s from audiences and thus gain more news exposure, fan page editors need to understand audiences’ preference, use cordial and less serious tone when making Facebook lead. In short, Facebook’s Fan Page service enables news organizations to reach not only just active readers. The more “likes” a post contains, the higher chance for passive and less interested readers to be exposed to the content and click on the news. Functions of headline When reading newspaper, headlines are easily recognizable by the readers at the top of the story, and usually in larger and bolder print. Readers regularly scan the headlines, and only occasionally stop to read the actual story because they are flooded with overly amount of information that they have neither the time nor the energy to consume on the daily basis. Newspaper headlines, in this case, help readers to get the maximum implication from the stories with the minimal cognitive effort. Several.

(13) 10. studies have pointed out the two functions of news headline. It is not just a simple combination of words but with a complex selecting mechanism behind. Semantically, headline is the summary that explains accurately the whole stories in short words; pragmatically, it is an attention-grabbing device, which relates readers to the story (Bowles and Borden 2000). Together, headline give readers the optimally relevance to the news stories. The study conducted by Dor (2003) investigated the functions of news headline by interviewing news workers, results found the choices for news headline can generally be reduced to the notion of relevance optimization, not matter it is a summarizing, localizing or quotation headlines. Tewksbury, Weaver and Maddex (2001) also pointed out the importance of news headline in piquing audience’s situational interest in computer-mediated news. They serve as cognitive shortcut and guide individual readers to those specific stories that are worthwhile for them to take time to read the full version. In the specific domain of science journalism, another research done by Lu (2008) also pointed out the importance of news headline and lead of science news in attracting audience’s attention. For instance, scientific findings which are published in “Science ” or “ Nature” magazines under strict scrutiny are not necessary suitable for science news reporting because they are not interesting, special or top-ranking enough. To make them suitable in the news coverage, science journalists usually include keywords such as “The first”, “Taiwan”, “breakthrough” in the headline so as to attract editor’s attention. It is said science news that is more related to public’s lives, more interesting or with attention-getting headlines are editor’s priorities. Analogous to headline, in the Facebook setting, leads dominate all the accompanying information such as hyperlink, and picture as a result of top-down layout design. When Facebook user scroll down the page, leads come to their first.

(14) 11. sight. Oftentimes, they make choice on the limited information provided by the Facebook lead. Presumably, this same effect occurs when people browse Facebook news leads: when reader’s interest is aroused long enough, they would click on the news link and read the accompanying story. This situation raises some important and interesting questions: how does Facebook lead attract readers’ attention? Do Facebook readers interpret the news text through the lens of the Facebook lead? And given the researcher’s interest in the science news, the study focuses specifically on science news. Framing effect of headline and lead As a matter of fact, studying the effect of headline/ lead on the news text cannot be done without the concept of framing. Because space limitations make it impossible for headlines and leads to tell the whole story, journalists emphasize certain aspects over others. Framing is the central practice in journalism, the frame journalist choose for any story is their most pivotal decision (Smith, 1997). It can be defined as a set of interpretive packages that give meaning to an issue “making sense of relevant events, suggesting what is at issue” (Gamson and Modigliana, 1989). Entman (1993) offers a unified definition: 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 promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described. (p. 52) Other than establishing the salience of certain aspect over others, news framing research also noted that how news is presented affected how people interpreted issues and events and what they thought about people (Iyengar 1991; Price and Tewksbury.

(15) 12. 1997; Price, Tewksbury, and Powers 1997; Scheufele 1999). Scholars have recognized the importance of looking at the relationship between news headline and the stories that follows (Geer and Kahn, 1993; Tannenbaum, 1953), and argued about the impression that news media leave on a person would vary when readers read all the available information provided in the news article and only headline. Current studies on the framing effect of headline on news have produced mixed results. For example, Blake (2007) in his study of political news found voters who scanned election headlines without paying close attention to the news article were supplied with a different set of heuristic cues compared to those who pay closer attention to the news content. Another example provided by Geer and Kahn (1993) in their experimental design also proved that different headline: issue, horse race and trait centered had significant impact on readers’ perception of the news article’s main point. However, in a quantitative study conducted by Condit et al (2001) found out that the frames offered by the headline, did not influence readers in a molecular genetics news experiment, because readers have the ability to discern headline from news text. As for the framing effect of lead on news, current studies are scare but still indicated some common mechanisms in producing leads. Other than just filling in plain facts, journalists usually include human drama, because human has the nature to know more about what happened around us. Toward that end, two prominent soft news frames- conflict frame and human interest frames are among the most frequently used attention-grabbing devices and are able to make hard issues more accessible to readers (Van Dijk, 1988; Jebril, et al, 2013). Firstly, conflict frame which includes combat is the most frequently used frame in piquing reader interest (MacDougall ,1982). It highlights incompatibilities, disagreements, or oppositional tensions between individuals, groups, and institution.

(16) 13. by labeling controversial situation (Neuman, Just & Crigler,1992). Secondly, human interest frame personalizes, dramatizes and emotionalizes a(n) event, issue or problem, through this mean, it contributes to the narrative character of news. Drawing on these, Chen and Zillmann (2002, 2004) conducted two researches that are most relevant to this paper. They studied the effect of lead framing on the reader’s attention in the print and online news respectively. Firstly, they tested the effect of lead framing on selective exposure in the printed magazine. Results found leads which detail conflict and agony over suffered misfortunes were found to foster attention, but leads that only emphasize misfortunes per se or its economic implication are with no result. Based on this study, they further tested the lead framing on the selective exposure in the Internet context. They presumed lead framing would have greater influence on news attention than in the printed magazine because online readers have different reading behaviors based on the fact that limited information such as headlines are provided in the Web page as the result of page limit, but they reached the same conclusion. Apart from studying the effect of soft news frame on news attention, others also studied their influence on news understanding. Baum (2002, 2006) found soft news that includes human interest and conflict elements could make uninterested citizens to be more aware of relevant political information, and news attention is the perquisite to political news learning. Jebril et al (2013) further indicated that the effect of soft news frame on political news learning was stronger for readers who were least motivated and less interested in politics. Science news frame Because this paper intends to investigate the relationship of lead framing on science news on Facebook, we need to understand how science coverage is usually.

(17) 14. framed first. Studies have shown that science news has its own recurring frames to communicate the results of scientific research to science- illiterate audiences (Nisbet & Scheufele, 2007). For example, Nisbet, Brossard, and Kroepsch (2003) discovered most news coverage of science was often framed in progress that intellectual progress in science and technology leads to social progress. Another study from McComas and Simone (2003) showed that there has been a growing number of science news that highlight the conflict of interest as a result of competitive news industry in a ten-year long content analysis of reputed American press. Conflict elements in the science news offer dramatic energy and thus the incentive to appeal news audiences. Besides, Stewart, Dickerson and Hotchkiss (2008) concluded the commonly employed frames in science news coverage included progress and economic prospects frames, political controversy and ethical dilemmas (Dahinden, 2002) as science are becoming more intertwined with politics and economics (Nisbet & Huge, 2006). The finding is similar to Weaver, Lively and Bimber (2009)’s, though specifically focused on Nano scale science and technology, they categorized the commonly used science news frame into four categories: progress, regulation, conflict, and generic risk. Based on these studies, we can see how science is framed in the media also gone through a change from overwhelmingly optimistic to diverse angles. After reviewing these literatures, we found despite the theoretical importance of Facebook lead, only limited attention has been paid to the relationship actual Facebook lead has with the stories they introduce. As lead contains more words than newspaper headline, this study presumes it will have bigger effect than headline in influencing people’s perception on the news and how it solicits reader’s situational interest. The study choose human interest and conflict frame as they are most widely.

(18) 15. adopted and are considered to be the most effective ways in softening hard issues. This paper has the following five hypotheses: H1: Readers who are exposed to Facebook lead with the conflict frame or the human interest frame will be more likely to click on the news feed, compared with those who are exposed to the technique frame. H2: Readers who are exposed to Facebook lead with the conflict frame or the human interest frame will pay more attention to the news, compared with those who are exposed to the technique frame. H3: Readers who are exposed to Facebook lead with the conflict frame or the human interest frame will have higher news understanding, compared with those who are exposed to the technique frame. H4: Readers would judge the main theme of the news article based on the frame of Facebook lead. H5: Readers who are exposed to Facebook lead with the conflict frame or the human interest frame will perceive the issue to be more related to themselves, compared with those who are exposed to technique frame..

(19) 16. Methods As this paper intends to investigate the framing effect of Facebook lead on science news and to ensure the generalizability of the results, this study conducted two onefactor between-subject experiments. Because science news spectrum is wide (Huang & Jian, 2006), two studies used bipolar science topics. Study 1 used genetic modification news, and study 2 used nanotechnology news. In this chapter, two studies were presented separately and would be discussed together in the next chapter. Study one: Genetic modification news Participants A total of 89 participants were recruited online at a large public university in Taipei, Taiwan. Respondents were offered NT $100 for their participation, which took about 15 minutes. The mean age was 22.51 years (SD = 3.25), with 56 people being female (62.9%). The majority of the participants (39.3%) reported that they were graduate students. Among all the participants, only 7 people (7.9%) were from the science background, and 34 people (38.2%) have subscribed to scientific Facebook fan pages. Participants had an average 4.43-year –Facebook using experience (S.D. = 1.17) Design A one-factor between-subject design was employed to measure the framing effect of Facebook lead on readers’ intention to click, news attention, news understanding, news theme and perceived relevance. To enhance internal validity of the study, the experiment was carried out in the field (online). This study chose online field experiment despite the argument has long been noted that laboratory experiment can ensure the success of causal relationship and make sure the process to be carried out.

(20) 17. without interruption and being confounded by other factors as a result of strict control. Due to the popularization of Internet and the growing online population, it is reasonable to adopt online-filed experiment because it is closer to the real situation, especially when the research medium is Facebook. Online experiment enables readers to access to the Internet in their familiar surroundings without the time and space restriction and it is rather cheap and can minimize the wrong coding problem by accidence. Besides, the advancement of the online survey platforms also avoids the concerning problems such as multiple long on from the same user. And to enhance external validity, the study employed actual online news. Material In the pretest, a convenience sample of 15 university and graduate students were asked to rank the most unfamiliar scientific news topics based on the science news categorization proposed by Huang and Jian (2006). The scientific news topic ranked as the most complex, least familiar and most important was medical science. As a result, a piece of genetic modification (GM) news from BBC. Chinese. com, released on march 20, 2013 about England legalizing the 3-person-IVF technic, detailing how the technic can help genetic dysfunctional mitochondria disease and related discussion on the controversy of this issue was chosen. Three respective Facebook lead were designed to meet the framed condition: technical, conflict and human interest. Each lead was similar in length and in terms of writing style. Each groups contained the same core elements- the legalization of the Act of mitochondrial DNA disease along with different information exclusive to that particular frame. Technique framed lead focused on the introduction to the disease and the medical technique itself, which contained professional medical jargons such as carbohydrate, Amino acid, and adenosine triphosphate, etc (236 Chinese words). Conflict framed lead included the.

(21) 18. opinions toward the Act from the proponent and opponent parties (256 Chinese words). Lastly, human interest framed Facebook lead described the personal misfortune of a hopeless mother whose beloved son suffered from the rare genetic disease (199 Chinese words). The full text of three framed leads are listed in the appendix. Measures (1) Independent variable Independent variables includes age, gender, study area, highest education, year of Facebook use, scientific media use, scientific promise, subscription to scientific Facebook fan page, familiarity with and interest in the scientific topic. Only scientific promise and scientific media use were constructed indices. The means of the independent variables in each of the experimental conditions are reported in Table 1. There were no substantial differences between the means and standard deviations of all the independent variables between groups determined by one way-ANOVA. As the independent variables were asked before the manipulation, this shows that the random assignment indeed resulted in equal groups. The index of science promise adopted the questions from Bann & Schwerin (2004)’s study, on a 5-point- scale (1= strongly disagree to 5= strongly agree) asking (1) science and technology are making our lives healthier, easier and more comfortable (2) most scientists want to work on things that will make life better for the public and (3) because of science and technology, there will be more opportunities for the next generation. Score from the three items were summed and then divided by three (M = 3.95, SD =0.46, Cronbach’s α =0.58). Participants’ attitude toward science was rather promising..

(22) 19. The index of science media use was constructed by summing the scores of exposed frequency, reading/watching/ browsing frequency, and the perceived attention while reading/watching/ browsing in newspaper, television, magazine, movie, Internet and social media (1= very low to 5= very high), and divided by 18 (M=3.00, SD=0.53, Cronbach’s α =0.91 ).. Table 1 Descriptive analysis of independent variables in GM news Genetic Modified news (n=89) Mean SD Gender Female (62.9%) _____ Age 22.51 3.25 Internet experience (year) 10.79 2.66 Facebook experience (year) 4.43 1.17 Subscription to science fan page Yes (38.2%) _____ Science promise 3.95 0.46 Science media use 3.00 0.53 Familiarity with scientific topic 2.81 0.73 Individual interest in the topic 3.37 0.72 (2)Dependent variables Intention to click. This variable is composed of four items based on the studies of Xu (2013) and Genova and Greenberg (1979). It was measured on 5-point-scale (1=strongly disagree, 5= strongly agree), asking (1) I would like to click on the news feed to access to the full story (2) the Facebook lead is very interesting and (3) I am very into the Facebook lead. Scores from three questions were summed and averaged to construct the index for intention to click (M = 2.83, SD =0.76, Cronbach’s α =0.86). News attention. News attention is measured by three questions (1= strongly disagree, 5= strongly agree) based on the studies of Wang (2006) and Chang (2012) asking (1) I am very focused on the news (2) the news grabs my attention and (3) I am very concentrated on the news. Scores from these three questions were summed and.

(23) 20. averaged to form the index of news attention (M = 2.84, SD =0.81, Cronbach’s α = 0.91). News Understanding. Measure for this variable were two items to assess perceived news understanding on five-point scale (1= strongly disagree, 5= strongly agree) proposed by Chang (2009), asking (1) I understand the news text and (2) the news is easy to read. Scores were summed and averaged to form the index of news understanding (M = 3.11, SD =0.37, correlation= 0.555. p=0.00). Thematic thought. This variable is measured by a multiple-choice question that asked participants to choose the most important theme of the news. Questions were designed to meet the frame treatment, three questions are as follows: (1) introduction to Mitochondria disease (Technic frame) (2) the controversy of 3-person IVF technic (conflict frame)(3) the story of British woman-Sharon (human interest frame). Perceived relevance. This experiment measured perceived relevance through three items adopted by Chang (2009) on the five-point scale (1= strongly disagree, 5= t strongly agree) asking (1) the news is relevant to me (2) the news is important to me and (3) the news is important to society. Scores were summed and averaged to form the index of perceived relevance (M = 2.83, SD =0.61, Cronbach’s α = 0.763). Table 2 Descriptive statistics of dependent variables in GM news experimental condition Genetic Modified news (N=89) Mean SD Cronbach’s α /correlation Intention to click 2.83 0.76 α =0.86 (1) I would like to click on the news feed to access to the full story (2) The Facebook lead is very interesting (3) I am very into the Facebook lead. News attention 2.84 0.81 α =0.91 (1) I am very focused on the news (2) The pieces of news grabs my attention (3) I am very concentrated on the news News understanding 3.11 0.37 r=0.56** (1) I understand the news text p=0.00.

(24) 21. (2) The news is easy to read (reversely coded) Thematic thought (1) Mitochondria and introduction to the illness (2) The conflict of 3 person IVF technic (3) The story of British woman- Sharon Perceived relevance (1) The news is relevant to me (2) The news is important to me (3) The news is important to society Note. **Correlation is significant at 0.01 level. Procedure. --. --. --. 2.83. 0.61. α =0.76. Participants were first recruited online 2 weeks before the official experiment took place (June 12-14, 2014), telling them briefly what the experiment was about. Then participants received a reminder mail two days before the official online experiment. On the designated date, participants received a mail that included the online survey link, and they were requested to complete the online experiment at any time and place with their own convenience within 3 days. Once they started the online experiment, they could not quit in the middle of the process because the survey system disable them from logging in for the second time, and answers without full completion would be treated as useless case. Respondents were randomly assigned to access one of the three unique Web addresses, each containing the same questionnaire but with different links which direct readers to their assigned Facebook Fan page: (a) genetic modified news with technical frame Facebook lead (n=32) (b) genetic modified news with conflict frame Facebook lead (n=29) (c) genetic modified news with human interest frame Facebook lead (n=28). After accessing their assigned Web site, participants first completed some basic questions asking age, gender, study area, highest education, year of Facebook use, scientific media use, scientific promise, subscription to scientific Facebook fan page, familiarity with and interest in the scientific topic. After they finish the first part, they.

(25) 22. were first shown a screen shot picture of a framed lead depending on which framed group they were assigned to. The reason why they were only exposed to a screen shot instead of the actual Facebook setting at this moment was to make sure the experimental manipulation was successful. Right after they saw the screen shot of the framed Facebook lead, they were asked to answer 6 questions concerning the frame manipulation. Afterwards, they were directed to the actual Fan page. At this stage, they were asked to read the article based on their personal reading behavior on Facebook setting. In the end, participants have to finish the following questions including the measures of intention to click, news attention, news understanding, news theme, and perceived relevance. After submitting their responses, they were exposed to a “thank you” Web page. For the full questionnaire please see the appendix..

(26) 23. Results One-way analysis of variance (ANOVA) tests for homogeneity revealed no significant differences between participants in scientific promise, media use, familiarity with and individual interest in the topic. One-way ANOVA and chi-square test are used in the following hypotheses testing. Manipulation check Facebook lead frame was manipulated in this experiment. Except for the first two items, other four items were chosen from the study of d’Haenens & Bosman’s (2003) to assess manipulation of the experimental conditions of the frames, regarding the frame groups each participant was randomly assigned to. In the following questions, participants were asked to answer, on a 5-point scale (1= strongly disagree to 5= strongly agree): (1) the Facebook lead contains a great amount of terminologies (2) the Facebook lead is easy to understand (reverse coded) (3) the Facebook lead refers to two sides or more than two sides of issue (4) the Facebook lead reflects disagreement between groups (5) the Facebook lead delves into the personal lives of the individuals it focuses on and (6) the Facebook lead emphasizes how individuals and groups are affected by the issue. Scores from item 1 and 2 were summed and divided by two to form the new index of “perceived technicality” (r= 0.75, p<.01). Item 3 and 4 were summed and divided by 2 to measure “perceived conflict”(r=0.4, p<.01). Lastly, item 5 and 6 were to summed and divided by 2 to measure “perceived human interest”(r=0.85, p<.01). One-way ANOVA assessed the success of the experimental manipulation. As shown in table 3, there was a main effect of technique frame on readers’ perceived technicality as determined by one-way ANOVA (F (2, 86)= 12.32, p=.00). A Turkey post- hoc test revealed that the scores of “perceived technique” was statistically significantly lower when readers were assigned to read.

(27) 24. the conflict framed lead (2.76± 0.60, p=0.00) and human interest framed lead (3.00± 0.68, p=.004) compared to technique framed lead group (3.55± 0.63). There were no statistically significant differences between the conflict and human interest framed lead groups (p=0.33). Besides, there was a main effect of conflict frame on readers’ perceived conflict as determined by one-way ANOVA (F (2, 86)= 7.76, p=.001). A Turkey post- hoc test revealed that the scores of “perceived conflict” was statistically significantly lower when readers were assigned to read the technique framed lead (3.04±0.68, p=0.012) and human interest framed lead (2.88± 0.85, p=.001) compared to conflict framed lead group (3.59± 0.60). There were no statistically significant differences between the technique and human interest framed lead groups (p=0.62). Lastly, there was also a main effect of human interest frame on reader’ perceived human interest as determined by one-way ANOVA (F (2, 86)= 8.44, p=.00). A Turkey post- hoc test revealed that the scores of “perceived human interest” was statistically significantly lower when readers were assigned to read the technique framed lead (2.28± 0.76, p=0.00) and conflict framed lead (2.60± 0.92, p=.031) compared to human interest framed lead group (3.20± 0.92). There were no statistically significant differences between the technique and conflict frame interest framed lead groups (p=0.32). Table 3 ANOVA: Perceived frame of GM news readers Type of Facebook Mean lead frame Perceived Perceived Technique Conflict Human interest Note. **p <.01. Technicality 3.55 3.05 2.28. Conflict 2.76 3.59 2.60. F Perceived Human interest 3.00 2.88 3.20. p. 12.315 .000** 7.76 .001** 8.439 .000**.

(28) 25. Hypothesis 1 assumed that readers who were exposed to conflict framed Facebook lead or human interest framed Facebook lead would have higher levels of intention to click on the accompanying news link compared with those who read the technical one. In table 4, as predicted, there was a statistically significant difference between groups as determined by one-way ANOVA (F (2, 86)= 5.09, p=0.008). A Turkey post- hoc test revealed that the intention to clcik was statistically significantly higher when readers were exposed to the conflict framed Facebook lead (3.18± 0.72) compared to technique framed (2.63± 0.62, p=.012) or human interest framed lead (2.69± 0.84, p=.032). However, there was no difference between the groups that were in the technique and human interest framed lead (p=. 954), so Hypothesis 1 was practically supported.. Table 4 ANOVA: Intention to click on GM news Mean Facebook framed lead. Technique Framed Facebook lead 2.63. Conflict framed Facebook lead 3.18. F. p. Human interest framed Facebook lead 2.69. 5.09. .008**. Note. **p <.01. Hypothesis 2 predicted that readers who were exposed to conflict framed Facebook lead or human interest framed Facebook lead would have higher level of news attention than those who were exposed to technique framed Facebook lead. Nevertheless, results showed there was no statistically significant difference between groups as determined by one-way ANOVA (F (2, 86)= .348, p=.707). Thus, H2: readers who are exposed to Facebook lead with the conflict frame or the human.

(29) 26. interest frame will pay more attention to the news, compared with those who are exposed to the technique frame, was not supported. Hypothesis 3 presumed readers who read the conflict or human interest framed lead would have higher level of understanding of news text compared with readers who read the technical framed lead. Results showed there was no statistically significance between groups as determined by one-way ANOVA (F (2, 86)= 1.821, p=.168). Accordingly, H3: readers who are exposed to Facebook lead with the conflict frame or the human interest frame will have higher news understanding, compared with those who are exposed to the technique frame, was not supported. Hypothesis 4 predicted readers’ judgment of the thematic thought of the news article would be affected by the framed treatment. More specially, readers who read the conflict framed lead would consider the conflicting element in the news article to be the main theme; readers who read the human interest framed Facebook lead would perceive the personal story to be the main theme and so did the technique frame. A Chi-square test of independence was performed to examine the relation between frame leads and readers’ thematic thought. Results in Table 5 showed Facebook lead frame did have an effect on reader’s thematic thought (χ2 (2, N=89) = 9.45, p <.01), so H4 was supported. Table 5 Results of Chi-square Test and Descriptive Statistics for readers’ thematic thought on GM news by Facebook lead frame(n=89) Perceived thematic thought Facebook lead frame TechniqueConflictHuman related related interest-related Technique frame 11 (12.4%) 21(23.6%) 0 Conflict frame 1(1.1%) 28(31.5%) 0 Human interest frame 5(5.6%) 23(25.8%) 0 2 Note.  = 9.45*, df = 2. Numbers in parentheses indicate column percentages. *p < .01.

(30) 27. Hypothesis 5 assumed that readers who read conflict of human interest framed lead would perceive the news to be more relevant to them, compared with those who read the technical framed Facebook lead. Results showed there was no statistically significance between groups as determined by one-way ANOVA (F (2, 86) = .408, p=.666). Thus, H5: readers who are exposed to Facebook lead with the conflict frame or the human interest frame will perceive the issue to be more related to themselves, compared with those who are exposed to technique frame, was not supported..

(31) 28. Study two: Nanotechnology news Participants A total of 90 participants were recruited online at a large public university in Taipei, Taiwan. Respondents were offered NT $100 for their participation, which took about 15 minutes. The mean age was 22.51 years (SD = 4.36), with 77.8 percent being female. The majority of the participants (36.7%) reported that they were graduate students. Among all the participants, only 5 people (5.6%) were from the science background. Participants had an average 4.46-year -Facebook using experience (S.D. = 1.37) and only one third of the total participant have subscribed to scientific Fan pages. Design As in the experiment 1, a one-factor between-subject factorial design was employed to measure the framing effect of Facebook lead on readers’ intention to click, news attention, news understanding, news theme and perceived relevance. Materials The second story released on February 16th, 2014 from Liberty times was about the potential threat of sunscreens with nanoparticles to the environment, creatures and human body and thus more related to general public’s life. Three respective Facebook lead were designed to meet the framed condition: technical, conflict and human interest. Each lead was similar in length and in terms of writing style. Each group contains the same core elements such as the core ingredient of the physical Nano particle sunscreen: ZnO2 and TiO2 along with different information exclusive to that particular frame. Technique framed lead focused on the introduction to the technique of Nano scale sunscreen itself, detailing of the spectrum of UVA and UVB and its special characteristics (230 Chinese words). Conflict framed lead included the.

(32) 29. different opinions toward the new product from the producer and medical practitioner (270 Chinese words). Lastly, framed human interest Facebook lead described the personal examples of two celebrities using Nano particle sunscreen, one foreign and one domestic (264 Chinese words). The full texts of three framed leads are listed in the appendix. Measures (1)Independent variables The means of the independent variables in each of the experimental conditions are reported in Table 6. There were no substantial differences between the means and standard deviations of all the independent variables between groups determined by one way-ANOVA. As the independent variables were asked before the manipulation, this shows that the random assignment indeed resulted in equal groups.. Table 6 Descriptive analysis of independent variables of Nanotech news Nano particle sunscreen news (n=90) Mean SD Gender Female (77.8%) _____ Age 22.51 4.36 Internet experience (year) 10.71 2.65 Facebook experience (year) 4.46 1.37 Subscription to science fan page Yes (33.3%) _____ Science promise 3.92 0.49 General media use 2.86 0.41 Familiarity with topic 2.53 0.75 Individual interest in the topic 3.03 0.81 (2) Dependent variables Intention to click. This variable is constructed as an index of four items based on the studies of Xu (2013) and Genova and Greenberg (1979) and was measured on 5point Likert scale (1=strongly disagree, 5= strongly agree), asking (1) I would like to click on the news feed to access to the full story (2) the Facebook lead is very interesting (3) I am very into the Facebook lead. Scores from four questions were.

(33) 30. summed and averaged to form the index of intention to click (M = 2.67, SD =0.76, Cronbach’s α =0.90). News attention. News attention is measured by three questions (1= strongly disagree, 5= strongly agree) based on the studies of Wang (2006) and Chang (2012) asking (1) I am very focused on the news (2) the news grabs my attention and (3) I am very into the news. Scores from these three questions were summed and averaged to form the index of news attention (M = 2.90, SD =0.79, Cronbach’s α =0.89). News Understanding. News understanding is constructed as an index of two fivepoint likert items (1= strongly disagree, 5= strongly agree) proposed by Chang (2009), asking (1) I understand the news text and (2) the news is easy to read. Scores were summed and averaged to form the index of news understanding (M = 3.07, SD =0.33 correlation=0.603, p=0.00). Thematic thought. This is a multiple choice question that asked participants to choose the most important theme of the news. Questions are designed based on the frame treatment. In the Genetic modified news, there are 3 choices: (1) Working mechanism and ingredients of sunscreen (2) The conflict of nanotechnology sunscreen (3) Hugh Jackman reveals skin cancer Perceived relevance. This experiment measured perceived relevance through three items adopted by Chang (2009) on the five-point scale ranging from (1= strongly disagree, 5= t strongly agree) asking (1) the news is relevant to me (2) the news is important to me and (3) the news is important to society. Scores were summed and averaged to form the index of perceived relevance (M = 3.36, SD =0.80, Cronbach’s α = 0.85)..

(34) 31. Table 7 Descriptive statistics of dependent variables in Nanotech news experimental condition Nanotechnology news (N=90) Mean SD Cronbach’s α /correlation Intention to click 2.67 0.76 α =0.90 (1) I would like to click on the news feed to access to the full story (2) The Facebook lead is very interesting (3) I am very into the Facebook lead. News attention 2.90 0.79 α =0.89 (1) I am very focused on the news (2) The pieces of news grabs my attention (3) I am very concentrated on the news News understanding 3.16 0.69 r=0.57** (1) I understand the news text p=0.00 (2) The news is easy to read (reverse coded) Thematic thought ---(1) Working mechanism and ingredients of sunscreen (2) The conflict of nanotechnology sunscreen (3) Hugh Jackman reveals skin cancer Perceived relevance 3.36 0.80 α =0.85 (1) The news is relevant to me (2) The news is important to me (3) The news is important to society Note. **Correlation is significant at 0.01 level. Procedure The procedure of study two was exactly the same as study 1, and 90 participants were randomly assigned to the following three groups: (1) nanotechnology news with technical Facebook lead (n=28) (2) nanotechnology news with conflict Facebook lead (n=32) and (3) nanotechnology news with human interest Facebook lead (n=30). The means of the independent variables in each of the experimental conditions are reported in Table 6. There were no substantial differences between the means and standard deviations of all the independent variables between groups. As the independent variables were asked before the manipulation, this showed that the random assignment indeed resulted in equal groups..

(35) 32. Results One-way analysis of variance tests for homogeneity revealed no significant differences between participants in scientific promise, media use, and familiarity with the topic and individual interest in the topic. One-way ANOVA is used in the following hypotheses testing. Manipulation check Facebook lead frame was manipulated in this experiment. Except for the first two items, other four items were adapted from the study of d’Haenens & Bosman’s (2003) to assess manipulation of the experimental conditions of the frames. In the following questions, participants were asked to answer, on a 5-point scale (1= strongly disagree and 5= strongly agree): (1) the Facebook lead contains a great amount of terminologies (2) the Facebook lead is easy to understand (reverse coded) (3) the Facebook lead refers to two sides or more than two sides of issue (4) the Facebook lead reflects disagreement between groups (5) the Facebook lead delves into the personal lives of the individuals it focuses on and (6) the Facebook lead emphasizes how individuals and groups are affected by the issue. Scores from item 1 and 2 were summed and averaged to form a new index of “perceived technicality” (r= 0.80, p<.01). Item 3 and 4 were summed and averaged to measure “perceived conflict”(r=0.50, p<.01). Lastly, item 5 and 6 were to summed and averaged to measure “perceived human interest”(r=0.85, p<.01). One-way ANOVA was used to assess the successfulness of the experimental manipulation. The comparison of the different experimental groups with respect to these six questions showed that the manipulation of this study was successful. As shown in table 8, there was a main effect of technique frame on readers’ perceived technicality (F (2, 87)= 37.38, p=.00). A Turkey post- hoc test revealed that the scores of “perceived technicality” was.

(36) 33. statistically significantly lower when readers were assigned to read the conflict framed lead (2.83± 0.62, p=0.00) and human interest framed lead (2.52± 0.50, p=.00) compared to technique framed lead group (3.55± 0.63). There were no statistically significant differences between the conflict and human interest framed lead groups (p=0.1). Besides, there was also a main effect of conflict frame on readers’ perceived conflict as determined by one-way ANOVA (F (2, 87) = 4.57, p=.013). A Turkey post- hoc test revealed that the scores of “perceived conflict” was statistically significantly lower when readers were assigned to read the technique framed lead (2.70±0.77, p=0.024) and human interest framed lead (2.73± 0.61, p=.036) compared to conflict framed lead group (3.20± 0.73). There were no statistically significant differences between the technique and human interest framed lead groups (p=0.98).. Table 8 ANOVA: Perceived frame of Nanotech news readers Type of Facebook Mean frame lead Perceived Perceived Perceived. F. p. Technicality Technique 3.80 Conflict 2.70 Human interest 2.02 Note. *p <.05, **p <.01. 37.38 4.57 24.60. .00** .013* .00**. Conflict 2.83 3.19 2.16. Human interest 2.52 2.73 3.26. Lastly, a main effect of human interest frame on readers’ perceived human interest was found as determined by one-way ANOVA (F (2, 87)= 24.60, p=.00). A Turkey post- hoc test revealed that the scores of “perceived human interest” was statistically significantly lower when readers were assigned to read the technique framed lead (2.02± 0.86, p=0.00) and conflict framed lead (2.16± 0.65, p=.00) compared to human interest framed lead group (3.26± 0.75). There were no statistically significant.

(37) 34. differences between the technique and conflict frame interest framed lead groups (p=0.76). Hypothesis 1 assumed that readers who were exposed to conflict framed Facebook lead or human interest framed Facebook lead were be more likely to click on the accompanying news link, compared with those who read the technical Facebook lead. In table 9, as predicted, there was a statistically significant difference between groups as determined by one-way ANOVA (F (2, 87)= 4.19, p=0.018). A Turkey post- hoc test revealed that the intention to click was statistically significantly higher when readers were exposed to the conflict framed Facebook lead (2.83± 0.82, p=.027) or human interest framed lead (2.80± 0.66, p=0.046) compared to technique framed (2.33± 0.69). There was no difference between the groups that were in the conflict and human interest framed lead (p=. 983), so Hypothesis 1 was supported.. Table 9 ANOVA: Intention to click on Nanotech news Mean Facebook framed lead. Technique Framed Facebook lead 2.33. Conflict framed Facebook lead 2.83. F. p. Human interest framed Facebook lead 2.80. 4.194. .018*. Note. *p <.05. Hypothesis 2 predicted that readers who were exposed to conflict framed Facebook lead or human interest framed Facebook lead would have higher level of news attention than those who were exposed to technique framed Facebook lead. Nevertheless, results showed there was no statistically significant difference between groups as determined by one-way ANOVA (F (2, 87)= 1.524, p=.224). Thus, H2: readers who are exposed to Facebook lead with the conflict frame or the human.

(38) 35. interest frame will pay more attention to the news, compared with those who are exposed to the technique frame, was not supported. Hypothesis 3 presumed readers who read the conflict or human interest framed lead would have higher level of understanding of news text compared with readers of read the technical framed lead. Results showed there was no statistically significance between groups as determined by one-way ANOVA (F (2, 87) = 1.346, p=.266). Accordingly, H3: readers who are exposed to Facebook lead with the conflict frame or the human interest frame will have higher news understanding, compared with those who are exposed to the technique frame, was not supported. Hypothesis 4 predicted readers’ judgment of the thematic thought of the news article would be affected by the framed treatment. More specially, readers who read the conflict framed lead would consider the conflicting element in the news article to be the main theme; readers who read the human interest framed Facebook lead would perceive the personal story to be the main theme and so does the technique frame. Results showed frame treatment does not have an effect on reader’s thematic thoughts (χ2 (4, N=90) = 2.936, p = .569), so H4 was not supported. Hypothesis 5 assumed that readers who read conflict of human interest framed lead would perceive the news to be more relevant to them, compared with those who read the technical framed Facebook lead. Results showed there was no statistically significance between groups as determined by one-way ANOVA (F (2, 87) = 1.442, p=.242). Thus, H5: readers who are exposed to Facebook lead with the conflict frame or the human interest frame will perceive the issue to be more related to themselves, compared with those who are exposed to technique frame, was not supported..

(39) 36. Discussion This paper aims to investigate the potential framing effects of Facebook lead on science news readers by employing the two most prominent soft news frames: conflict and human interest. Because this study is an explanatory research, it included as many variables as possible, five variables were tested: intention to click, news attention, news understanding, perceived relevance, and thematic thoughts. There were two major outcomes. Firstly, in accordance with the hypothesis, the primary variable in this study- intention to click on the news reaffirmed the importance of arousing reader’s situational interest when presenting news contents. More specially, conflict framed Facebook lead, in both studies, have the highest mean compared to human interest and technique framed groups. Besides, in the study 1, only conflict frame successfully enhanced reader’s situational interest to click on the news link for accessing full news, but no substantially difference between technique and human interest frame were found. While in the study 2, conflict frame and human interest frame achieved the same result. To delve into the difference between the two studies in terms of intention to click, one possible explanation lied in the text of human interest frame. In order to make the stimulus material the closest to the real situation, human interest framed Facebook lead in the study 1 adopted the real personal story of an English woman- Helen, a strong advocate of the 3-person-IVF Act and a beloved mother of a rare genetic dysfunctional child. Her story was widely covered in the Internet when she spoke out loud her hope to legalize the Act and stood out for others who have experienced the same situation. Notwithstanding, the emotional climax of the English mother did not raise the interest of readers in this study. Dor (2003) in his deliberate study of.

(40) 37. headline as relevance optimizer, pointed out readers would probably perceive a local news as more relevant to their lives than foreign news because the cognitive effort to process a piece of foreign news is larger, thus they may skip the foreign news in most of the cases, only if they have some subjective interest in the country, story topic or when the impact of the foreign story could be glocalized. This could help explain why participant in the framed human interest lead group of Genetic modification news didn't feel attached to the British mother Helen in the story, especially when all the participants are all young and healthy students, rare disease may not be of their interest and it is also not directly related to their lives. And in the study 2, reader’s intention to click between framed human interest Facebook lead and conflict framed Facebook lead only displayed slight difference, and were significantly higher compared to the plain technique framed Facebook lead. Unlike the content of human interest frame in the study 1, study 2 included the personal stories of one foreign and one domestic figures- Australian Hugh Jackman and Taiwan “Small S”. Even though Hugh Jackman is a foreigner, he is a worldwide well-known figure in view of his active movie performance, together with “Small S”, a local celebrity who appeared extremely frequent in the Taiwanese media, the human interest framed Facebook lead displayed certain degree of attractiveness to readers despite the fact that readers may have subjective varying attitude towards the two celebrities. These findings not only reaffirmed that soft news value can be applied to science journalism as well (Hanson, 1994; León, 2008), they were also in line with the past studies that conflict frame and human interest frame are the most two effective frame in enticing reader’s interest (Price, et al 1997; Semetko & Valkenburg, 2000). The second primary outcome of this paper was the framing effect of Facebook lead on reader’s thematic thought, regardless of the fact that the result was only.

(41) 38. significant in genetic modification news. In the first study, result was as predicted that frame can influence how people perceive the news theme. One third of the respondents who read technique framed lead thought the technical aspectintroduction to the mitochondria illness and the 3- person- IVF technique was the main point of the whole news story. Only one person in the conflict framed lead group didn't consider the conflict of the 3- person- IVF new technique to be controversial. And 17.5 percent of participants in the human interest framed Facebook lead considered the personal story of the English woman- Helen to be the main point, despite the fact that not a single word about Helen’s story was mentioned in the actual news story. These findings were consistent with a lot of extant studies conducted by Iyengar (1987) and Price et al (1997), which proved that how journalist select and present an issue have effect on reader’s issue- specific judgment, and had already been discussed in the literature review. On the other hand, in the study 2, reader’s perception of the news theme of Nano scale sunscreen did not reach the same result. At least one third of the respondents in each group reported conflict-related thoughtthe conflict of the nanotechnology sunscreen as the main point of the piece of news article. Result from study 2 was similar to the finding of Condit, et al (2001) on a study of the framing effect of headline on reader’s genetic determinism. Contrary to what they expect, readers who were supplied with a highly deterministic headline in genetic issue did not change their interpretation of a less deterministic news content, because readers had the ability to discount headline. Another interesting discovery from study 2 is that there was one participant in the human interest framed group actually select the answer “Hugh Jackman reveals skin cancer” when being asked to choose the news theme of Nano scale sunscreen news. Even though the case was extremely rare, evidence indicated there is still chance that readers only rely on.

(42) 39. heuristic cue to interpret news. To put it briefly, when considering the results of thematic thought from the both studies together, we can tell human interest frame only serve to arouse reader’s intention to access to the whole story but doesn't affect reader’s thematic thoughts; conflict frame not only can enhance reader’s intention to click but also reinforce their interpretation of the main text. Apart from these two primary findings, no relationships were found between frames and readers’ perceived news attention, news understanding, and relevance. The discussion would continue on from news attention. In spite of the frame manipulations and the fact that readers were told to read the news based on their personal usual news reading behavior on Facebook, all groups in either Genetic modification news or nanotech news showed similar levels of perceived attention, they are two possible explanations. Firstly, the failure of this variable may due to the fact that participants were aware of their experimental participation. Although this research used online field experiment and pursued the natural situation as far as possible, participants were carefully instructed step by step throughout the whole process, these instructions may somehow kept making participant remind themselves of being in an experiment and thus pay closer attention to the news article than usual. Secondly, news attention was measured based on participants’ subjective perception, how much attention they actually paid to the news article was not measured objectively and documented in this research. As a matter of fact, there are different ways to measure news attention. For example, Chen & Zilmann (2004) tested the effect of lead framing on selective exposure to print and Internet news by measuring the time each participant spent on the news, and found readers who read agony and conflict leads displayed longer reading time while misfortune and economic frames did not. Though their study did not focus on science news specifically, it provided.

(43) 40. another way to measure participants’ objective and actual news attention. Next, results of perceived news understanding from both studies, albeit they were both as the hypothesis predicted- participants in soft news frame groups: conflict frame and human interest frame reported higher levels of perceived understanding than technique framed groups, the results were not significant. But this result was not surprising; scholars have indicated attention arousal as the perquisite to learning (Graber, 2001; Baum, 2002). Again, news understanding was only measured by participants’ subjectively perceived understanding instead of answering actual questions related to the news content, to what extent they actually understand the news content was no unknown. Lastly, it is surprising that soft news frame- either conflict nor human interest frame make reader feel they are more relevant to the issues discussed in the news. To examine the result in more detail, means in each framed group in Nanotechnology news were in generally higher than genetic modification group did. One possible reason is the complexness of the content. In the Genetic modification news, a new medical technique that only took place in the lab with complex working mechanism behind, which aim to serve as the antidote to rare genetic dysfunctional mitochondria disease were introduced. Because it is a technique to cure a specific rare disease, the information may only be of socially minority’s interest. On the contrary, the new product discussed in the Nanotechnology was sunscreen, a daily common product, which was more familiar to the general public. The other possible reason for the mean difference between the two pieces of news may due to the result of the proportion of female participants. In study 1, female participant accounted for 62.9 per cent of total participants, while the percentage of female participant in the study 2 even amounted.