GATEKEEPING JOURNALISTSâ€™ WEBLOGS: THE INFLUENCE OF THE MEDIA ORGANIZATIONS AND INDIVIDUAL FACTORS OVER U.S. JOURNALISTSâ€™ PERCEIVED AUTONOMY By HYERI CHOI A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA 2006
Copyright 2006 by Hyeri Choi
To my parents, In-Soon Choi and Young-Hee Seo.
iv ACKNOWLEDGMENTS My first thanks should go to my chair, Professor Melinda McAdams. She was not only a very knowledgeable scholar in journa lism but also a warmhearted instructor. Without her consistent faith in me, this docum ent could not have been completed. Also as a human being, her endless aspirations toward life were enough to challenge me. I would like to give my thanks to my committee members, Dr. Cory Armstrong and Dr. Johanna Cleary. Dr. Armstrong willingly shared her we alth of knowledge in methodology for me. She was the one who enriched this document . Whenever I was in deep frustration, Dr. Cleary was there. She sincerely understood my struggles as an inte rnational student and cared for me with her heart. I should give my thanks to the Kwanjeong Educational Foundation, who has supported me during my masterâ€™s study at the University of Florid a. I would continue challenging myself in a new academic setting and would thus love to become a proud student of the Kwanjeong.
v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii ABSTRACT....................................................................................................................... ix CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................4 Weblog......................................................................................................................... .4 Definition...............................................................................................................4 Elements................................................................................................................5 Differences............................................................................................................5 Journalistsâ€™ Weblog...............................................................................................6 Gatekeeping Theory......................................................................................................7 Historical Review..................................................................................................7 Dimensions............................................................................................................9 Individual-level......................................................................................................9 Organization level...............................................................................................11 Ownership and Blogsâ€™ Position...........................................................................12 Size and Satisfaction............................................................................................14 Ad-dependency....................................................................................................15 Editor...................................................................................................................16 Bureaucracy.........................................................................................................18 3 METHODOLOGY.....................................................................................................19 Sampling.....................................................................................................................19 Initial Data Description...............................................................................................21 Conceptualization and Operational Definitions..........................................................23 Weblog Autonomy as Dependent Variable.........................................................23 Independent Variables.........................................................................................30 Data Analysis..............................................................................................................35
vi 4 RESULTS...................................................................................................................36 Findings......................................................................................................................3 6 Additional Findings....................................................................................................43 5 DISCUSSION AND CONCLUSION........................................................................46 Review of the Present Study.......................................................................................46 Summary and Discussion of Results..........................................................................47 Limitations and Future Research................................................................................54 Conclusion..................................................................................................................55 APPENDIX A SURVEY QUESTIONNAIRE...............................................................57 LIST OF REFERENCES...................................................................................................64 BIOGRAPHICAL SKETCH.............................................................................................70
vii LIST OF TABLES Table page 1 Frequency for Blogger Journalistsâ€™ Career..............................................................22 2 Frequency for Blogger Journali stsâ€™ Beats (multiple check).....................................23 3 Reliability Test.........................................................................................................27 4 Correlation Matrix Variables...................................................................................28 5 Factor Analysis for Journalists â€™ Perceived Freedo m of Autonomy.........................29 6 Analysis of Variance for Complexity of Ownership................................................36 7 Mean Difference for Complexity of Ownership......................................................36 8 Analysis of Variance for Ownership of Weblogs....................................................37 9 Mean Difference for Ownership of Weblogs...........................................................37 10 Analysis of Variance for Size of Organization........................................................38 11 Mean Difference for Size of Organization...............................................................38 12 Regression Result Job, Ad-dependency, and Editors in Positive Autonomy...........39 13 Regression Result Job, Ad-dependency, and Editors in Negative Autonomy..............40 14 Multivariate Analysis of Covariate Bu reaucracy of Organization in Positive Autonomy.................................................................................................................41 15 Mean Difference for Bureaucracy of Organization in Positive Autonomy..............41 16 Multivariate Analysis of Covariate Bureaucracy of Organization in Negative Autonomy.................................................................................................................42 17 Mean Difference for Bureaucracy of Organization in Negative Autonomy............42 18 Frequency for Weblog Update.................................................................................43
viii 19 Frequency for Issues of Public Importance..............................................................44 20 Frequency for Interactio n between Weblog Readers...............................................44
ix Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Mast er of Arts in Mass Communication GATEKEEPING JOURNALISTSâ€™ WEBLOGS: THE INFLUENCE OF THE MEDIA ORGANIZATIONS AND INDIVIDUAL FACTORS OVER U.S. JOURNALISTSâ€™ PERCEIVED AUTONOMY By Hyeri Choi August 2006 Chair: Melinda McAdams Major Department: Journalism and Mass Communications This study explored if j-bloggers are infl uenced by seven factor s: complexity, blog ownership, size, job-satisfaction, ad-dependenc y, editor, and bureaucr acy. It starts from the question that even if, in general, weblogs are consid ered a symbol of freedom, jbloggers might not have autonomy. Seven hypotheses were delivered. Seven influencing factors in this st udy were guided by gatekeeping theory, following the assumption that gate keeping theory can also be applied to certain types of new media, including weblogs. This study proposed a Web-based survey me thod to be completed by j-bloggers. To test seven hypotheses, one-way analysis of variance (ANOVA), multiple regression, and multivariate analysis of covariate were performed. As a result, it was found out that ad-dep endency influence journalist blogger. It means journalistsâ€™ weblogs are not free from th e oversight of their organizations but are
x partially influenced by organizational interests. Therefore, these findings seem to indicate the journalistsâ€™ weblogs might function as alternative media in a way similar to citizensâ€™ weblogs.
1 CHAPTER 1 INTRODUCTION The rapid rate of adoption of the Internet indicates that it is one of the fastest growing innovations in huma n history (Rogers, 2003). Rogers asserts that in 2001, there were an estimated one hundred million Internet users over the age of fifteen in the U.S. Focusing particularly on weblogs, several researchers predict that weblogs will become an alternative form of media. Gillmor (2004), in his book We the Media , uses the term â€œgrassroots journalismâ€ with reference to weblogs and spends pages attempting to demonstrate the significance of weblogs a nd their potential as an alternative media format. He contends that the low cost of bl ogs coupled with their open source nature will aid in this Internet format becoming a substi tute for traditional media. â€œDecentralized informationâ€ ranging over location and time is the key to understanding this grassroots media. New technologies are ma king this possible due to the fact that ideal media are audience-centered rather than journalistcentered (Singer, 1998). For these reasons, Gillmor encourages former audiences to establish their own media and to join the media aggressively. There are even handbooks for beginning bloggers being published, including those directed toward blog gers and cyber-dissidents published by Reporters Without Borders , providing readers with suggestions fo r the fastest and easiest ways of building their own cyber homes. Hewitt (2005) also shares th e same view as Gillmor for weblogsâ€™ alternativity. He insists that the weblogâ€™s communicatory reformation is greatly benefited from broad availability of informati on and it has a very similar aspect with the
2 religious reform back in the medieval ages since religious refo rm was also greatly benefited from a new printing technology de veloped by Gutenberg. A few major presses have been trying to monopolize the inform ation, however, their dreams seem unreal as weblogs and other grassroots media are in their way. This characteristic of blogs as an al ternative media medium may not only be relevant to citizens but to journalists as well. In the media field, the blog is considered a viable alternative medium due to merits su ch as its low operating costs and real time update capabilities (Gillmor, 2004) . For these reasons, some me dia owners and publishers believe that blogs represent a fundamental attack on journalism (Gillmor, 2004). Some say blogs are the only media form that conveys â€œreal journalismâ€ fr ee of the censorship that often filters traditional media (Pain, 2005). Moreover, some me dia researchers and well-known bloggers assume blogs will eventu ally change the landscape of todayâ€™s journalism (Gillmor, 2004). Despite weblogsâ€™ potential as an alternative media, there is a dearth of published academic research concerning weblogs. To da te, most of the scholarly literature addressing blogs focuses on how they can be ut ilized in educational endeavors as a means of expression and for evaluating material pulled from Internet sources. Furthermore, research analyzing journa listsâ€™ weblogs is rare. While in general a weblog is considered a symbol of freedom, journalistsâ€™ blogs are still censored (Palser, 2003). Whoever has the desire to express their thoughts, whether they be objective or biased and opinionated, can present their viewpoi nt on oneâ€™s private online room: the blog. However, journalists are held to different standards, because their blogs may be censored by their editors (Palser, 2003). Furthermore, j ournalists belong to
3 media companies, so their blogs are also re gulated by the owners or editors of their publications (Palser, 2003). C onsidering these checks placed on journalistsâ€™ blogs, this study poses the research questions: (1) How much freedom of expression do journalists perceive themselves as having in the crea tion and running of their own weblogs? (2) Do journalists believe that their weblog s are censored by certain factors? One of the initial steps unde rtaken in this study is to identify what constitutes a weblog and who is using them, and to explore and elaborate on the fa ctors that influence journalistsâ€™ blogs. Specifically, this study will apply factors guided by gatekeeping theory. Following the assumption that gatekeepi ng theory can also be applied to certain types of new media, including weblogs, I will determine if general gatekeeping factors that influence traditional me dia may also affect weblogs. In addition, this study will examine how journalists perceive thei r freedom of expression when blogging.
4 CHAPTER 2 LITERATURE REVIEW Weblog Much research has proclaimed a rosy picture of new media. Jones (2002) announced that new media has many positive aspects as a communication tool. Due to advances in communication technologies coupled with the deregulation of the telecommunication industry, a rapidly incr easing number of people can now receive substantial amounts of information quickly a nd communicate with othe rs freely (Singer, 1998). These positive factors of new media are very different from those of traditional media (Jones, 2002). This transitional phase for media represents a healthy advance for democracy. Today, old media outlets are losing a udiences to a variety of new media sources such as cable television programs, In ternet web sites, a nd chat rooms (Jones, 2002). Definition Weblogs or â€œblogsâ€ are one of the newest media forms that have rapidly spread. The way in which scholars define a weblog and differentiate it from other Internet communications varies according to the particul ar frame of reference that underlines their approach. The simplest descrip tion of weblogs defines them as â€œonline journalsâ€ (Klotz, 2004). Winer (2003) offers a more technical definition, describing a weblog as â€œa hierarchy of text, images, media objects and data, arranged chronol ogically, that can be viewed in an HTML browserthe center of the hierarchy, in some sense, is a sequence of weblog â€˜postsâ€™ that forms the index of the weblog.â€ To further differentiate blogs
5 from other types of online content, it can be said that blogs inject a philosophical element, meaning that there is a purpose dr iving the postings made . Many insist that blogs should convey a unique, human voice that communicates a distinct point of view or a distinct set of opinions (Grossman, 2004). Elements Some of the common features of weblogs incl ude (1) titles that describe each post, (2) links that lead to other Web sites, (3) comments about the links , (4) a tool allowing readers to add comments regarding each post, and (5) content in the form of stories, pictures, and audio and video clips. In a handbook for bloggers (2005), a blog is illustrated as â€œa personal websiteâ€ (1) contai ning mostly news, (2) regularly updated, (3) in the form of a diary with most of the posts also arranged in categor ies, (4) set up using a specially-designed interactive tool, and (5) usually created and run by a single person, sometimes anonymously. Differences Weblogs are different not onl y from old media but also from any other new media. The explosion in blogging over 2003 has incr eased the impact that blogs have on traditional media (Carr, 2004), with bloggers referred to as a genui ne alternative to mainstream news outlets (Grossman, 2004). Wi ner (2003) believes that a true blog must allow the personalities of the writers to sh ine through. That is the essential element of weblog writing. Trammell (2004) describes webl ogs as having five differences from other web pages: (1) hyperlink as the ga teway, (2) hyperlink as specific forms of interaction, (3) reverse chronological order, (4 ) maintaining changes or additions to the blog in the archive, and (5) communities that surround them. The â€œblogosphere,â€ the label applied to the online universe of interconnected weblogs (McFedries, 2003), has
6 expanded at an amazing rate, growing from th e twenty-three known blogs in existence in 1999 (Blood, 2003) to more than four million blogs today (Perseus, 2003). The proliferation of free blogging technology a llowing even the most technologically unsophisticated users to post their thoughts inst antly to the Internet has significantly helped fuel this weblog boom. Journalistsâ€™ Weblog Being a new type of media, journalistsâ€™ weblogs, there is only a small amount of research on journalistsâ€™ weblogs to be found, but that sm all amount is rapidly increasing. Some of the research done on journalistsâ€™ weblogs was to understand why journalists began weblogs and their relationship to tradit ional media. Journalistsâ€™ started writing weblogs in order to provide daily news of traditional media to their website (Robinson, 2006). According to Wall (2004), by 2004, many of journalistsâ€™ weblogs were ones hosted by traditional media by journalists th emselves and these numbers reach over one hundred. Discussions about jour nalistsâ€™ weblogs often descri be it as a function of a new channel to break the norms and outlets that traditional media dominantly have had. Robinson (2006) insisted that to journalists, weblog writing is a means in which they may break away from traditional style of reporti ng and writing of news articles, agendas, and frame which has been dominantly controlled by traditional media. Armstrong and McAdams (2006) found that whether or not the weblog writer is a reporter is not an influencing factor upon the credibility of the weblog. Also, Singer (2005) examined political journalist bloggers â€™ objectivity in weblogs writing and found that they do express their own opinions in weblog writing. That is to say, journalist bloggers hold the gatekeeping role themselves in weblog writing.
7 Gatekeeping Theory Historical Review The first scholar who used the concept â€œgatekeeperâ€ was sociologist Kurt Lewin (1947) in his manuscript â€œFr ontiers in Group Dynamics: 2 Channels of Group Life; Social Planning and Action Researchâ€ in the Journal of Human Relations . A gatekeeper is a subject who observes that news flows in a channel, and a channel normally contains several gates that are controlled by gatek eepers (Lewin , 1947). Th e initial concept of gatekeeping was applied to a model of how food items pass through channels on their way to the table (Lewin, 1947). According to th is original model, food served on the table passes through several channels (buying a nd gardening channels), and each channel includes several sections along the way to delivering food to the home, such as the grocery store, transportation, planting, ha rvesting, and so forth (Lewin,1947). Throughout the process of passi ng through sections, some food items may be lost or wasted. Lewin denoted every channel or secti on of this delivery procedure as a â€œgate,â€ with each process within the channel or s ection being controlled by a â€œgatekeeper.â€ Lewin proposed that this model could be applic able not only to food channel but to other general groups by regarding the channel, section, and ga te concepts as physical structures. Lewin was correct. Not long afte r his research, White (1950) applied Lewinâ€™s research to the field of journalism. White (1950) pioneered the first gatek eeping theory in journalism and mass communication research by applying the idea of individuals acting as gatekeepers. White named the editor â€œMr. Gates.â€ Mr. Gates was asked to give the background reasons why he had selected certain examples of wire news and also why he threw out unused ones. According to White, Gatesâ€™ reasons for sele ction were very subj ective, including the
8 editorâ€™s personal evaluation of the stories as well as lack of space for them. In 1977, Snyder compensated for the weak point of White â€™s research, the fact that White analyzed only a single gatekeeper, by studying multiple gatekeepers. McNelly (1959), in his expanded model, identified multiple gates th at international news items pass through on their way to an audience. About thirty years later, McQuail and Windahl (1981) developed the gatekeeping model, which held that news sources send news items to the gatekeepers, and items selected by gatekeepers are then sent to the audience. A key aspect in the process of sending news items to the gatekeepers is the idea that some of them may be lost and changed; therefore, when selected items are sent on to the audience, their initial contents may be altered. Bass (1969) also approached gate keeping in terms of individual aspects; however, he differentiates individuals into two categories: news gatherers (writers, bureau chiefs , reporters, or city editors) and news processors (editors, copyreaders, and translators). In contrast to White, who concluded th at selection of news items is up to individualsâ€™ personal prefer ence, there were a couple of scholars who found other influencing factors, including media organizations and routin es. Gieber (1964) compared the news decision-making process to mechan ism, media to machine. Gieber concluded that size and time pressures are more crucial and decisive factors in gatekeeping. Epstein (1974) and Dimmick (1974) found that the ideo logy of media owners, media routine, and official sources affect the selection of news items. In their model of the mass communication process, Westley and McLean (1957) proposed the idea that organization act as gatekeepers, not individuals.
9 Dimensions Throughout the history of gatekeeping, scholar s have approached characteristics of gatekeepers in several levels of analysis. Compared to Gieber (1964) who studied only the individual-level influen ce, Schramm (1963) approached the analysis of gatekeepers on three levels: individual, rou tine, and organizational levels . Based on previous studies, gatekeeping was further elaborat ed into a five-level hierarch ical model of influences on content, including individual, media routine, organizational, extramedia, and ideological levels (Shoemaker and Reese, 1996). This study will focus on the individual and organization levels. Individual-level Individual-level analysis was the focus of the earliest version of gatekeeping studies. Journalists are inclined to avoid the issue of censors hip yet are likely to forget about the connection between their personal values and their behaviors (Shoemaker and Reese, 1996). Shoemaker and Reese (1996) assume that j ournalists will have more autonomy to release their work if they have fewer restra ints, based on their pers onal beliefs and values constructed from their personal backgrounds and experiences. The general public often forgets about the fact that journalists as individuals have persona l values, beliefs, and attitudes influencing their work just as ordinary people do (1996). As Breed (1955) noted, journalists tend to adjust th eir professional views in th e process of socialization. When examining individual gatekeepers, sc holars have studied theories of thinking, decision-making, and characteristics of th e individual gatekeeperâ€™s personality, background, and values (Shoemaker, 1991). Fo r instance, Hews and Grahamâ€™s (1989) model of cognitive process, Second Guessing, in which an individual attempts to correct
10 a message by introducing prior knowledge to re interpret the manifest message, is one of the applicable individual-le vel analyses of gatekeepi ng. Gans (1979) has identified several social values for guiding the production of news such as progressivism, ethnocentrism, altruistic democracy, responsib le capitalism, small-town pastoralism, individualism, moderatism, social order, and national leadership (Shoemaker 1991). Although there are a series of individual-related theories, models, and concepts, this study will focus on Weaver and Wilhoitâ€™s ( 1986) three journalistic role concepts: The interpretive function, the dissemi nation function, and the advers ary function in order to keep track of autonomy, which is the purpose of this study. According to Weaver and Wilhoit (1986), the interpre tive function, which involves investigating official claims, analyzing complex problems, and discussing national policy, is the dominant professional role of mode rn U.S. journalists. They examined that more than 60% of surveyed journalists scor ed very high on the interpretive scale. Weaver and Wilhoit (1986) explained th e dissemination function, comprised of getting information to the public quickly while concentrating on the widest audience, is also very important, with more than half of U.S. journalists scoring very high on this scale. They examined that about a thir d of journalists scored high on both the investigation and dissemination functions. Last, Weaver and Wilhoitâ€™s (1986) de scribed the adversary function, which involves journalists acting as adversaries of officials or businesses, plays a relatively minor role. They examined that only 17% of U.S. journalists scored very high on the adversary scale.
11 Organization level Not all decisions journalists make ar e done from the individual perspective (Shoemaker and Reese, 1996). The power of hiring or firing does not belong to the journalists or gatekeepers, but rather to media organizations that have the greatest power to shape the rules and change them (Ste wart and Cantor, 1982). Journalists, as employees, avoid fully stating any unorthodox va lues they may possess in order to keep their jobs (Zelizer, 2004). In this context, it can be assumed that successful journalistgatekeepers are the individuals who best sta nd for organizationsâ€™ interests because they do not want to lose their jobs (Shoemaker, 1991). It is for th is reason that this study of gatekeeping has not been limited to individual-level analysis. Media organizations have the power to control individual gatekeepers. Many scholars have determined that certain factor s result in organizational levels having the greatest gatekeeping power. Some have proven that a media publisher possesses the single most decisive power in news ma king (Donohew et al, 1972). The less the bureaucratic structure, the more decision-ma king power journalists seem to have with regards to what they do or do not write (T uchman, 1972). Larger organizations seem to more often apply their own rules and policies innewsrooms than do smaller ones (Gieber, 1960). Adams (1980) explains gatekeeping by me ntioning a concept of organizational boundary roles as applied to the AP wire se rvice. Adams identifie s gatekeepers in the wire service system as the persons who filter inputs and outputs. Adams stated that criteria established by media organizations affect the gatekeeper , a boundary-role person, who selects or rejects news items. Gatekeep ing occurs in the mo ment when a boundary role person decides which news items are eith er selected or rejected in terms of both
12 fulfilling and avoiding transgression against th e criteria set forth by the organizations. Following this selection process, those chosen items may be manipulated, edited, rewritten, added or deleted, and so forth. A couple of scholars have addressed the c ontrolling power that media organizations have upon journalists, which ultimately contra dicts the concept of â€œdemocracy.â€ Breed (1955) verified that despite high expectati ons of democracy in the newsroom, daily newspapers have set policies established by the publisher that journalists are forced to follow. Schudson (1995) pointed out this cr ucial contradiction between democratic societies and journalis tsâ€™ autonomy. Schudson argued that ironically, in this liberal society, organizations constrain journalists in their decision-making power. Epstein (1974) addressed such organiza tional constraints more direct ly by maintaining that most news is determined by media organizational restraints rather th an through journalistsâ€™ autonomy. This organizational le vel power in media is not lo calized only in the United States. Some scholars have researched the broadcasting system in the United Kingdom and observed that organizational imperatives in television producti on threaten autonomy in the system (Bantz, 1985; Roshco, 1975). There are a number of fact ors that influence news c ontent. Influencing news content corresponds with influencing journa lists. Specifically, this study will measure journalist-influenced factors at the organi zational level, includ ing ownership, size, perceived ad-dependency, allegiance of th e editor, bureaucracy, and blog affiliating. Next, the study will address the individual level by assessing journalistsâ€™ job satisfaction. Ownership and Blogsâ€™ Position Types of media ownership are considered influencing factor s over journalists. Many researchers have demonstrated that ownership affects news content. Some
13 newspaper researchers have drawn the conclu sion that chain-owned newspapers tend to influence news content more than indivi dually owned papers do (Lacy, 1991; Picard, 1994; Shoemaker and Reese, 1996). Picard (1994) concluded that inst itutional owners are likely to influence news content more than individual owners due to their stronger profitseeking motivations. Bagdikian (2004) found th at when chains take over a newspaper, they tend to limit the hard news presen ted. A good illustration of this phenomenon is Gannettâ€™s takeover of The Louisville Courier-Journal , with overall news increasing, but stories became shorter and hard news decreasing (Coulson and Hansen, 1995). This tendency of institutions to put out so fter news likely extends to other types of media, including television and magazines. According to Williams (2002), media conglomerates increased the am ount of company-related news. ABCâ€™s Good Morning America spent two hours covering Disney Worldâ€™s 25th Anniversary, including an interview with CEO Michael Eisner (Kaufman 2000). Anchor Tom Brokaw repeatedly promoted his book, â€œThe Greatest Generationâ€ on NBC News shows, but neglected to mention that N BC owns nearly 25% of the bookâ€™s profits (Rosenwein 1999/2000). And the May 20, 1996 cover of Time featured a movie still from Time-Warnerâ€™s â€œTwisterâ€ for a science story on tornadoes, coinciding with the movieâ€™s release (p. 457). Shoemaker & Reese (1996) hypothesized that, â€œthe personal attitudes and values of news media owners may be reflected not only in editorials and columns but also in news and featuresâ€ (p. 267). Chain owners or publishers, who demand higher profit margins than do independent media sources, are likely to exercise control over journalistsâ€™ weblogs if they conclude that a certain weblogâ€™ story could negatively impact their
14 money-making potential. Thus, this study assu mes that media ownership is likely to influence the freedom of expression that jour nalists perceive themselves as having over their weblogs. Therefore, the following hypothesis is posited: H1: The complexity of the media owners hip will be negatively associated with journalistsâ€™ degree of pe rceived autonomy in managing their weblogs. From the organizational level, the studyâ€™s second hypothesis is drawn. If journalistsâ€™ weblogs are run by media companies that journalists bel ong to rather than by the journalists themselves, it is likely the organizational rule will be applied to weblogs too. In this situation, companie s will consider journalistsâ€™ weblogs as a part of their organization and if they conc lude that a certain weblogâ€™s story could negatively impact their profits, they will likely enact constraints on the journali stâ€™s weblog. Therefore, this study assumes that whether journalistsâ€™ we blogs are considered part of the media organizations will likely influence journalist sâ€™ perceived freedom of expression over their weblogs. Thus, the literature outlined a bove led to the following hypothesis: H2: Journalist-owned weblogs will be asso ciated with higher perceived autonomy as compared to company-owned weblogs. Size and Satisfaction The size of the media company is one of the factors influe ncing journalists. Specifically, research has proven that la rger companies have more influence on journalists than do smaller ones (Shoemaker and Reese, 1996). If the companies are larger, they need more advertisemen ts to support their companies and will be more greatly affected by major advertisers (Shoemaker and Reese, 1996). In order to attract advertisers, who pref er softer news content, media organizations may wield more influence over journalists (1996).
15 Much of the research addressing compa ny size has to do with journalistsâ€™ job satisfaction. In a survey of j ournalists at daily newspapers in the United States, it was found that journalists who belong to larger companies were le ss satisfied with their work than journalists at smaller companies (Johnstone, Slawski, and Bowman, 1976; Samuelson, 1962). From this literature, a third hypothesis is drawn. Larger media organizations are more likely to constrain journa listsâ€™ weblogs than are smalle r ones. Therefore, this study assumes that the size of a medi a organization is likely to in fluence journalistsâ€™ perceived freedom of expression over their weblogs. T hus, the research reported above led to the following hypothesis: H3: The size of the media organization will be negatively correlated with journalistsâ€™ degree of perceived autonomy in managing weblogs. Journalists at larger companies were likely to have less feelings of selfaccomplishment (Cook and Banks, 1993; Chan, Pan, and Lee, 2004). Journalistsâ€™ job satisfaction is representative of a combin ation of values, ideas , and the beliefs of journalists themselves (Bergen and Weav er, 1988). From this supposition, this study draws the conjecture that as journalistsâ€™ job satisfaction is lowered, their preference toward weblogs will grow. Thus, based on the literature review, this discussion leads to the following hypothesis: H4: The extent to which journalistsâ€™ perceived job satisfaction will be negatively correlated with journalistsâ€™ degree of perceived autonomy in managing weblogs. Ad-dependency Because advertising is the biggest mone y-making source in media, current media could not survive without adve rtisements. Consequently, adve rtisers indirectly influence
16 journalists, and the strength ening interconnection between media and capitalism becomes stronger. Investment and expenditures on media appear to be central to overall economic growth (McChesney, 2004). Advertising is an indispensable element in the media because it is the biggest source of profit for the industry, and it also acts as a cultural force (McChesney, 2004). The content of th e media presented has become linked with the interests of those w ho fund it (Altschull, 1995). Shoemaker and Reese (1996) noted this connection by comparing media organi zations to other profit-making business companies. Shoemaker and Reese stated that media companies, which are profit-making enterprises, try to make a product that can be sold for more than the costs of production. It is not surprising th at the power of advert isers in media becomes an impediment to journalistsâ€™ autonomy (Weaver and Wilhoit, 1996). According to Price (2003), only 9 out of 131 correspondents answered that they ha d never been pressured by advertisers. From this literature, an additional hypot hesis for this study is drawn. Advertisers are likely to constrain journa listsâ€™ weblogs if they concl ude that a certain weblog story could negatively affect thei r potential profits. Therefore, this study assumes that the degree of ad-dependency of media organiza tions is likely to influence journalistsâ€™ perceived freedom of expression over their weblogs. Therefore, the following hypothesis is posited: H5: The extent to which journalists percei ve ad-dependency of their organization will be negatively associated with journalistsâ€™ degree of perceived autonomy in managing weblogs. Editor The allegiance of editors is considered to be one of the influencing factors over journalists. Ken Sands, a reporter for Washingtonâ€™s Spokesman-Review , confesses that
17 editors often monitor reportersâ€™ weblogs to check stories out (Heyboer, 2004). Furthermore, Heyboer expects that as the num ber of blogger journalis ts grows, there will be increasing need to redefine the ed itorâ€™s role in blogging. As Heyboer (2004) maintained, newspaper workers have traditiona lly experienced a hier archical relationship with editors. That is, edit ors normally function as audien ces and arbiters of what a reporter does. As the media have faced d eclining audiences and increasing market competition, editors have attempted to increase productivity of their companies (Shoemaker and Reese, 1996). Underwood (1998) cited an example of an executive editor at the Gannett-owned Seattle Times who claimed that he spends 40% of his time monitoring newsroom budgets and coordinati ng their marketing role. In this modern media management environment, journalists mu st get approval for re leasing their stories by their editors, which in the end constrains journalistsâ€™ freedom (Shoemaker and Reese, 1996). Shoemaker and Reese (1996) hypothesized that â€œupper level media management personnel whose background is on the business side of the organization are more likely to make decisions based on economics rather than on professional cons iderationsâ€ (p. 267). From this literature, this study draws a suppos ition that editors w ho are strongly businessoriented are likely to constrain journalists. Mo reover, if the editors have strong loyalties to their companies, they are likely to c onsider the benefit of the company more and exercise gatekeeping over journalists more extensively. Thus, th e literature outlined above led to the following hypothesis: H6: The extent to which journalists believe that their editor pr ioritizes business interests will be negatively correlated with journalistsâ€™ de gree of perceived
18 autonomy in managing weblogs. Bureaucracy Bureaucracy is another variable that re gulates journalists. In more flexible companies, journalists may feel free to sp eak their minds and express their thoughts. Shoemaker (1991) stated that the gatekeeper model involves the idea of accepting physical limits. That is, given the number of stories and limited space available, the gatekeeper must make decisions to funnel a great number of news events down to a few. Shoemaker pointed out that organizations deve lop patterns and habits of gathering news and that the media should discover wa ys of gathering news effectively. From this literature, this study assume s that bureaucracy will be applied to journalistsâ€™ weblogs. The le vel of bureaucracy in an or ganization influences the flexibility of the newspaper and will thereby affect journalistsâ€™ weblogs. Journalists will be unlikely to post unfavorable comments a bout their companies on their weblogs. Thus, based on the literature review, this disc ussion leads to the following hypothesis: H7: The level of bureaucracy demonstrated in an organization will be negatively associated with journali stsâ€™ degree of pe rceived autonomy in managing weblogs.
19 CHAPTER 3 METHODOLOGY This thesis proposed a survey method in order to investig ate the hypotheses, specifically utilizing a Web-based survey to be completed by journalists who run weblogs. Babbie (2006) proposed that â€œsurvey research is probably the best method available to the social researcher who is interested in collecting original data for describing a population too large to observe directly, and surveys are also excellent vehicles for measuring attitudes and orientat ions in a large populationâ€ (p. 243). Thus, this current study anticipate d that a survey would be adequate for exploring the journalistsâ€™ use of weblogs considering the fact that the population was too large to observe directly. Specificall y, the current study employed a We b-based survey to collect the data because the topic of this paper is highly relevant to the Internet, and it enables quick and accurate gathering of survey information with minimal cost as compared to a traditional paper and pencil method. Furthermore, it is assumed that blogger journalists may favor a Web-based survey over a mailed survey. Sampling This study was based on two sampling methods, stratified random sampling and snowball sampling. The first step involved the stratified random sampling of 406 journalistsâ€™ blog addresses collected from Cyberjournalist.net, the news and resource Web site that focuses on â€œhow the Inte rnet, convergence and new technologies are changing the mediaâ€ (http://www.cyberjour nalist.net/news/000162, 2005. These initial journalist â€œj-bloggersâ€ were divided into four sub-categories based on the manner in
20 which their blogs are published: published by news sites (165 blogs), published by news sites-spot news and events (82 blogs), pub lished independently (98 blogs), and personal sites (61 blogs) as of October 24, 2005. Am ong these 406 blogs, 51 journalistsâ€™ blogs were discarded since some we re non-U.S-based blogs, had unid entified addresses, did not reveal bloggerâ€™s private e-mail addresses or had double addresses. Ul timately, a total of 355 blogger journalistsâ€™ e-mail addresses were obtained. Upon UF IRB approval, the survey link titled â€œJournalism Blog Survey â€ was e-mailed to blogger journalists (who posted e-mail addresses on their blogs) asking if the blogger would click to the Informed Consent leading to the Web-based survey. Th e following text was provided to the blogger and written for optimal results: I have worked as a reporter in Korea, and currently I am researching U.S. journalism blogs at the University of Fl orida with Professor Melinda McAdams. As a fellow journalist, I truly understand how busy you are. But please take a moment to complete this Web survey. Y our response would be very useful for this journalism blog research. You just click the link below and open a survey agreement form. When you click the agreem ent link, it will lead you directly to the survey. Your responses are completely confidential. This Web survey does no t collect any information about you and your computer. If you want to have the research results, pl ease e-mail me your request, and I will mail the results when this research is done. In order to increase the res ponse rate, the researcher firs t appealed to the potential participants as a fellow journalist rather th an emphasizing status as a graduate student with the intent of drawing th eir attention to the survey. S econd, the researcher stated every journalistsâ€™ full name excluding those th at were unidentified. Third, the e-mail was sent out on either a Tuesday or Wednesday. These approaches were aligned with Adams and Clearyâ€™s (2006) proposed techniques for increasing response rate in Web-based surveys, including sending personalized fo llow-up messages and choosing the most effective days to send emails Wednesda y, Tuesday, and Thursday, in order.
21 Additionally, a snowball sampling method wa s implemented by asking the original survey respondents to recommend other blogge r journalists. Snowba ll sampling is a nonprobability method whereby each person sampled is asked to suggest additional people for responding. Babbie (2006) proposed that â€œsno wball sampling is appropriate when the members of a special populati on are difficult to locateâ€ (p. 1 84). For this reason, at the end of the e-mail invitation, respondents were asked: â€œIf you know any other blogger journalists who might be interested in this survey, would you please forward this e-mail to them?â€ For this study, a Web-based survey was conducted as a following step. First, the participants in this research were directed to read the informed consent by clicking the link shown in the e-mail statement. The surv ey was conducted during the period of time from March 22, 2006, to April 5, 2006. The informed consent and actual survey link were open for 15 days. Follow-up messages were se nt out to 274 people excluding 81 people who already responded for one week. After one week, additional 27 responses were obtained. A total of 108 blogger j ournalists participated in the survey. All 108 responses were used. Initial Data Description Data for this study was collected on April 5, 2006, from journalists working in this field. A total of 108 subjects completed the survey. More than 77% (83 people) of the samp le were male, and 23% (25 people) were female. The ages of participants ranged from under 25 years through over 40 years: 7% (8 people) were under 25 years, 12% (13 people) were 25-30 year s, 17% (18 people) were 30-35 years, 10% (11 pe ople) were 35-40 years, 54% (58 people) were over 40 years.
22 Table 1. Frequency for Blogge r Journalistsâ€™ Career The duration of time that par ticipants have worked as journalists ranged from under 1 year to over 20 years: 41% (44 people) of th e sample answered that they had worked in journalism for more than 20 years. 19% ( 20 people) worked 10-15 years, 16% (17 people) worked 15-20 years, 13% (14 peopl e) worked 5-10 years, 10% (11 people) worked 1-5 years, and 2% (2 people) worked under 1 year. With regards to the amount of time the part icipants had been wo rking as journalists for their current organizations, response s ranged from under 1 year through over 20 years: 26% (28 people) of the sample answer ed that they had been working in their current organizations for 1-5 years. 22% ( 24 people) had been working for 5-10 years, 18% (19 people) for more than 20 years, 12% (13 people) for 15-20 years, 10% (11 people) for 10-15 years, and 9% (10 people) had been work ing as journalists for the organization for less than 1 year. Regarding beats covered by the journalis ts participating, in Table 2, 17% (18 people) of the sample answered that they cover city/metro, 16% (17 people) cover arts and entertainment, 15% (16 people) cover f eatures, 14% (15 people) cover sports, 13% Item Journalism career (%/ N) Current organization (%/ N) Less than 1 year 2% (2) 9% (10) 1, but less than 5 10% (11) 26% (28) 5, but less than 10 13% (14) 22% (24) 10, but less than 15 19% (20) 10% (11) 15, but less than 20 16% (17) 12% (13) More than 20 years 41% (44) 18% (19) No answer 0% (0) 3% (3) Total 100% (108) 100% (108)
23 (14 people) cover business, 12% (13 people) cover technol ogy, 7% (8 people) cover national, and 7% (8 people) co ver international issues. Multiple answers were allowed. Table 2. Frequency for Blogger Journalistsâ€™ Beats (multiple check) Item Frequency of Response (%/ N) City/ metro 17% (18) Sports 14% (15) Features 15% (16) Arts & entertainment 16% (17) Business 13% (14) National 7% (8) International 7% (8) Technology 12% (13) Other 40% (43) Conceptualization and Operational Definitions As defined by Babbie (2006), operational de finition is â€œhow a concept will be measuredâ€ (p. 125). This study conceptuali zes seven independent variables and one dependent variable to be operationally defi ned. The dependent vari able is journalistsâ€™ degree of perceived autonomy in managi ng weblogs. There are seven independent variables: (1) types of owne rship, (2) ownership of the we blogs, (3) size of the company, (4) journalistsâ€™ job satisfac tion, (5) perceived ad-dependenc y of the media organization, (6) allegiance of the editors, and (7 ) bureaucracy of the media organization. Weblog Autonomy as Dependent Variable There have been some research define autonomy in journalism. The concept of autonomy, in general, was defined which is based on notions of control over oneâ€™s own destiny (Lieber, 1994) and a main characterist ic of profession (Sc holl and Weischenberg, 1999). More specifically, workplace autonomy is the amount of freedom workers have to schedule their work and determine the procedur es to be used in carrying it out (Hackman and Oldham, 1976). Furthermore, â€œjob autonomy implies that the individual has some
24 objective control in the speci fic domain of task process and accomplishment.â€ Hodson (1991) has argued that such individual creativit y is central to work and that autonomy is an essential requirement for th e viability of an organization. Applying to journalism, Scholl and Weisch enberg (1999) described â€œautonomy is a concept, which means freedom to shape jour nalistsâ€™ work without being controlled by internal and external powers.â€ (p. 1). Furthermore, they mainta ined that journalists need a great amount of scope to do a good job because their autonomy is directly connected to freedom of the media. Journalistic autonom y has been also examined by Weaver and Wilhoit (1991). Palser (2003) st ated that journalists are he ld to different standards, because their blogs may be censored by their editors. Furthermore, journalists belong to media companies, so their blogs are also re gulated by the owners or editors of their publications (Palser, 2003). From these previous research, this study pursue to measure if journalists still have autonomy when it comes to weblogs. The dependent variable in this study, th e journalistsâ€™ degree of perceived autonomy in managing weblogs, is divided into two di mensions: positive and negative autonomy. Positive autonomy could be defined as freedom of expression whereas negative autonomy was defined prevention of writing what journalist wants. Thus, the questionnaire statements used fo r measuring the depende nt variable were grouped into two categories based on the result s of factor analysis: either positively or negatively worded. In order to prevent response biases in the survey and thereby strengthen the reliability of the measures (Horan, DiStefano, and Motl, 2003) this study used the strategy of addressing the same issue with some items positively and some negatively
25 worded. However, exploratory factor analys is did not confirm the existence of the intended underlying variables. The factor an alysis sorted the items in two separate factors: positively worded statements in one factor and negatively worded items in another. This is a well known effect of wording (Horan, DiStefano, and Motl, 2003; Marsh, 1996). The positive worded items seem to have in common encouragement for running weblog issue. On the other hand, th e negative worded items are describing suppression in the issue. A factor analysis wa s performed on the data to measure the journalistsâ€™ degree of perceive d autonomy in managing weblogs with the variables being measured on a seven-point Likert scale. As shown in Table 3, although the widely-accepted social science cut off is that alpha should be .70, the alpha level of .60 could be considered as lenient acceptable level in this study (Cronbach, 1951). Some scholar argued that the alpha level of .70 with standard deviations level of .55 could be accep table in social science research (Cronbach, 1951). Alpha level with under .60 is considered poor, alpha between 6 and 7 is moderate, alpha between 7 and 8 is good, and alpha ove r .80 is highly reliable (Cronbach, 1951). Nunnally (1978) states that a satisfactory level of reliability is dependent upon how the measure is used as well as in some cases a reliability value of .50 or .60 is acceptable. Kerlinger and Lee (2000) argue d that there is no evidence to support the cut-off alpha level of .70 arbitrary rule as well as a low reliability value may be acceptable if the measuring instrument has high validity (p. 662). Gronlund (1985) emphasized that researchers should go into deciding based on the used test and measuring instrument whether a reliability value is acceptable. Prakash a nd Lounsbury (1983) argued that current social scientists should examine empiri cally the quality of th eir measure in terms
26 of classical reliability concepts as well as inquiry into relate d measurement topics such as generalizability theory . Thus, in this study, ad-dependency close to .60 and alpha level of allegiance of editors is below .60. However, al l standard deviations of items used two factors (e.g., ad-dependency and allegiance of editors) were satis fied with .55. Thus, these variables could be include d in analysis because two fact ors play an important role in the relationship between these fa ctors and journalistsâ€™ autonomy. The internal consistency of the measur es was evaluated with Cronbachâ€™s alpha: positive autonomy was .74; negative autonomy was .37 ( r of two items: .23). In the negative autonomy, alpha value was .37 and Simple correlation was to estimate the relationship between the two items, r is .23. If the correlation coefficient value is on the range from 0 to .30, it can be said that that is weak relationship. A lthough there is a weak association of the rela tionship between two items, correlati on is significan t at .05 level. Allegiance of the editors was .43; in the dimension of ad-dependency, the statement â€œMy organization depends on advertisingâ€ from ad-dependency items was dropped from the analysis and the alpha of ad-dependency wa s increased to .69 after deleting this item. And in the dimension of jour nalistsâ€™ job satisfaction, the st atement â€œI enjoy adopting new technologyâ€ from journalistsâ€™ job satisfacti on was dropped from the analysis and the Cronbachâ€™s alpha journalistsâ€™ job satis faction was then increased to .69. The objective of factor analysis was to e xplore the dimensions underlying six of the autonomy variables. The data reduction obj ective requires grouping variables. Thus, Rtype factor analysis was a ppropriate. The original variables were: â€œencouragement,â€ â€œpositive reactions,â€ â€œnegative reactions ,â€ â€œmonitor,â€ â€œcensor,â€ and â€œopinion.â€
27 Thus, this study can be examined a corre lation matrix of the six variables to determine whether the data were ap propriate for a factor analysis. Table 3. Reliability Test Latent Variables Items Mean SD Cronbachâ€™s alpha 1) My company encourages journalists to run weblogs. 5.37 1.59 2) I have received positive reactions to my weblog from my supervisor, co-workers, and other people in the company. 5.36 1.65 Positive Autonomy 4) I believe my weblog is monitored by upper-level managers in the company. 3.97 1.89 .74 3) I have received negative reactions to my weblog from my supervisor, coworkers, and other pe ople in the company 5.71 1.50 Negative Autonomy 5) My weblog has been censored by upper-level managers in the company 1.87 1.44 .37 1) My editor has power in the organization* 5.38 1.34 2) My editor stresses that freedom of expression is important in journalism 5.27 1.60 3) My editor cares about bottom line* 4.94 1.51 4) My editor is open to the opinions of reporters 5.53 1.53 Allegiance of the Editors 5) My editor supports my work. 5.72 1.47 .43 1) I might be reprimanded if I write against advertisers. 2.65 1.76 2) My company cares about bottom line. 5.97 1.32 Ad-Dependency 3) Advertisers influe nce contents in my companyâ€™s news products. 2.26 1.45 .69 1) I feel loyal toward my current company. 5.29 1.37 3) I am satisfied with my current work. 5.11 1.39 4) I express my opinion freely in my work place. 5.52 1.49 journalistsâ€™ job satisfaction 5) I want to leave journalism* 2.53 1.73 .69 * Three items denote reversed-scores: the fo llowing items; My editor has power in the organization, My editor cares about bottom line, and I want to leave journalism. KMO is to measure the sampling adequac y. The value of KMO (.62) is high. KMO and Bartlett's test of sphericity ( X2=113.5, df =15 , p <.05), indicate that correlations
28 between pairs of variables can be explained by other variables. As shown the Bartlett's test of sphericity, significant result indicates matrix is not an identity matrix. We can say that the varimax is better because it has more ambiguous correlation figures to each factor. In Table 4, the correlation matrix revealed that 6 of the 10 correlations â€“ or 60 percent â€“ were significant at the .05 level. Although the Bartle tt test of sphericity shows that the overall correlations are significant at the .000 level, the measure of sampling adequacy revealed that the variables were poorly predicted without error by the other variables (KMO= .618). An examination of the KMOs for the individual variables also revealed that â€œencouragement,â€ â€œpositive reactions,â€ â€œnegative reactions,â€ â€œmonitor,â€ â€œcensor,â€ and â€œopinionâ€ had KMO values over .50. Table 4. Correlation Matrix Variables Correlations 1 2 3 4 5 1 Encou 1 .631** -.037 .399** -.036 2 Posi 1 .023 .453** .004 3 Nega 1 -.252** -.293** 4. Monit 1 .273** 5. Censo 1 **p<.05 Table 5 illustrates the results of a princi pal component factor analysis with the varimax rotated methods, displaying the following factors: â€œencouragement,â€ â€œpositive reactions,â€ â€œnegative reactions,â€ â€œmonitor,â€ â€œcensor,â€ and â€œopinion.â€ As demonstrated in Table 3, three variables, â€œencouragement, â€ â€œpositive reactions,â€ and â€œmonitor,â€ are highly correlated with Factor 1, while the va riables of â€œnegative reactionsâ€ and â€œcensorâ€ are highly correlated with Factor 2. Factor 1 could be considered positive autonomy and Factor 2 negative autonomy. One variable, â€œopinion,â€ was deleted based on its low communality of .21. Originally, â€œcensorâ€ item had negative loading within the same
29 factor, either the variable with the positive or the negative loadings must has its data value reversed (Hair et. al., 1995). Thus, â€œcenso râ€ original factor loading -.773 now have the reversed score of .773. Table 5. Factor Analysis for Journali stsâ€™ Perceived Freedom of Autonomy Unrotated Varimax Autonomy F1 F2 F1 F2 Communalities Encouragement .778 .372 .861 .043 .74 Positive reactions .795 .388 .883 .051 .78 Negative reactions -.308 .682 -.021 .748 .56 Monitor .775 -.110 .673 -.401 .61 Censor .296 .714 .003 .773 .60 Opinion -.228 .399 -.057 .456 .21 Eigenvalues 2.07 1.43 1.97 1.53 % of Variance 34.5 23.9 33.2 25.5 Cumulative Variance Explained 59 KMO=.618, B/S X2=113.5, df=15, p .001 In this study, the dependent variable is the journalistsâ€™ degree of perceived autonomy in managing weblogs. To measure this variable, six statements were presented in conjunction with a Likert scale with po ssible responses ranging from â€œ1,â€ very strongly disagree, to ,â€ very strongly agre e. The statements posed are as follows: (1) My organization encourages journalists to run weblogs, (2) I have received positive reactions (comments, etc.) to my weblog fr om my supervisor, co-workers, and other people in the organization, (3) I have received negative reacti ons (comments, etc.) to my weblog from my supervisor, co-workers, a nd other people in the organization, (4) I believe my weblog is monitored by upper-level managers in the organization, (5) My weblog has been censored by upper-level mana gers in the company, and (6) I freely express my opinion on my weblog.
30 Independent Variables All of the seven variables were derived by adaptation from previous literatures. The first variable, complexity of ownership, distinguishes between journalists as working for chains, conglomerate owners or individual, minor owners. Williams (2002) stated that media conglomerates increased the amount of company-related news quantity and news items. Consequently, this study assumes that ch ain owners are likely to exercise control over the contents of journalistsâ€™ weblogs for profit reasons. In or der to measure this variable, one question was deli vered: Which corporation ow ns your news organization? Twenty checking options were provided, ei ghteen of which were titles of specific organizations with an additional option for noting that the organization is independently owned or a final choice of â€œother.â€ The ei ghteen major conglomerates were taken from Bagdikian (2004). This study defined as eighte en conglomerates could be considered as major chain, independently owned considered as independent, and other considered as minor conglomerates. Thus, eighteen conglom erates were coded as 1, independently owned were coded as 2, and others were c oded as 3. Then these three groupsâ€™ mean differences were compared by using analysis of variance. The second variable, ownership of the we blogs, is defined as distinguishing if journalistsâ€™ weblogs are either journalistowned or organization-owned. The hypothesis related to this variable is drawn from the same assumption as was applied to the first variable. This study supposes that organizati on-owned weblogs can be considered as the chain ownerâ€™s phases compared to journali st-owned weblog as individually owned media. In order to measure this variable, in itially questions were delivered: (1) Does your company host your weblog? (2) Do you host your weblog separately from any employer? (3) Does your organization put journalistsâ€™ weblogs on the website of the organization?
31 However, after delivering frequencies of th ese three questions, two questions: (1) Does your company host your weblog? (3) Does your organization put jour nalistsâ€™ weblogs on the website of the organization? were not used because of their apparent contradictions to one question: (2) Do you host your weblog separately from any employer? Regarding the second question, which was included in this study, among 108 respondents, 33 people (30.6%) responded they host their weblog sepa rately from any employer and 75 people (69.4%) do not. In the contrary, 74 peopl e (68.5%) responded their organization host their weblogs (question 1) and 91 people (84.3%) responded their organization put journalistsâ€™ weblogs on the website of the or ganization (question 3) . Respondents could answer either yes which were coded as 1 or no which were coded as 0. Then to see these two groupsâ€™ mean differences analysis of variance were performed. The third variable assessed is size of the company. This is defined as the number of staffs in the newsroom. This idea is de livered from Shoemaker and Reeseâ€™s (1996) findings that larger companies have more infl uence on journalists than do smaller ones. From this supposition, this study proposes that larger size media organizations are more likely to control the contents of journalistsâ€™ weblogs than are smaller ones. In order to measure this variable, one question was as ked: How many journalists work in your newsroom? Five employee ranges were given as answering options: Less than 20, 20 to 60, 60 to 100, 100 to 140, and more than 140. This study defined as companies which have less than 20 to 60 journalists could be considered as small size company, 60 to 140 considered as medium size company, and more than 140 considered as large size company. Then, less than 20 and 20 to 60 were coded as 1, 60 to 100 and 100 to 140
32 were coded as 2, and more than 140 were coded as 3. Then these three groupsâ€™ mean differences were compared by using analysis of variance. The fourth variable, journalis tsâ€™ job satisfaction, is defi ned as journalistsâ€™ overall evaluation of their current wor k. This variable is highly rela ted to the third variable, size of the company. In previous research, it is found that journalists who belong to larger companies were less satisfied with their curre nt work as compared to those at smaller companies. From this assumption, this study su pposes that if journa listsâ€™ job satisfaction is decreased, they would demonstrate stronger motivation to run personal media, such as weblogs. In order to measure this variable, fi ve statements were pr esented in conjunction with a seven-point Likert scale with possibl e responses ranging from ,â€ very strongly disagree to ,â€ very strongly agree. The statements presen ted are as follows: (1) I feel loyal toward my current organization, (2 ) I enjoy adopting new technology, (3) I am satisfied with my current work, (4) I express my opinion freely in my work place, and (5) I want to leave journalism. The items used in this study could be appropriate to measure journalistsâ€™ job satisfaction. C onvergent validity which is one of the common methods of assessing the validity of a scale is to see how well it correlated with alternative measures of the same attitude (Campbe ll and Fiske, 1959). All of thes e items have a tendency to agree with items that express job satisfacti on. Thus, the relationshi p between these items could have reflected common sources. The fifth variable, perceived ad-dependenc y of the media organizations is defined as the journalistsâ€™ percepti on of the influence of adve rtising over their current organizations. Altschull (1995) stated that the content of media has become linked with the interests of those who fund it. Based on this assumption, this study proposed that
33 advertisers are likely to control journalists â€™ weblogs with organizations determining if weblogsâ€™ contents may negatively affect agains t their profits. In order to measure this variable, four statements were presented in conjunction with a seve n-point Likert scale with possible responses ranging from ,â€ ve ry strongly disagree to ,â€ very strongly agree. The statements presented are as fo llows: (1) I might be reprimanded if I write against advertisers, (2) my organization cares about the bottom line, (3) advertisers influence contents in my organizationâ€™s news products, and (4) my organization depends on advertising. The items used in this st udy could be appropriate to measure addependency. All of these items have a tende ncy to agree with ite ms that express the degree of ad-dependency of organizations . Thus, these statements have a fitting translation of the construct perceive d ad-dependency of organizations. The sixth variable, allegiance of the editors, is defined as the j ournalistsâ€™ perception of their editorsâ€™ degree of loyal toward their news organizations. Shoemaker & Reese (1996) stated that â€œupper level media mana gement personnel whose background is on the business side of the organization are more likely to make decisions based on economics rather than on professional considerations â€ (p. 267). Drawing on this assumption, this study proposed that editors may consider the be nefit of their organizations more and as a result will exert control over j ournalistsâ€™ weblogs. In order to measure this variable, five statements are delivered in conjunction with a seven-point Likert scale with possible responses ranging from ,â€ very strongly disa gree to ,â€ very strongly agree. The statements presented are as follows: (1) My editor has power in the organization, (2) my editor stresses that freedom of expression is important in journalism, (3) my editor cares about the bottom line, (4) my editor is open to the opinions of reporters, and (5) my editor
34 supports my work. The items used in this st udy could be appropriate to measure editorâ€™s business interests. All of these items have a te ndency to agree with items that express the degree of editorsâ€™ busin ess interests. Thus, the measurem ent tools have higher convergent validity. The last variable is bureaucracy of the media organization, which is defined as the number of layers of approval that an articl e must pass through prior to being published by a news organization. In gatekeep ing, a gatekeeper acts as a fu nnel selecting news content. From this finding, this study assumed that jour nalists who have less layers of approval in organizations experience lower levels of pe rceived freedom in running their weblogs. In order to measure this variable, one questi on was delivered: On average, how many people must approve your stories before they are publ ished in the news organization or broadcast in your news program? Possible answer range s included: none, 1, 2, 3, and more than 3 people. This study grouped range s as two: existing layers of approval and do no existing. Thus, none were coded as 0 and 1, 2, 3, and more than 3 were coded as 1. In addition, this study was delivered idea that journalistsâ€™ ty pes of employment and size of organization might work as control variable s. In order to measure journa listsâ€™ types of employment, one question was delivered: I am employed by or ganizations either full time, part time, free lancer, or not employed by any organization. Response to full time were coded 1, part time were coded 2, free lancer were coded 3, and I am not employed by any organization were coded 4. Size of organization variable, an other control variable, was already coded in H2: Less than 20 and 20 to 60 were coded as 1, 60 to 100 and 100 to 140 were coded as 2, and more than 140 were coded as 3. Thus, to see group differences between existing layers of approval and non existing controlling the influence of
35 covariates such as size of organization and type of employment, multivariate analysis of covariate was performed. Data Analysis The Statistical Package for the Social Sciences (SPSS 14 for Windows) was utilized in this study for stat istical computer analysis. To test H1, H2, and H3, One-Way Analysis of Variance (ANOVA) was used in th e analysis of mean differences for several groups. A One-Way Analysis of Variance ( ANOVA) was uses variances to test the equality of three or more means at one tim e, meaning that ANOVA can detect significant differences between means (Agresti and Finla y, 1997). To test H4, H5 and H6, a multiple regression analysis was used to find the relationship between job satisfaction, addependency, and allegiance of editors and depe ndent variable. In H7, size of organization and type of employment were treated as the co ntrolled variables. Also the covariates are assumed to be linearly related to the depende nt variables (Hair et al., 1995). Thus, to adjust the influence of covariates such as size of organization a nd type of employment, multivariate analysis of covariate was performed.
36 CHAPTER 4 RESULTS Findings Tables 6 and 7 illustrate the results of one-way analysis of variance, showing means, standard deviations, and analysis of variance results. Table 6. Analysis of Variance for Complexity of Ownership Positive Autonomy Level M S.D N Individual 4.24 1.64 17 Chain 5.08 1.20 40 Minor 4.98 1.40 51 Complexity of ownership Total 4.90 1.39 108 Negative Autonomy Level M S.D N Individual 2.06 1.32 17 Chain 2.19 1.31 40 Minor 2.12 1.09 51 Complexity of ownership Total 2.13 1.20 108 Table 7. Mean Difference for Complexity of Ownership Positive Autonomy df S.S M.S F Sig. Between Groups 2 9.19 4.60 2.45 .09 Within Groups 105 196.87 1.88 Total 107 206.06 Negative Autonomy df S.S M.S F Sig. Between Groups 2 .22 .11 .08 .93 Within Groups 105 154.58 1.47 Total 107 154.80 H1 predicted that the complexity of the media ownership will be negatively associated with journalistsâ€™ degree of perceived autonomy in managing weblogs. In the dimension of positive autonomy, as shown in Table 6, individual ownership group had a mean of 4.24 and standard deviation of 1.64, chain owners hip group had a mean of 5.08 and standard deviation of 1.20, and minor co mplexity ownership group had a mean of
37 4.98 and standard deviation of 1.40. In Table 7, th e analysis of variance for complexity of the ownership revealed that there is a marg inally significant difference between the three groups based on levels of autonomy ( F = 2.45(2, 105), p =.09). The results were not significant; thus H1 was not s upported for positive autonomy. In the dimension of negative autonomy, as shown in Table 6, individual ownership group had a mean of 2.06 and standard de viation of 1.32, chain ownership group had a mean of 2.19 and standard deviation of 1.31, and medium size ownership group had a mean of 2.12 and standard deviation of 1.09. In Table 7, the analysis of variance for complexity of the ownership revealed that there is no significant difference between the three groups based on levels of autonomy ( F = .08 (2, 105), p =.93). Table 8. Analysis of Variance for Ownership of Weblogs Positive Autonomy Level M S.D N Company 5.39 1.09 75 Individual 3.79 1.37 33 Ownership of the weblog Total 4.90 1.39 108 Negative Autonomy Level M S.D N Company 2.18 1.29 75 Individual 2.03 .98 33 Ownership of the weblog Total 2.13 1.20 108 Table 9. Mean Difference for Ownership of Weblogs Positive Autonomy df S.S M.S F Sig. Between Groups 1 58.90 58.90 42.43 .00 Within Groups 106 147.15 1.39 Total 107 206.06 Negative Autonomy df S.S M.S F Sig. Between Groups 1 .51 .514 .35 .55 Within Groups 106 154.29 1.46 Total 107 154.80 H2 predicted that journalist-owned webl ogs would be associated with higher perceived autonomy as compared to comp any-owned weblogs. In the dimension of
38 positive autonomy, as shown in Table 8, company-owned weblogs group had a mean of 5.39 and standard deviation of 1.09, individu ally-owned weblogs group had a mean of 3.79 and standard deviation of 1.37. In Table 9, th e analysis of varian ce for ownership of weblogs revealed that there is a significant difference be tween the two groups based on levels of autonomy ( F = 42.43(1,106), p <.05). Yet, H2 was not supported. Companyowned weblogs associated with higher aut onomy as compared to individually-owned weblogs, which is in the opposite direction of H2. In the dimension of negative autonomy, as shown in Table 8, company-owned weblogs group had a mean of 2.18 and standa rd deviation of 1.29, individually-owned weblogs group had a mean of 2.03 and standard deviation of .98. In Table 9, the analysis of variance for ownership of weblogs revealed that there is no significant difference between the two groups based on levels of autonomy ( F = .35(1,106), p =.55). Table 10. Analysis of Variance for Size of Organization Positive Autonomy Level M S.D N Small 4.45 1.85 29 Medium 5.19 1.19 34 Large 4.98 1.16 45 Size of organization Total 4.90 1.39 108 NegativeAutonomy Level M S.D N Small 1.89 .98 29 Medium 2.47 1.35 34 Large 2.04 1.19 45 Size of organization Total 2.13 1.20 108 Table 11. Mean Difference for Size of Organization Positive Autonomy df S.S M.S F Sig. Between Groups 2 8.98 4.49 2.40 .10 Within Groups 105 197.08 1.88 Total 107 206.06 Negative Autonomy df S.S M.S F Sig. Between Groups 2 6.09 3.05 2.15 .12 Within Groups 105 148.71 1.41 Total 107 154.80
39 H3 predicted that the size of the organi zation will be negative ly associated with journalistsâ€™ degree of perceived autonomy in managing weblogs. In the dimension of positive autonomy, as shown in Table 10, small size of organization group had a mean of 4.45 and standard deviation of 1.85, medium size of organization group had a mean of 5.19 and standard deviation of 1.19, and large size of organization group had a mean of 4.98 and standard deviation of 1.16. In Table 11, the analysis of variance for the size of organization revealed that there is a margin ally significant difference between the three groups based on levels of autonomy ( F = 2.40(2,105), p =.10). The results were not significant; thus H3 was not s upported for positive autonomy. In the dimension of positive autonomy, as shown in Table 10, small size of organization group had a mean of 1.89 and standard devia tion of .98, medium size of organization group had a mean of 2.47 and standard deviati on of 1.35, and large size of organization group had a mean of 2.04 and standard devia tion of 1.19. In Table 11, the analysis of variance for the size of organization revealed that there is no significant difference between the three groups based on levels of autonomy ( F =2.15(2,105), p = .12). Table 12. Regression Result Job, Ad-dependency, and Editors in Positive Autonomy I.V Standardize d Coeffificient s (Beta) T P R R2 Standard Error of the Estimate F (Constant) 3.02 .003 Job Satisfaction .23 -1.10 .273 Ad-dependency -.10 2.16* .033 Allegiance of Editors .14 1.32 .189 .35 .12 1.31 4.78 * p .05
40 Table 13. Regression Result Job, Ad-dependency, and Editors in Negative Autonomy I.V Standardize d Coeffificient s (Beta) t P R R2 Standard Error of the Estimate F (Constant) .96 .341 Job Satisfaction .15 1.60 .114 Ad-dependency .18 1.60 .114 Allegiance of editors -.07 -.60 .547 .22 .05 1.19 1.83 In the dimension of positive autonomy, H4 pred icted that higher levels of journalist job satisfaction will be negatively associated with journalistsâ€™ perceived freedom of expression in blogging. However, as shown in Table 12, job satisfacti on is not related to perceived freedom of expression in blogging ( = .23, t = -1.10, p =.27). The effect of job satisfaction on perceived freedom of expre ssion in blogging would not have been found to be statistically significant. Thus, H4 was not supported. H5 predicted that the extent to which j ournalists perceive their companyâ€™s revenue as being dependent on advertising sales will be negatively associated with journalistsâ€™ degree of perceived autonomy in mana ging weblogs. As shown in Table 12, addependency is closely related to percei ved freedom of expression of blogging ( =-.10, t =2.16, p <.05). The effect of ad-dependency on perceived freedom of expression in blogging would have been found to be statistic ally significant. Yet, it was statistically significant only for positive autonomy, not for negative autonomy. Thus, H5 was partially supported. H6 predicted that the extent to which journalists believe that their editor prioritizes business interests will be nega tively associated with journa listsâ€™ degree of perceived autonomy in managing weblogs. However, as displayed in Table 12, the allegiance of editors is not related to perceive d freedom of expression of blogging ( =.14, t =1.32,
41 p =.19). The effect of editorâ€™s business interest on perceive d freedom of expression in blogging would not have been found to be st atistically significant. Thus, H6 was not supported. In the dimension of negative autonomy, as shown in Table 13, there were no statistically significant relationships betw een journalistsâ€™ perceived freedom of expression in blogging and job satisfaction, ad-dependency, or allegiances of editors ( =.15, t =1.60, p =.11 for job satisfaction, =.18, t =1.60 , p =.11 for ad-dependency, and =-.07, t= -.60 , p =.55 for editorâ€™s business interest). There have not been found to be statistically significant relations between journalistsâ€™ perceive d freedom of expression in blogging and job satisfaction, ad-dependency, and editorâ€™s business interest in the dimension of negative autonomy. Table 14. Multivariate Analysis of Covariate Bureaucracy of Organization in Positive Autonomy M S.D N No layers of approval 4.50 1.81 16 Layers of approval existed 4.97 1.30 92 Total 4.90 1.39 108 Table 15. Mean Difference for Bureaucracy of Organization in Positive Autonomy df S.S M.S F Sig. Size of organization 1 2.39 2.39 1.26 .26 Types of employment 1 3.74 3.74 1.98 .16 Bureaucracy 1 1.27 1.27 .67 .41 Total 108 2800.44 df S.S M.S F Sig. Size of organization 1 2.69 2.69 1.41 .24 Bureaucracy 1 1.87 1.87 .98 .32 Total 108 2800.44 df S.S M.S F Sig. Types of employment 1 4.04 4.04 2.13 .15 Bureaucracy 1 2.13 2.13 1.27 .29 Total 108 2800.44
42 Table 16. Multivariate Analysis of Covariate Bureaucracy of Organization in Negative Autonomy M S.D N No layers of approval 2.38 1.27 16 Layers of approval existed 2.09 1.19 92 Total 2.13 1.20 108 Table 17. Mean Difference for Bureaucracy of Organization in Negative Autonomy Df S.S M.S F Sig. Size of organization 1 .34 .34 .23 .63 Types of employment 1 .77 .78 .53 .47 Bureaucracy 1 1.55 1.55 1.06 .31 Total 108 646.75 Df S.S M.S F Sig. Size of organization 1 .39 .39 .27 .61 Bureaucracy 1 1.32 1.32 .90 .34 Total 108 646.75 Df S.S M.S F Sig. Types of employment 1 .82 .82 .56 .45 Bureaucracy 1 1.33 1.33 .91 .34 Total 108 646.75 H7 predicted that bureaucracy of organizations will be negatively associated with journalistsâ€™ degree of perceived autonomy in managing weblogs. In the dimension of positive autonomy, as shown in Table 14, no la yers of approval group had a mean of 4.50 and standard deviation of 1.81, layers of approval group had a mean of 4.97 and standard deviation of 1.30 . In Table 15, the analysis of varian ce for bureaucracy of an organization revealed that there is no si gnificant difference between the two groups based on levels of autonomy ( F =.67(1,108), p =.41) across group size ( F =1.26(1, 108), p =.26), and employment ( F = 1.98(1,108), p =.16). When only controlling for the size of the organization, there is no si gnificant difference between la yers of approval of an organization on positive autonomy ( F =.98(1, 108), p =.32), and group size ( F = 1.41(1,108), p =.24). When only controlling for the types of employment, there is no
43 significant difference between layers of approval of an organization on positive autonomy ( F =1.27 (1, 108), p =.29), and employment ( F = 2.13(1,108), p =.15). Thus, H7 was not supported. In the dimension of negative autonomy, as shown in Table 16, no layers of approval group had a mean of 2.38 and standa rd deviation of 1.27, layers of approval group had a mean of 2.09 and standard deviation of 1.19 . In Table 17, the analysis of variance for bureaucracy of an organization revealed that there is no significant difference between the two groups based on levels of autonomy ( F =1.06 (1,108), p =.31) across group size ( F =.23(1,108), p =.63), and employment ( F = .53(1,108), p =.47). When only controlling for the size of the organiza tion, there is no significant difference between layers of approval of an or ganization on positive autonomy ( F =.90(1,108), p =.34), and group size ( F = .27(1,108), p =.61). When only controlling fo r the types of employment, there is no significant difference between la yers of approval of an organization on positive autonomy ( F =.91(1, 108), p =.34), and employment ( F = .56(1,108), p =.45). In sums, the relationship between the bureaucracy and the positive autonomy, bureaucracy and the negative autonomy would have not been found to be statistically significant. Thus, H7 was not supported. Additional Findings Table 18. Frequency for Weblog Update Item Frequency of Response (%/ N) Rarely 6% (7) About once a month 1% (1) About 2-3 times a month 5% (5) About once a week 12% (13) About 2-3 times a week 31% (34) Once a day or more often 44% (47) No answer 1% (1) Total 100% (108)
44 When participants were asked how freque ntly they updated their weblogs, 44% (47 people) of the sample answered one or mo re times a day, 31% (34 people) indicated about 2-3 times a week, 12% (13 people) res ponded about once a week, 6% (7 people) noted that they rarely update their weblogs , 5% (5 people) update about 2-3 times a month, and 1% (1 person) answered that he updated his weblog about once a month. Table 19. Frequency for Issues of Public Importance Item Frequency of Response (N/%) Very strongly disagree 4% (4) Strongly disagree 3% (3) Disagree 4% (4) Neutral 20% (22) Agree 21% (23) Strongly agree 27% (29) Very strongly agree 21% (23) No answer 0% (0) Total 100% (108) Regarding the degree to which participants believe their weblogs deal mostly with non-personal issues, 27% (29 people) of the sa mple chose the strongly agree response, 21% (23 people) of each sample answered agre e or very strongly agree, 20% (22 people) of the sample indicated a neutral response, 4% (4 people) answ ered very strongly disagree or disagree, and 3% (3 people) of sample answered strongly disagree. Table 20. Frequency for Interaction between Weblog Readers Item Get reactions (N/%) Respond to reactions (N/%) Not at all 2% (2) 3% (3) Sometimes 51% (55) 22% (24) Often 23% (25) 33% (36) Frequently 24% (26) 41% (44) No answer 0% (0) 1% (1) Total 100% (108) 100% (108)
45 When participants were asked to indicat e the frequency with which they receive reactions to their weblogs from readers, responses ranged from not at all to frequently: 2% (2 people) of the sample indicated that they never get reactions from their readers, 51% (55 people) sometimes get reactions, 23% (25 people) often get reactions, and 24% (26 people) frequently get reactions. Responses to the question asking particip ants how often they respond to their weblogsâ€™ readers ranged from not at all through frequently: 3% (3 people) of the sample answered that they never re spond to their readers, 22% ( 24 people) sometimes respond to their readers, 33% (36 people) often res pond to their readers, and 41% (44 people) frequently respond to their readers. In addition, in terms of the locations where participants most frequently update their weblogs, 50% (54 people) of the sample indicated that they upda te their weblogs in the office, and 35% (38 people) of the sample answered that they update their weblogs at home.
46 CHAPTER 5 DISCUSSION AND CONCLUSION Review of the Present Study The purpose of this research was to exam ine if U.S. journalists who run weblogs have autonomy in managing them. This rese arch proposed to examine the link between weblogs and journalistsâ€™ autonomy. For investig ating this research, gatekeeping theory was applied as a theoretical frame work. Based on previous gatekeeping research, Shoemaker and Reese (1996) elaborated gatekeep ing into a five level hierarchical model of influences on content: individual, media routine, organization, extramedia, and ideological level. Among theses levels, this research has focused on two of the levels: individual and organization, examining if gate keeping theory applie s as well to the nontraditional medium of weblogs. Specifically, this research focused on some influencing factors: types of ownership, the status of the blogs as either being company or individually run, size of the company, the journalistsâ€™ job satisf action, perceived addependency of the media organization, allegian ce of the editors, and bureaucracy of the media organization. These factors were all independent variables, while journalistsâ€™ perceived autonomy was the dependent vari able. A total of seven hypotheses were delivered. In order to test these hypothes es, a Web-based survey was conducted, and a total of 108 blogger journalists responded to this survey ove r the course of two weeks from March 22, 2006, to April 5, 2006. To determin e statistical significance, analysis of variance, regression tests, and multivariate analysis of covariate were performed.
47 Summary and Discussion of Results This study discovered journalistsâ€™ per ceptions of their organizationsâ€™ addependency act as influencing factors over jo urnalistsâ€™ perceived autonomy in managing weblogs. Complexity of ownership and size of organization may have an effect but were not supported. That is to say, th is study concluded that weblog s, a form of non-traditional media may be, influenced by ad-dependency factor. The first hypothesis, which assumed that th e complexity of media organizations has a negative relationship to blogger journalistsâ€™ autonom y, was not supported, but may have an effect. Blogger journalists who work for chain-owned organizations perceive less autonomy in blogging than those working fo r medium chain-owned organizations, and those working in medium chain-owned organi zations perceive less autonomy than those blogging in independently-owned organizations . Even though the differences in this study were not significant, further research ma y clarify that relationship. The significance of this finding is that this influencing fact or, organization complex ity, is as well applied to weblogs as would be anticipated by previ ous research indicating that ownership works as an influencing factor in traditional media (Bagdikian, 2004; Coulson and Hansen, 1995; Lacy, 1991; Picard, 1994; Shoemaker and Reese, 1996). At present, radio, television, cable, newspapers, magazines, and other media organizations attempt to grow through limitless mergers and strategic coalitions in order to compete for higher positions in the media market and obtain more sizable profits (McManus, 1994). As a result, these profit-seeking consolidat es may influence media content if the content is seen as being potentially harmful to the organizationsâ€™ pr ofits. Consequently, blogger journalists will also be influenced in their writing and blogging by their organizationsâ€™ profit-seeking
48 nature. However, this study did not find th at blogger journalists who work for more complex organizations possess less freedom of expression. The third hypothesis, which assumed that the size of the media organization has a negative relationship with blogger journali stsâ€™ autonomy, was not supported, but may have an influence. Larger companies are like ly to need more advertisements to support their companies, and for this reason, larg er organizations cannot ignore the profits brought in by their advertisers (Shoemaker a nd Reese, 1996). The result showed marginal support for this hypothesis, however, and perhap s the result would have been significant with a different sample composition. The fifth hypothesis, which assumed th at the ad-dependency of the media organization has a negative relationship with blogger journalistsâ€™ autonomy, was partially supported. Blogger journalists who perceived their organizationsâ€™ ad-dependency as being high have less perceived autonomy in blogging than those who do not. This finding is perhaps the most representative impli cation of the overall re sults of this study. That is to say, media cannot exist apart from advertising, and it follows that media cannot be completely free from profit-seeking motiv ations. Because advertising is the main profit avenue for media (McChesney, 2004), media cannot be free from advertisersâ€™ influence (Altschull, 1995), and this economic force thus impacts journalistsâ€™ writing (Price, 2003). This study conjectured that three factor s out of seven, which are ad-dependency, size, and complexity share a common similarity when compared to the other four factors, which are blogs ownership, job satisfacti on, editor, and bureaucracy. The first three factors were involved in orga nizational concept, whereas th e other four are involved the
49 more personal concept. However, when it come s to the concept of other four factors, it comes with human subjects. Thus, it is pr esumed that organiza tion involved factors impacting more on weblogs than personal factors. The second hypothesis, which assumed that journalist-owned weblogs will be associated with greater perceived autonomy as compared to company-owned weblogs, was statistically significant but in the opposite direction. That is to say, this study discovered that blogger journalists whose blogs belonged to their or ganizations perceived greater autonomy than journalists whose bl ogs were their own. This finding is possibly due to the limitation of the composition of the sample. Among the 108 participants in total, 74 people (69%) responded that thei r organizations host their weblogs and 67 people (63%) of participants responded that they freely e xpress their opinions on their weblogs. Although this study initially star ted to sample blogge r journalists who individually run their weblogs, snowball sa mpling allowed responses from numerous blogger journalists who write for company-run weblogs, and in the final sample, they accounted for 69% of the respondents. That is to say, because of the snowball method, company-supported blogger journa lists responded in greater numbers. Thus, it affected the data results. The fourth hypothesis, which assumed that journalistsâ€™ job sa tisfaction in their media organizations has a negative relations hip with blogger jour nalistsâ€™ autonomy, was not supported. Initially this hypothesis was base d on the previous rese arch indicating that, job satisfaction was referred to journalistsâ€™ overall evaluation of their works. Journalistsâ€™ job satisfaction represents combination of values, ideas, and beliefs of journalists themselves (Baron 1976; Bergen & Weaver 1988) . From this supposition, this study drew
50 the conjecture that as journalistsâ€™ job satisfaction is lowered, their preference toward weblogs will grow. This re jection can be explained by noting that besides the job satisfaction matter there are other reasons fo r blogging such as personal attachment and personal interests about blogging. Simply by th e journalists own preference, they may lean more towards weblog writing. They might enjoy receiving readersâ€™ feedback who write their personal views. From this consider ation, it is concluded th at there is no clear relationship between job satisfac tion and journali stsâ€™ autonomy. In testing of the sixth hy pothesis, it was discovered th at there was no relationship between the editorâ€™s business interests and blogger journalistsâ€™ perceived autonomy. In traditional media, editors who have more business interest tend to have a greater influence on journalistsâ€™ writi ng. Yet, when it comes to webl ogs, regardless of the degree to which editors tend to focus on the bottom line, they do not have an actual right or duty to approve their journalist-o wned weblog. However, this al so was a limitation of sample composition. Among responses, 74 people (69%) out of 108 wrote for company-owned weblogs, thus this led 67 people (63%) that responded that they freely express their opinions on their weblogs. It a ppears from the results that th e survey questions used to measure editor allegiance may not have accurately captured the concept. The seventh hypothesis examined whether heightened leve ls of bureaucracy of the organization have a negative relationship with blogger journalistsâ€™ autonomy, but was not supported. Bureaucracy issue is related to job satisfaction. While organizationsâ€™ bureaucracy was measured in terms of layers of approval, these layers of traditional approval are not actually applied to cyberspa ce or the approval of weblog writing, even when the blog is owned by the news organizatio n. In general, webl og writings might not
51 be subject to concrete layers of approval as is the case with traditional media writing. That would explain why the layers of bureau cracy showed no relationship to perceived autonomy in weblogs. In summary, the interesting point of this study is that out of the original seven hypotheses, the one partially supported hypothesis, H5, and the two marginally significant results for H1 and H3, involve circ umstances that typically occur in rather large organizations. The questi on of organizational size (H3), the presumption that chainowned media is larger than independent media (H1), the significant result for ad dependency (H5), and the positive rela tionship between ad-dependency and organizational size in previous research (M cManus, 1994), all are re lated to and support the same idea. Therefore, it might also be presumed that some of the limitations to journalistsâ€™ perceived autonomy that have long existed within larger organizations might also be found in cyberspace with relation to weblogs. There are indications that journalists in larger organizations tend to have more restrictions than the ones in smaller organizations (Cook and Banks, 1993; Cha n, Pan, and Lee, 2004). In conclusion, journalistsâ€™ weblogs are not media complete ly free from control of their affiliated organizations and are instead partially influenced by these organizations. There are some reasons for such a tendency. The first reason is that media organizations recognize the influence of weblogs in presenting issues and setting agendas in current society. Although the weblog is c onsidered to be a rather personal media format that focuses on individualized subjec ts, the impact of weblogs can never be overlooked because their messages are delivered using the most dominant network of all: the Internet (Hewitt, 2005).
52 When an issue is exposed on the web, small voices can easily turn into a massive roar, and the strength of the message may outmatch the voice of conventional media. Profit-focused organizations must be concer ned with the tone and messages presented by their weblog journalists sin ce these have the potential to work unfavorably against organizationsâ€™ ultimate goals. In fact, there have been a few incidents to make headlines that were ignited by citizen bloggersâ€™ activities. Trent Lottâ€™s resignation from th e Senate leadership in 2002 was a prime example of the power of weblog activity over traditional media (Gillmor, 2004). During his fellow senatorâ€™ s birthday party, Lott utte red words which could be considered racially insensitive, and his comments set off a controversy (Gillmor, 2004). Most of the major presses treated the incide nt lightly, but it was ex acerbated and widely spread through weblogs and eventually led to Lottâ€™s resignation as Se nate Majority leader (Gillmor, 2004). Another incident demonstrati ng the power of weblogs is that of New York Times reporter Jayson Blairâ€™s plagiari sm scandal in 2003, which received much coverage by conservative bloggers who were displeased by the liber al news tone of New York Times (Rosen, 2006). Weblogs also spread the fact that Dan Rather, CBS News anchor, relied on falsified documents in a report about President Bush during the 2004 presidential election (Eberhar t, 2005). From these incident s, media organizations have discovered the significant and farreaching impact of bloggers. Second, as media organizations focus more on profit-seeking a nd competing with other media organizations, they begin to view journalistsâ€™ weblogs as another means for their publicity or an extension of their media, rather than seeing it as an independent form of media. That is, media organizations envision journalistsâ€™ weblogs as suitable
53 marketing tools. In reality, many media organizations host th eir journalistsâ€™ weblogs on their website and run many current events su ch as Olympic Games weblogs for attracting readers or viewers who are losing interest in traditional media. Therefore, these media organizations often feel the need to overs ee their journalistsâ€™ webl og writing and regulate any content that may negatively affect profit s derived from advertisers. Hewitt (2005) emphasized that when it comes to the jour nalistic function of we blogs, constructing a transparent blogging system should be more focused on freedom of expression for the blogger journalists. One of the survey questions delivered in this study asked participants to cite reasons for shutting down their weblog if they had previously run one. Out of 108 respondents, 17 (16%) in dicated that they had run a webl og but chosen to shut it down, with one respondent giving the following reason: â€œStated policy is that journalists are not to have Weblogs without prior approval. I di dn't know who to ask and had no interest in being grilled about it.â€ His/her answer would not be generalized to others. However, it is worth to taking notice of the fact that as the number of j ournalistsâ€™ weblogs con tinue to grow, some media organizations are becoming more concer ned with their journalistsâ€™ weblog writing. Obviously some factors limit the range of topics blogger journalists write about; however, it is important to note that some ot her factors, such as weblog ownership, job satisfaction, editor allegiance, and bureaucracy, apparently do not affect the perceived autonomy of the journalist bloggers. This finding suggests that the expansion of journalistsâ€™ weblogs might functi on as a form of alte rnative media, enabling journalists to produce work that would not appear in the traditional news outlets.
54 In addition, this study did not support any hypotheses in the dimension of negative autonomy, which was defined as the preventio n of writing what a journalist wants. Addependency (H3) as well was only supporte d in the dimension of positive autonomy, which was defined as freedom of expression. This may be due to blogger journalists who responded to the survey might feel uncomfor table with answering negative questions. That is, they might not have responded to th e survey honestly because of concerning their editors or just their own biases, without interventi on of anything or anybody. Limitations and Future Research First, we might compare the journalist-owned and company-owned weblogs as two groups, if we had a bigger sample of each one. Among the 108 participants in total, 74 people (69%) responded that th eir organizations host their weblogs and 67 people (63%) of participants res ponded that they freely express their opinions on their weblogs. Although this study initially st arted to sample blogger jour nalists who individually run their weblogs, snowball sampling allowed re sponses from numerous blogger journalists who write for company-run weblogs. It mean s, in the process of snowball sampling, untargeted bloggers might be included and this thus skewed data. Second, the ad-dependency of organizatio ns may be more accurately measured based on organization revenues. In this st udy, the ad-dependency of organizations was based mainly on journalistsâ€™ perception of their organizatio nsâ€™ level of involvement in this category. Third, other media gatekeeping factors s hould be used to test weblogs. According to Shoemaker & Reese (1996) , there are some other dimensions aside from the organizational and individual level factors that were examined in this study that could be considered.
55 Conclusion Theoretically, this study proposed to investigate the relationship between gatekeeping theory and weblogs. Overall, or ganizational level gatekeeper existed in weblogs but â€œMr. Gatesâ€ named the editor (W hite, 1050) did not. News items selected by gatekeepers are then sent to the audience. A key aspect in the process of sending news items to the gatekeepers is the idea that some of them may be lost and changed (McQuail and Windahl, 1981). This study found out th at organizations having higher addependency performed more of a gatekeeping role in weblogs. However, â€œMr. Gatesâ€ apparently had no influence in weblogs. Acco rding to White, Gatesâ€™ reasons for selection news were very subjective, including ed itorâ€™s personal evaluation of stories. Gatekeeping is an established theory in mass communication research. Singer (1998) described that â€œalthough only a few publis hed studies have specifically addressed gatekeeping in the online environment, there is some evidence that journalists see that function as evolving and adapting rather than disappearingâ€ (p. 3). This study examined whether the six levels of orga nizational and one level of indi vidual factors which acted as gatekeeping role in traditional media worked as well in weblogs. The study concluded that organizations expanded their gatekeei ng roles onto weblogs when they were addependent. In other words, gatekeeping theory applies to weblogs ev en if only partially. Weblogs, a new type of media, are consid ered as alternative media form due to merits such as low running costs and real-t ime update capabilities. Based on these notions, this study examined whether blogger journalists are influenced by traditional media factors in the same manner as traditio nal journalists. Findings of this study indicate that some traditional factors, such as media owners hip, size of the organization, and ad-dependency, limit journalistsâ€™ percei ved autonomy even in new types of media
56 like weblogs. On the other hand, some traditional factors, such as weblog ownership, journalistsâ€™ job satisfaction, allegiance of the editors to their organization, and bureaucracy of the organization, did not influence blogger journalists. In conclusion, journalistsâ€™ weblogs ar e not free from the oversight of their organizations but are partially influenced by or ganizational interests. However, j-bloggers indicate they might have more freedom of expr ession in their blogs than in their writings for traditional media. Therefore, these findi ngs seem to indicate the journalistsâ€™ weblogs might function as alternative media in a way similar to citizensâ€™ weblogs.
57 APPENDIX A SURVEY QUESTIONNAIRE 1) Which corporation owns your news organization? Advance____ Bertelsmann ____ Copley Newspapers____ Cox____ Disney____ Gannett ____ Hearst____ International Data____ Knight Ridder ____ McGraw-Hill____ News Corporation____ Panax____ Reed Elsevier____ Scripps-Howard____ The New York Times Company ____ The Tribune Company (Chicago) ____ Time Warner____ Viacom____ My organization is independently owned ____ Other____ 2) Does your company host your weblog? Yes____ No____ 3) Do you host your own weblog separately from any employer?
58 Yes____ No____ 4) Does your organization put journali stsâ€™ weblogs on the website of the organization? Yes____ No____ 5) How many journalists work in your newsroom? Less than 20____ 20-60____ 60-100____ 100-140____ more than 140____ 6) I am employed by organization Full time ____ part time ____ free lan cer ___ I am not employed by any organization___ *Please check the answer that best indi cates the degree to which you agree with each statement. Very strongly disagree Strongl y disagre e disagre e neutral agree Strongl y agree Very strongl y agree 7) I feel loyal toward my current company 8) I enjoy adopting new technology. 9) I am satisfied with my current work. 10) I express my opinion freely in my work place. 11) I want to leave journalism. *Please check the box that best indicates the degree to which you agree with each statement. Very Strongldisagreneutral agree StronglVery
59 strongly disagree y disagre e e y agree strongl y agree 12) I might be reprimanded if I write against advertisers. 13) My company cares about bottom line. 14) Advertisers influence contents in my companyâ€™s news products. 15) My company depends on advertising. *Please check the box that best indicates the degree to which you agree with each statement. Very strongly disagree Strongl y disagre e disagre e neutral agree Strongl y agree Very strongl y agree 16) My editor has power in the organization. 17) My editor stresses that freedom of expression is important in journalism 18) My editor cares about bottom line. 19) My editor is open to the opinions of reporters 20) My editor supports my work.
60 21) On average, how many people must approve your stories before they are published in the news organization, or broadcast in your news program? None____ 1____ 2____ 3____ more than 3 people____ Weblog adoption (Dependent Vari ablefreedom of expression) 22) How long have you been publishi ng your weblog? ____years and ____months. 23) If you used to run a we blog, but shut it down, why? No time____ no more interest____ pressure from bosses____ other____ 24) How often do you update your weblog? rarely____ about once a month____ about 2-3 times a month____ about once a week____ about 2-3 times a week____ once a day or more often____ *Please check the box that best indicates the degree to which you agree with each statement. Very strongly disagree Strongl y disagre e disagre e neutral agree Strongl y agree Very strongl y agree 25) My work mostly deals with issues of public importance (nonpersonal issues). 26) Where do you usually update your weblog? Home____ office____ other____
61 27) How often do you get reactions (comme nts, e-mails, etc) from your weblog readers? Not at all ____ sometimes____ often____ frequently____ 28) How often do you respond, if they do receive comments? Not at all ____ sometimes____ often____ frequently____ *Please check the box that best indicates the degree to which you agree with each statement. Very strongly disagree Strongl y disagre e disagre e neutral agree Strongl y agree Very strongl y agree 29) My company encourages journalists to run weblogs. 30) I have received positive reactions (comments, etc) to my weblog from my supervisor, coworkers, and other people in the company. 31) I have received negative reactions (comments, etc) to my weblog from my supervisor, coworkers, and other people in the company. 32) I believe my weblog is monitored by
62 upper-level managers in the company. 33) My weblog has been censored by upper-level managers in the company. 34) I freely express my opinion on my weblog. Demographics 35) Please check your gender. Male____ Female____ 36) How old are you? _____years old 37) How long have you been working as a journalist? less than 1 year____ at least 1, but less th an 5 years____ at least 5, but less than 10 years____ at least 10, but less than 15 years____ at least 15, but less than 20 years____ more than 20 years____ 38) How long have you been working at your current company? Or, if free lancer, how long have you been working? less than 1 year____ at least 1, but less th an 5 years____ at least 5, but less than 10 years____ at least 10, but less than 15 years____ at least 15, but less than 20 years____ more than 20 years____
63 39) Which beat do you cover? (Check all that apply) city/ metro____ sports____ features____ arts & entertainment____ business____ national____ international____ technology____ other____
64 LIST OF REFERENCES Adams, J. S. (1980). Interorg anizational processes and or ganizational boundary spanning activities. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior (Vol. 2, pp. 321-355). Greenwich, CT: JAI. Adams, T., and Cleary, J. (2006). Surveyi ng broadcast newsrooms: using Web-based technology to reach reluctant respondents. Presented to the Broadcast Education Associationâ€™s Annual Confer ence, Las Vegas, April 2006. Agresti, A., and Finlay, B. (1997). Statistical methods fo r the social sciences (3rd ed.). Upper Saddle River, NJ: Prentice Hall. Altheide, D. L. (1976). Creating reality: How TV news distort events, Beverly Hills. CA: Sage. Altschull, J. H. (1995). Agents of power: The role of the media in human affairs. (2nd ed.). New York: Longman. Armstrong, C. L., and McAdams, M. J. (2006). Believing blogs? Examining the influence of gender cues on credibility. Will be presented to Mass Communication and Society Division of the Association fo r Education in Journalism and Mass Communication Annual Conference in San Francisco, CA, Aug. 2-5, 2006. Babbie, E. (2006). The practice of social research. (11th ed.). Belmont, CA: Wadsworth/Thomson Learning. Bagdikian, B. H. (2004). The new media monopoly . Boston, MA: Beacon. Bass, A. Z. (1969). Refining the â€œgat ekeeperâ€ concept: A UN radio case study. Journalism Quarterly, 46. 69-72 Bantz, C. R. (1985). News organizations : conflict as a crafted cultural norm. Communication , 8, 225-244. Benson, R., and Neveu, E. (2005). Bourdieu and the journalistic field . City Polity Press. Bergen, L. A., and Weaver, D. H. (1988). Job satisfaction of daily newspaper journalists and organizaton size. Newspaper Research Journal, 9 (2), 1-13. Blood, R. (2002). The weblog handbook: Practical advic e on creating and maintaining your blog . Cambridge, MA: Perseus Publishing.
65 Blood, R. (2003, June 11). The wild world of "open-source media," Business Week Online, Technology section (1467 words). Breed, W. (1955). Social control in newsroom: A functional analysis. Social Forces , 33, 326-335. Campbell, D. T., and Fiske, D. W. (1959). Convergent and discriminate validation by the multitrait-multimethod matrix. Psychological Bulletin , 56, 81-105. Carr, P. (October 2004). The terrifying power of the bloggers, The Guardian, http://www.mediaguardian.co.uk/me diaguardian/story/0,7558,1248609,00.html Chan, J., Pan, Z., and Lee, F. (2004). Prof essional aspirations and job satisfaction: Chinese journalists at a time of change in the media, Journalism and Mass Communication Quarterly, 81 (2). Cook, B. B., and Banks, S. R. (1993). Predic tors of job burnout in reporters and copy editors. Journal of Quarterly, 70 , 108-117. Coulson, D. C., and Hansen, A. (1995). The L ouisiville Courier-Journal's news content after purchase by Gannett. Journalism and Mass Communication Quarterly, 72 , 205-215. Cronbach, L. J. (1951). Coefficient alpha and the internal st ructure of tests. Psychometrika , 16, 297-334. Dimmick, J. (1974). The gatekeeper: An uncertain theory. Journalism Monographs, 37 , 1-39. Donohue, G. A., Tichenor, P. J., and Olien, C. N. (1972). Gatekeeping: mass media systems and information control. Beverly Hills CA: Sage Publications. Eberhart, D. (January 2005). How the blogs torpedoed Dan Rather, Newsmax, http://www.newsmax.com/archives/arti cles/2005/1/28/172943.shtml accessed: January 10, 2005. Epstein, E. (1974). News from nowhere. New York: Vintage Books. Gans, H. (1979a). Deciding whatâ€™s news . New York: Pantheon. Gillmor, D. (2004). We the media: Grassroots journalis m by the people, for the people . Sebastopol, CA: Oâ€™Reilly Media. Gieber, W. (1964). News is what newspapermen make it. In L. A. Dexter & D. M. White (Eds.), People, society and mass communication. New York: Free Press. Gronlund, N.E. (1985), Measurement and evaluation in teaching (5th ed.). New York, Macmillan.
66 Grossman, L. (October 2004). Time , Available online at www.time.com/time/magazine/article/o,9171,1101040621-6507321,00.html,accessed: October 20, 2004. Hackman, J. R., and Oldham, G. R. (1976). Mo tivating through the design of work: test of a theory. Organizational Behavior and Human Performance , pp. 250-276 Hair, J. F., Anderson, R. E., Tatham, R. L., and Black, W. C. (1995). Multivariate data analysis . (5th ed.). Upper Saddle River, NJ: Prentice Hall. Hewitt, H. (2005). Blog: Understanding the informa tion reformation thatâ€™s changing your world. Nashville, TN: Thomas Nelson, Inc. Heyboer, K. (2004). Bloggi nâ€™ in the newsroom, American Journalism Review , 25(8). Hodson, R. (1991). The active worker: complia nce and autonomy at the workplace. Journal of Contemporary Ethnography , pp. 47-78. Horan, P. M., DiStefano, C. and Motl, R. W. (2003). Wording effects in self-esteem scales: Methodological artif act or response style? Structural Equation Modelling, 10(3), 435-455. Johnstone, J. W. C., Slawski, E., & Bowman, W. (1976). The news people: A sociological portrait of Americ an journalists and their work. Urbana, IL: University of Illinois Press, 184. Jones, D. A. (2002). The polarizi ng effect of new media messages. International Journal of Public Opinion Research, 14 (2). Kerlinger, F., and Lee, H. (2000). Foundations of Behavioral Research , Orlando, FL: Harcourt College Publishers. Klotz, R. J. (2004). The politics of Internet communication, Lanham, MD: Rowman and Littlefield Publishers. Lacy, S. (1991). Effects of group ownership on daily newspaper content. Journal of Media Economics, 4, 35-47. Lewin, K. (1947). Frontiers in group dynamics: 2. Channels of group life; social planning and action research. Human Relations, 1 , 143-153. Liebler, C. (1994). How race and gend er affect journalistsâ€™ autonomy. Newspaper Research Journal , 7. Marsh, H. W. (1996). Positive and negativ e global self-esteem: A substantively meaningful distincti on or artifactors? Journal of Personality and Social Psychology, 70(4), 810-819.
67 McChesney, R. W. (2004) The problem of the media: U .S. communication politics in the twenty-first Century. Monthly Review Press. McFedries, P. (2003, December). Blah, blah, blog , IEEE Spectrum, p. 60. McManus, J. H. (1994). Market driven journalism: Let the citizen beware. Thousand Oaks, CA: Sage. McNelly, J. T. (1959). Intermediary communicat ors in the internati onal flow of news. Journalism Quarterly, 36, 23-26. McQuail, D., and Windahl, S. (1981). Communication models for the study of mass communication. New York: London. Nunnally, J. C. (1978). Psychometric theory , New York: McGraw-Hill Inc. Pain, J. (2005). Boggers, the new heralds of free expression. Handbook for bloggers and cyber-dissidents. Palser, B. (2003). Free to blog? American Journalism Review, 25 (5), 62. Perseus Development Corporation's Blog Survey (October 2003). White Paper â€œThe Blogging Iceberg: Of 4.12 Million hosted weblogs, most little seen, quickly abandonedâ€ www.perseus.com/blogsurvey/thebloggingiceberg.html accessed: October 18, 2003. Picard, R. G. (1994). Institutional ownershi p of publicly traded U.S. newspaper companies. Journal of Media Economics, 7 (4), 49-64. Prakash, Ved and Lounsbury, W John (1983). A reliability problem in the measurement of disconfirmation of expectations. Advances in Consumer Research , 10(1), 244249. Price, C. (2003). Interfering owners or meddling advertisers: How network television news correspondents feel about ownership a nd advertiser influen ce on news stories. Journal of Media Economics, 16 (3), 175-188. Robinson, S. (2006). The mission of the j-blog. Journalism , 7(1), 65-83. Roshco, B. (1975). Newsmaking . Chicogo, IL: University of Chicago Press. Rogers, E. M. (2003). Diffusion of Innovations. (5th ed.). New York: Free Press. Rosen, J. (March 2006). Great blogger bake off, Press Think, http://journalism.nyu.edu/pubzone/weblogs /pressthink/2006/03/24/bd_wapo.html accessed. March 3, 2006. Samuelson, M. (1962). A standardized test to measure job satisfaction in the newsroom. Journalism Quarterly, 39 , 285-291.
68 Scholl, A., and Weischenberg, S. (1999). Aut onomy in journalism: how it is related to attitudes and behavior of media professionals. Web Journal of Mass Communication Research. 2(4). Schramm, W. (1963). The challenge to communi cation research. In R. O. Nafziger & D. M. White (Eds.), Introduction to mass communication research (pp.3-31). Baton Rouge: Louisiana State University Press. Schudson, M. (2005). The power of news. Cambridge, MA: Havard University Press. Singer, J. B. (1998). Online journalist: Foundati ons for research into their changing roles. Journal of Computer-Mediated Communication , 4(1). Singer, J. B. (2005). The political j-blogger. Journalism, 6(2), 173-198. Shoemaker, P. J. (1991). Communication concepts 3: Gatekeeping . Newbury Park, CA: Sage. Shoemaker, P. J., and Reese, S. D. (1996). Mediating the message: Th eories of influences on mass media content. White Plains, NY: Longman. Trammell, K. D. (2004). Celebrity blogs: Investigation in the persuasive nature of twoway communication regarding politics. Doctoral dissertation, Un iversity of Florida, Gainesville. Tuchman, G. (1972). Objectivity as strategic ritu al: An examination of newsmenâ€™s notions of objectivity. American Journal of Sociology, 77(4), 660-679. Tuchman, G. (1977). The exception proves that ru le: The study of routine news practice in Hirsch, Miller, and Kline (Eds.). Strategies for comm unication research, Beverly Hills, CA: Sage, 43-62. Underwood, D. (1998). When MBAs rule the newsroom. Columbia Journalism Review , 23-32. Wall, M. (March 2004). Blogs as black mark et journalism: a new paradigm for news, Berglund Center for Internet Studi es, (consulted: 11 January 2005): http://bcis.pacificu.edu/journal/ 2004/02/wall.php accessed: March 5, 2004. Weaver, D. H., and Wilhoit, G. C. (1986). The American Journalists : A portrait of U.S news people and their work. Bloomington, IN: Indiana University Press, 163. Weaver, D. H., and Wilhoit, G. C. (1996). The American journalist in the 1990s: U.S news people at the end of an era . Mahwah, NJ: Erlbaum. Westley, B. H., and McLean, M. S. (1957). A conceptual model for communications research. Journalism Quarterly , 34(4), 31-38.
69 White, D. M. (1950). The Gatekeeper: A case study in the selection of news. Journalism Quarterly, 27, 383-390. Williams, D. (2002). Synergy bias: conglom erates and promotion in the news. Journal of Broadcasting and Electronic Media , 46(3), 453-472. Winer, D. (May 2003). "What Makes a Weblog a Weblog?" Weblogs at Harvard Law, http://blogs.law.harvard.e du/whatmakesaweblogaweblog accessed: May 2, 2003. Zelizer, B. (2004). Taking journalism seriously , Thousand Oaks, CA: Sage.
75 BIOGRAPHICAL SKETCH Hyeri Choi was born in Korea. In academic aspect, she earned her B.A. in English language & literature from Kangwon National University, double majoring in communications. She graduates from Univer sity of Florida in August 2006 with M.A.M.C. She will pursue her doctoral study in journalism at the University of Texas at Austin from August 2006. In profession aspect , she worked as a political reporter in Kangwon, Korea, for three years.