USF Libraries
USF Digital Collections

Content differences between print and online newspapers


Material Information

Content differences between print and online newspapers
Physical Description:
Smith, Jessica E
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:


Subjects / Keywords:
Local news
Content analysis
Dissertations, Academic -- Mass Communications -- Masters -- USF
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


ABSTRACT: The Internet provides the opportunity to develop a new way to present journalism, but many scholars say newspaper Web sites do nothing but mirror their print parents. This study used content analysis to compare the content of stories in five newspapers with their Web counterparts, and it examines whether reporter affiliation or a story's geographic emphasis has a relationship with the story's amount of contextual elements. These elements could include photos, graphics, or multimedia or interactive components online. This approach applied gatekeeping theory to publications that have editions in two media. This study examined the five largest newspapers in the South over 14 days, collecting a sample of 635 stories on the front pages and metro section front pages of the papers. Nearly all stories in the sample appeared on the newspapers' Web sites, and story content was the same 96% of the time. The study found that 85% of print stories were published with at least one contextual element, but only 58% of online stories had at least one such element. About a third of the sample had at least one contextual element in common between print and online versions of a story, while about 20% of the sample had entirely unique sets of contextual elements in print and online. Newspapers are no more likely to publish additional contextual elements with local stories than any other type of content. This effort focused on storytelling components; it examined whether print and Web editions of newspapers tell stories differently---whether they are complementary or competitive.
Thesis (M.A.)--University of South Florida, 2005.
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Jessica E. Smith.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 87 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001912585
oclc - 173987143
usfldc doi - E14-SFE0001332
usfldc handle - e14.1332
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001912585
003 fts
005 20071010140853.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071010s2005 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001332
1 100
Smith, Jessica E.
0 245
Content differences between print and online newspapers
h [electronic resource] /
by Jessica E. Smith.
[Tampa, Fla] :
b University of South Florida,
3 520
ABSTRACT: The Internet provides the opportunity to develop a new way to present journalism, but many scholars say newspaper Web sites do nothing but mirror their print parents. This study used content analysis to compare the content of stories in five newspapers with their Web counterparts, and it examines whether reporter affiliation or a story's geographic emphasis has a relationship with the story's amount of contextual elements. These elements could include photos, graphics, or multimedia or interactive components online. This approach applied gatekeeping theory to publications that have editions in two media. This study examined the five largest newspapers in the South over 14 days, collecting a sample of 635 stories on the front pages and metro section front pages of the papers. Nearly all stories in the sample appeared on the newspapers' Web sites, and story content was the same 96% of the time. The study found that 85% of print stories were published with at least one contextual element, but only 58% of online stories had at least one such element. About a third of the sample had at least one contextual element in common between print and online versions of a story, while about 20% of the sample had entirely unique sets of contextual elements in print and online. Newspapers are no more likely to publish additional contextual elements with local stories than any other type of content. This effort focused on storytelling components; it examined whether print and Web editions of newspapers tell stories differently---whether they are complementary or competitive.
Thesis (M.A.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 87 pages.
Adviser: Kenneth C. Killebrew, Ph.D.
Local news.
Content analysis.
Dissertations, Academic
x Mass Communications
t USF Electronic Theses and Dissertations.
4 856


Content Differences Between Print and Online Newspapers by Jessica E. Smith A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts School of Mass Communications College of Arts and Sciences University of South Florida Major Professor: Kenneth C. Killebrew Jr., Ph.D. Timothy E. Bajkiewicz, Ph.D. Randy E. Miller, Ph.D. Date of Approval: November 15, 2005 Keywords: gatekeeping, convergence, shovelware, local news, content analysis Copyright 2005, Jessica E. Smith


Acknowledgments Thanks to Deena Kemp and Erika Llenza for several hours of coding help; thanks to Norma Smith and Nancy Smith for reading headlines to me; thanks to Dr. Cheryl Bacon and Susan Lewis for brainstorming a nd encouraging; thanks to Charles and Deanne Smith for not asking too often when Id be done; and thanks to the USF faculty members who have shaped the way I understa nd mass communication. Special thanks to my thesis committee members, Dr. Randy Miller and Dr. Tim Bajkiewicz, and chair, Dr. Ken Killebrew, for their availability, patie nce, and guidance through this process.


i Table of Contents List of Tables iii Abstract iv Chapter 1: Introduction 1 Chapter 2: Literature Review 7 Gatekeeping 7 Broadening definitions 13 Framing 14 Television gatekeeping 15 Online gatekeeping 16 Trends in online newspapers 20 Shovelware 21 Interactivity and multimedia use 22 Local news online 25 Chapter 3: Method 29 Chapter 4: Results 36 Chapter 5: Conclusion 46 Limitations 49 Future research 51 References 54 Appendices 65


ii Appendix A: Print and online sample articles 66 Appendix B: Coding sheet 84 Appendix C: Coding instructions 85


iii List of Tables Table 1 Distribution of stories by newspaper 36 Table 2 Comparison of contextual elemen ts between print and online stories 37 Table 3 Elements in common between print and online stories 38 Table 4 Number of stories with t ypes of print elements present 39 Table 5 Number of stories with t ypes of online elements present 40 Table 6 Print reporter affiliation and mean print elements 42 Table 7 Online reporter affiliation and m ean frequency of photos and infoboxes 42 Table 8 Online reporter affiliation and mean types of elements represented 43 Table 9 Story geographic emphasis and m ean frequency of print elements 44 Table 10 Story geographic emphasis and m ean types of elements represented 44


iv Content Differences Between Print and Online Newspapers Jessica E. Smith ABSTRACT The Internet provides the opportunity to develop a new way to present journalism, but many scholars say newspaper Web sites do nothing but mirror their print parents. This study used content analysis to compare the content of stories in five newspapers with their Web counterparts, and it examines whether reporter affiliation or a storys geographic emphasis has a relationship with th e storys amount of contextual elements. These elements could include photos, graphics, or multimedia or interactive components online. This approach applied gatekeeping theo ry to publications that have editions in two media. This study examined the five largest newspapers in the South over 14 days, collecting a sample of 635 stor ies on the front pages and metro section front pages of the papers. Nearly all stories in the sample appe ared on the newspapers Web sites, and story content was the same 96% of the time. The study found that 85% of print stories were published with at least one contextual element, but only 58% of online stories had at least one such element. About a third of the samp le had at least one c ontextual element in common between print and online versions of a story, while about 20% of the sample had entirely unique sets of contextual elements in print and online. Newspapers are no more likely to publish additional contextual elements with local stories th an any other type of


v content. This effort focused on storytelli ng components; it examin ed whether print and Web editions of newspapers tell stories diffe rentlywhether they are complementary or competitive.


1 Chapter 1: Introduction The Internet provides the techni cal capability for a user to read an article, examine its sources, and interact th rough a natural conduit that ot her media do not provide. A newspapers Web site can provide e-mail a ddresses or discussion forums that make journalists accessible while fo stering community discussion. A site can offer audio clips from interviews, text of government records, and interactive maps that all can change the way a reader understands a story. News online provides the opportunity to develop a whole new way to present journalism, and Jan Schaffer (2001) of the Pew Center for Civic Journalism suggested that this be done with a much more interactive toolbox. Intera ctivity is one of the things that gives the Internet value as a medium. Ne wspapers can provide in -depth stories, and television gives pictures and sounds. When providers offer th ese elements online, users often confront technological limitations. These technical obstacles may be overcome in a few years, but Web journalism still should be able to offer something more, something unique now. If interactivity is the Internets outstanding ch aracteristic (Mo rris, 2001), it seems that online journalism should be taking ad vantage of it with stories that engage the reader and with tools such as e-mail links to journalists and officials, documents available to view, databases, and hyperlinks. Some media have been slow to us e interactive elemen ts, though (Dibean & Garrison, 2001; Greer & Me nsing, 2004; Singer, 2002; Tankard & Ban, 1998). The current environment has media trying to determine whether online news is


2 complementary or competitive for thei r operations (Dutta-B ergman, 2004) while journalists are trying to understand how to work in multiple media platforms (Huang, Davison, Shreve, Davis, Bettendorf, & Nair 2003; Killebrew, 2004). Others say the high-tech footnotes (Weinberg, 1996) pr ovided by documents and audio clips with stories online provide credibility but reduce on line journalists to repackagers of news. In a time when scholars and practit ioners have such questions, it seems crucial to examine the ways newspapers and their Web sites dist ribute news. Interactivity is a buzz word used about news Web sites, and the ability fo r a user to give feedback or choose which elements of a story to examine certainly distinguishes a story on a site from its print counterpart. Perhaps an equally popular word fo r added features of a newspaper Web site is multimedia, a form of presentation that uses audio, video, graphics, or other methods to give users different pieces of a story. Interactivity and multimedia capability are integrated features of the In ternet, and it is important to understand how online news sites use these techniques. Newspapers have a tradition of seek ing, gathering, proces sing, and producing news in a one-way daily delivery, but the Intern et can give users the ability to make the reporting process more transparent if site vi sitors can see, hear, or read the sources reporters relied on. In order to get to a poi nt where sites make full use of the online functions that make the Web unique, Lowrey ( 1999) said both journalists and users must develop new schemas for processing news on line instead of viewing it as a modified version of print journalism. This is an opportunity for newspapers to move past shovelwarecontent pushed directly from the print product to a Web pageand convert stories into forms that make them worthwhile for the online user. Online journalists have


3 no template for accomplishing that goal, howev er. Greer and Mensing (2004) make an excellent point: Newspapers are still working to find interactive elements that function well in an online news environment (p. 109). Finding an online model that is valuable to users and cost-effective for newspapers will not happen overnight, but it is important in producing print and online media th at continue to be viable. Although many newspapers have separate staffs for their online product, often the Web site is essentially the same as the pr int edition (Tewksbury & Althaus, 2000) or has fewer stories than what appears in prin t (Peng, Tham, & Xiaoming, 1999; Regan, 1995; Singer, 2001). With a news hole limited only by the size of a newspa pers servers, the Web offers a logical home for more inform ation than appears in the newspaper. A newspaper Web site that does not differentiate itself from the print newspaper in daily content has no unique quality to draw users, an d without this, a site has no leverage to make a profit with advertiser s or through paid co ntent (Chyi, 2005; Chyi & Sylvie, 1998). Several scholars (Fortunati, 2005; Pav lik, 1997) suggested that news Web sites develop in three stages. A first-stage site mirro rs its print parent, and sites in the second stage add some interactive, multimedia, or cu stomizable features. Pavlik (1997) said the third stage, a rarity at that time, would pr esent new forms of storytelling and a different understanding of what constitutes a community. A few years later, Fortunati (2005) said the mature site (2005, p. 30)one that has mastered multimedia presentation and is trying to develop or improve its economic returnwas a pr esent reality, at least for European media. A site that has learned to use multimedia techniques, however, may apply them only sporadically. This could indi cate that news sites still do not exploit the characteristics that make them a new, different medium. The evolution of online


4 newspapers may not be complete, so it is im portant for scholars to evaluate how news sites function now to see the di rection they might move as online newspapers come into their own. Although companies are undoubted ly at the stage of wan ting a sustainable return on investments in all products, online or othe rwise, many newspaper s ites at first glance appear to be little more than reflections of the days edi tion with perhaps some standing extra features. Research suggest s that there is an audience fo r both print and Web as they currently exist (Dutta-Bergman, 2004), but Weiss (2004) said online newspapers must reinvent themselves as the novelty of their innovation wears off or risk failure. Online newspapers are competitors if their content is free shovelware from the print edition, but news online can be complementary if sites o ffer different information or features. This study examines daily news content in print a nd online editions to determine whether the editions are competitive or complementary. At some level, the law of diminishing returns suggests that a newspaper will put money and effort into its Web presence until the point that more resources would not improve the financial bottom line, even if the product still could be improved. Shoemaker and Reese (1996) address the desire of news organizations to efficiently use resources; the scholars media content research fits into a fourth phase of agenda-setting research (McCombs & Shaw, 1993). Shoemaker and Reese (1996) created a hierarchical model of influences on media content that include s routines and organizational forces. The structural logic (p. 37) of a newspapers staffing structure and reporting norms shape the stories that readers see. If a newspapers workflow calls for reporters and their editors to produce stories, editors and designers to produce display type and add photos and


5 graphics, and Web editors simply to format that work for the Web, users will see a mirror of what appears in print. This structure creates an assembly line mentality, and Shoemaker and Reese suggest that the person at the end of the line, the editoror in this case, the Web editormay have little investme nt in the final product. If Web editors can ask reporters for source materials, such as documents and taped interviews, they face the challenge of creating a presentation that was not grown up in an online medium, (DeJean, 1995). This kind of stor ytelling in effect deconstruc ts a story and presents the whole as well as its parts. It can require several hours to produ ce one story like this online, and the outcome is a story that mi ght be different than one produced by a Web editor involved with the reporter in the be ginning. The final Web content may depend on the employee structure the organization has in place for its newspaper and Web site. The content that makes it onto a newspape r site might be only part of what the print edition offers that day, which is a process recommended by the online editor of The Christian Science Monitor (Regan, 1995). Singer (2001) also found that print editions ran more than twice as many stories as their corresponding online editions. Both the stories selected from the print edition to be posted online and the elements such as photos and graphics selected to go with those stories undergo some kind of gatekeeping process to determine whether they will move from the print medium to the Web. Gatekeeping, one of the oldest mass communicati on theories, has interesting imp lications for online media. Singer also found a strong local orientat ion of online newspapers, which makes sense considering that staff writers w ould produce most of the content about a newspaper's metropolitan area. A newspaper might find it relatively simple to post wire stories from the nation or region, but its lo cal content is the product that other news


6 sources do not offer. In theory, a newspaper only has to invest its employees time into posting extra resources online wi th the local stories that ra n in the print edition. This creates a situation where, as Shoemaker (1991) said, cost becomes value (p. 23). This study compares the content of storie s in selected newspa pers with their Web counterparts and examines whether locally wr itten stories are more likely than non-local stories to have additional content. Additiona l or different content could include photos, graphics, or multimedia or interactive components. The results of this study should offer a picture of how industry-leading newspapers use their Web sites and whether users can find unique information there. The results ha ve implications for models of newspaper Web sites and discussions about paid content online, so it is im portant to examine whether print and Web editions of newspapers tell stories different lywhether they are complementary or competitive. This effo rt focuses on storytelling components and should provide a clearer picture of any relati onship between newspapers practice of print and online gatekeeping. This study continues in Chapter 2 with a review of relevant literature about the areas of gatekeeping, online newspapers, and local news. Chapter 3 describes the method and reliability of this content analysis, Chapter 4 provides results from the study, and Chapter 5 includes conclusions, limitations and suggestions for future research.


7 Chapter 2: Literature Review This study draws from work in several areas. Traditional gate keeping literature examines why particular news items make it to the public. Studies about online news explore unique features of Web delivery and how it matches up with older media. The intersection of these areas provide s excellent context for this study. Gatekeeping Gatekeeping has a long history as a mass communication theory, and the definitions of who or what can be a gatekeep er and the scope of the gatekeeping process have shifted through the year s. The concept of information gatekeeping began with Lewin (1947), who described forc es that acted on the select ion of groceries and produce as they moved through channels to the family dinner table. He said decisions about food fall to persons in key positions (p. 143) who must weigh opposing forces, such as the cost and desirability of particular types of foods, and select what to bring home. A gate was any area in the channel where the for ces changed enough to make that a decisive point for making it through the channel, and Lewin said gatekeepers or impartial rules govern the gates. The first study that applied Lewins gatekeeping concept to mass communication was Whites (1950), and his appr oach established a focus on th e individual as gatekeeper in mass communication literature. In Whites study, a mid-career wire editor at the Peoria Star (Reese & Ballinger, 2001) saved all storie s that he rejected from three wire services in one week and recorded why he di d not select the storie s to run. White found


8 that Mr. Gates was able to run about one -tenth of the 12,000 column inches of copy he saw and tended to choose stories written conservatively that came across early in the evening. The study noted a gatekeeping eff ect both from individual preferences and constraint by the organization s production process. If Mr. Gates was representative of other wire editors, White concluded that the community shall hear as a fact only those events which the newsman, as the representative of his culture, believ es to be true (p. 390). Snider (1967) repeated Wh ites study with the same wi re editor 17 years later to see whether time tempered Mr. Gates approach to news selection. Mr. Gates remained in the same job at the same newspaper, but his situation changed somewhat in the intervening time. The citys evening news paper took control of the morning paper Mr. Gates worked for, and the paper had a smaller news hole because it had increased advertising. The paper used only one wire serv ice in 1966 compared to the three used in 1949. Mr. Gates newspaper had to compete with radio and television news, which Snider said had not been as competi tive 17 years previously. Snider studied Mr. Gates for five days instead of the week Wh ite (1950) did and found that Mr. Gates chose about a third of nearly 2,000 column inches of wire type he saw in 1966. Snider said the wire editor chose more international war news than a ny other category of story in 1966, perhaps reflecting the ongoing Vietnam War, compared to a preference for human interest stories in 1949. Mr. Gates still chose the stories he liked and thought his r eaders wanted, Snider said, adding that the wire service offered a nd Mr. Gates chose a better balanced news diet (p. 424) in 1966 th an they did in 1949.


9 Several other early studies of gatekeepi ng focused on wire editors and wire stories (Gieber, 1956; Jones, Troldahl & Hvistendahl, 1961). Instea d of creating a case study of one editor, Jones, Troldahl, and Hvistendahl (1961) analyzed the state wire content provided to and run in 23 Mi nnesota newspapers. They foun d that the newspapers used about one-third of the state content provided by the Minneapolis bureau of the Associated Press and that stories provided just before the newspapers d eadlines were unlikely to see print, a finding similar to White s (1950) gatekeepers approa ch to filling his newspaper. Contrary to the Mr. Gates studies, Giebe rs (1956) study of 16 telegraph editors found that their personal opinions about events and people in the news had no effect on their selection of newspaper stories. The results of interviews, mock story selection, and analysis of spiked copy in Giebers study s howed a passive group of gatekeepers, each of whom was caught in a strait jacket of mech anical details (p. 432). He said the driving force for these editors was simply filling the news hole. Giebers (1956) study aligns with the a pproach to gatekeeping that Westley and MacLean (1957) took in their communication model. In addition to parties A and B passing information through channels, they creat ed C as an intermediary and designated that as a gatekeeping position. Those gatekeep ers survive to the extent that they satisfy [audience] needs (p. 34). This model uses entire organizations as gatekeepers rather than individual journali sts, who are merely interchangeable cogs in the media machine (Shoemaker, 1991) in this understand ing of gatekeeping. Giebers approach also de-emphasizes the importance of th e individual gatekeeper by describing gatekeeping as a mechanistic process governed more by the norms and routines of news production than the purposive decision of individuals.


10 The Mr. Gates study became a model fo r many gatekeeping st udies by putting the individual gatekeeper at center stage and looking at how th e process worked around him. Some scholars (Bass, 1969; Brown, 1979), how ever, criticized Whites (1950) emphasis on the individual because of what it left out of Lewins (1947) in itial description of gatekeeping. Bass (1969) examined the staff structure of United Nations Radio, where stories came to a central news desk and then were sent to various language departments for translation and broadcast. He said st udying a wire editor as a gatekeeper was analogous to studying one of the language editors at UN Radi o, and he argued that the news desk made coverage decisions, while wire editors and language editors processed stories. Bass divided the news flow process into two parts: news gathering, which would include reporters and their editors, and news processing, which would include other editors, copyreaders, and translators. Bass said significant news decisions happened at the news gathering level, a contention that differed from Whites, who saw any decisionmaker as a gatekeeper. This disagreement ove r terms formed the greater part of Basss discussion of gatekeeping, but he also argue d that White pared aw ay a bit of Lewins gatekeeping concept to make it applicable to mass communication instead of small-group communication. Lewin (1947) initially said that gatekeeping holds for the traveling of a news item through certain communication channels in a group (p. 145), and Bass (1969) pointed out that Whites omission of the phrase in a group changed a group dynamics concept. Although Bass suggested that Lewins gatekeeping concept might more accurately be applied to a familys consumption of news rather than an organizations production of it, Bass did nothing to change Whites original application of the term in his own analysis of gatekeeping.


11 Brown (1979), however, said Whites (1950) analysis of stories rejected by Mr. Gates missed Lewins statement that gatekeep ing applies to some, not all, information. Brown said this ignored the forces Lewin described in his original conception, so Brown examined the stories included about one topic in national news mag azines over the course of 30 years. He looked for correlations be tween Census data and the frequency of coverage of family planning, and Brown found that gatekeeping decisions mirrored societal perceptions. This finding would seem to indicate that gate keepers are in tune with public opinion. However, Sasser and Russell (1972) concluded that there is no such thing as news of the day important to the public or at the least, editors did not have the training to recognize what the public judged to be important stories. Their study of several newspapers and television st ations found that media organi zations consistently covered major news events but in gene ral shared few topics. Stempel (1985) also found that when stories were divided into br oad subject categoriessuch as politics and government, crime, and general human interestnewspap ers and television stations selected approximately equal mixes of the types, al though they varied widely in which stories actually were selected. An an alysis of Whites (1950) study suggested that the number of stories the wire services released in each cat egory of news influenced Mr. Gates to select roughly the same mix (McCombs & Shaw, 1976) Dimmick (1974) also suggested that gatekeepers were uncertain about what stories to choose. These findings support Giebers (1964) cont ention that news is subjective to the gatekeeper, and Adhikari, Everbach, and Fahm y (2002) also observed little to suggest that editors are detached and objective professionals (Significance and discussion


12 section, ). Whether su bjectivity is a positive or negativ e thing, it is a theme of many studies of individual gatekeepers dating back to Whites (1950) Mr. Gates, who rejected several stories as B.S. and wrote on another, Dont care for suicides. Gatekeeping studies later expanded to l ook at the roles of groups, organizations, and routines in gatekeeping decisions rather than the influence of the individual gatekeeper only (Berkowitz, 1990; Berkow itz, 1991; Dimmick, 1974; Donohue, Olien, & Tichenor, 1989; Shoemaker, 1991; Shoemake r, Eichholz, Kim, & Wrigley, 2001). None suggested that the individuals were unimporta nt in the process, but rather that the structural context within which individua ls operate affects gatekeeping decisions (Donohue, Olien, & Tichenor, 1989, p. 807). That can happen when an organization reinforces a gatekeepers ow n values (Dimmick, 1974), issu es such as deadlines and space constraints impose (Donohue, Olien, & Tichenor, 1989; Shoemaker, 1991), or media from different platforms must decide what to do with the same story (Abbott & Brassfield, 1989; Epstein, 2000). Routine and organizational influences fr equently are discussed separately; both are part of the hierarchy of influences on media content created by Shoemaker and Reese (1996) that works its way outward from the individual level through media routines, organization, extramedia, and ideological levels. Shoemaker (1991) differentiates between routines and organi zational factors by including communication practices that are common across many communication organiza tions in the category of routines and factors that vary in the organizational leve l (p. 53). A good story that comes across the wire 10 minutes before a newspapers deadline is unlikely to get the same play it would have earlier in the day; the im mediate pressure to publish can constrain the level of detail


13 and context in a breaking story being pos ted to the Internet. Those time and space constraints occur at many organizations, and Shoemaker and Reese (1996) said routines help explain how content is shaped in response to those limits (p. 118). Organizational norms and policies also dict ate the amount of time, space, and other resources available for certain projec ts, but those vary among workplaces. Journalists produce news within orga nizations and bureaucracies, and this sociology of news has been of interest to a number of scholars (Hirsch, 1977; Reese & Ballinger, 2001; Tuchman, 1978) who built on the work of Breed (1955). Breed talked about policy as a more or less consistent orientation shown by a paper (p. 327) that never was directly communicated to new em ployees but was widely known. Journalists become socialized to their workplace and become subject to organizational level influences on content. When Hirsch (1977) re-evaluated Whites (1950) gatekeeping data, he argued that professional and organizational norms had far more effect on Mr. Gates decisions than any personal bias the wire editor brought Like McCombs and Shaw (1976), Hirsch pointed out the similarity in the overall mix of stories chosen to the mix of wire stories initially sent to Mr. Gates. Broadening definitions. Although research of routines as gatekeepers appeared more slowly than the work about individua ls, it fit with Lewins contention from the beginning that rules can act as impartial gatekeepers. Shoema ker et al. (2001) provide a useful definition of gatekeep ers that includes both people and processes: Gatekeepers are either the individuals or the sets of routine procedures that determine whether items pass through the gates (p. 235). The research ers performed a conten t analysis of the coverage of 50 Congressional b ills and then surveyed the newspaper reporters and editors


14 included in the sample of stories. Their study concludes that routines of news work shape content more significantly than individual influence. Shoemaker (1991) suggested that this could be all the more true at a la rge organization where many more people and processes might be acting as gatekeeper s than an organization like a community newspaper. Gatekeeping at a larger organization would depe nd more on its rules than the idiosyncratic logic of the individual (p. 56). This organizationa l level is important because it puts the people and pr actices in place that determine what gets past the gate and how it is presented (Shoemaker, 1991, p. 53). The idea that gatekeeping pl ays a role in news presenta tion, not just selection, is key in this study. Whereas Wh ites (1950) study and others focused on story selection, Donohue, Tichenor, and Olien (1972) broadene d information gatekeeping to include selection, shaping, display, ti ming, withholding, or repeti tion of entire messages or message components (p. 43). Shoemaker (1991) argued that items that made it through the gate often had a strong positive force, which meant they were likely to be packaged attractively, get good placement, and be repeate d. Stories with negativ e forces that still progressed through the gate were more likely to have unfavorable shaping, display, timing, or repetition (p. 25). The problem with focusing on selection alone, Reese and Ballinger (2001) said, was that the in-or-out decision ignored the effect of framing messages. Time and space constrain the structure, or framing, of a story, and these routine factors can act as gatekeeping forces. Framing. The pieces of information included in a story, their position in the story, pictures, infographics, or anyt hing that contributes to conten t can have framing effects. A frame organizes everyday reality, Tuch man (1978, p. 193) said, and it makes sense of


15 events that would otherwise have no contex t for many people. Reese (2001) defined it by including the interests, comm unicators, sources, and culture that combine to create an understanding of the world, which a person devel ops from all available verbal and visual symbolic resources (p. 11). Th is suggests that every piece of information included in a story compounds to create a diffe rent picture of an issue or event than if the newspaper had included different pieces of information. Even if all of the same pieces are included in two different presentations, the way they are structured affects the way a reader processes the news (Reese, 2001). Pavlik (2001) suggested that the Web, with its hypermedia and multimedia capabilities, presents many framing possibili ties that differ from traditional media. Hypermedia, or the abilit y to link among online objects, provides additional background, detail and, most importantly, co ntext (p. 316), and layering multimedia elements, such as audio and video files, can give extra content to many elements of a given story. Pavlik did not discount the role of the journalist as gate keeper of information related to a story, but he said the role of the journalist onl ine emphasizes inclusiveness of perspectives much more than in tradi tional journalism (p. 319). Perspectives for stories might include source documents, transc ripts and recordings of sources views, links to past stories on the i ssues, links directly to source information online, or other such items that might have been merely a reference for the journalists synthesis for traditional media. Television gatekeeping. Scholars applied gatekeeping theory to print media for years, then to radio (Bass, 1969), and fina lly to television news. Seeing how scholars adapted the theory among media provides inte resting context in a time when researchers


16 examine whether gatekeeping works on the Inte rnet. The question of local focus was the most important factor for both newspaper a nd television gatekeeper s in a study by Abbott and Brassfield (1989), but television gate keepers were more likely to worry about timeliness than newspaper editors and usually weighed visual storyt elling criteria more heavily. Berkowitz (1990) echoe d this finding and said tele vision gatekeeping was very much a group process rather than an indivi dual one. Harmon (1989) found that traditional news values applied to local television news, but even though stations were capable of enterprise reporting, their role more often was that of a repackager of news (p. 861). The focus shifted to production and technol ogical capability. The study did not address the quality of the content, but Livingston a nd Bennett (2003) said they could make no guarantees that technologies will not be us ed simply as glitz factors (p. 364). Online gatekeeping. Researchers have questioned the viability of gatekeeping theory in the new media era because of the va st amount of information available from so many sources, and most studies have conclude d that journalists role in information gatekeeping is not dying but evolving (Blake, 2004; Cassi dy, 2002; Singer, 2001; Singer, 2005). This concern about shifting roles seem s to point more st rongly at potential changes in agenda setting and agenda building, however. Williams and Delli Carpini (2004) heralded the collapse of gatekeepi ng (p. 1208), going against the tide who speak of a gatekeeping evolution. The researchers argued that mainstream mediawhich they viewed as a monolithic group rather than ma ny individual gatesno longer have control of story selection because they must follo w as a pack stories that come to public knowledge through alternative media, an issue that Wigley (2004) also discussed. Bill Clinton became a case study for the researcher s; Williams and Delli Carpini said that a


17 tabloid first reported the alle gations by Gennifer Flowers, that a trade magazine first wrote about Paula Jones, a nd that the Drudge Report br oke the story about Monica Lewinsky. These three situations persuaded Williams and Delli Carpini that mainstream journalism [had] lost its position as the centr al gatekeeper of the nations political agenda in a six-year period (p. 1225). From a broader view of gatekeeping, however, it seems that as long as journalists are selecting some content and rejecting other items, they have a gatekeeping role even in a media environment with more choices for the audience. The audience is of greater concern to Webmasters than to print gatekeepers, though, Beard and Olsen (1999) found. They studied Webmasters of college and university sites and found that the Webmasters previous work experience affected their focus or main goalsuch as editing, design, or communicationi n managing the site. These online gatekeepers also f aced constraints consistent with those mentioned in earlier gatekeeping research, such as inadequate resources, heavy workloads, degree of autonomy, and sharing gatekeeping responsib ilities. Beard and Olsen said these constraints, particularly la rge amounts of work, created a gatekeeping role limited to selecting and processing exis ting messages (p. 207) rather than creating news ones. Overall, the researchers qua litative study of eight Webmas ters concluded that these people had many of the same responsibilities, values, and constraints as traditional gatekeepers. Several researchers have compared print and online editions to examine gatekeeping effects (Blake 2004; Singer, 2001). Singers sample of six Colorado newspapers showed that despite the unlim ited news hold availabl e online, editors of


18 Web products were whittling down the print pa ckage for online distribution (p. 71); the study found more than twice as many stories in the print editions th an the papers posted online. Stories online typically were identical or had minor changes to the print versions, and although 48% of the print st ories ran with some art, only about 18% of stories had art online. The study revealed no daily news content made just for the site, but it showed that a greater proportion of online stories were about the papers metropolitan area than the amount in print editions. Of the news, sports and business stories coded, about 45% of online stories were metro items, and about 31% of print stories in the sample were metro items. Singer concluded that the Web, with its global potential, was becoming a local niche for online newspapers. A number of constr aints that could be c onsidered routine or organizational factors in the Shoemaker and Reese (1996) scheme we re listed. Most of the papers in the sample had few online staff members, and locally written stories represented content that the organization al ready had paid for by employing its reporters and editors. The newspapers would have ha d to pay more money to find or create additional content. Singer observed that journalists had a seeming willingness to abandon their traditional gatekeeping responsib ilities, (p. 78), which she said might be because they faced organizationa l constraints, reordered prio rities, or recognized the high level of personalization the Web offers its users. The contention that the Internet bl ows open the whole no tion of a gate remains in Singers (2005, p. 3) work, but she argued that journalists give credibility to information online and therefore retain a ga tekeeping role online. The study targeted editors of online political cont ent at some of the countrys largest newspapers during the 2004 presidential campaign, and 47 responded. Mo st content originated in the print


19 medium, but Singer found journalis ts increasingly likely to enhance the original content. Three ways that journalists stepped back fr om traditional gatekeeping roles included the ability for users to persona lize content, the presence of chats, message boards, or discussion forums, and the addition of blogsonline journals. The first two methods gave a great deal of freedom to users, and the last allowed journalists or local opinion leaders to analyze politics. These results indicate that journalists may see their gatekeeping role as a responsibil ity to offer tools for analysis and interact with readers. Singer said this interaction could save prof essional journalists from online irrelevancy. The interaction and vetting of information pr ovides an even more valuable gatekeeping role in todays rowdy, unbounded in formation environment (p. 24). Blake (2004) said gatekeeping was a va lid construct for evaluating different media with common content ownership. Ra ther than saying j ournalists abandoned gatekeeping roles online, Blake suggested that at least one gate existed to select news to appear online because his study of two papers found one that focused its content on global and national issues and one that, like Singer (2001) found, focused its content on local and state issues. In both cases, however stories appearing online were more likely to have been written by a staff member than some other source. Blake searched online for all stories appearing in the front and local sections of the newspapers and found about 78% on the Weba higher percentage but an even smaller sample than Singersbut almost all of the stories that had photographs in the print edition lacked them online. For the most part, researchers agree th at online gatekeeping is important but evolving. Many have concluded that quantit y is not quality, as Arant and Anderson (2000) said: If a news organi zation simply publishes everythi ng it can get its hands on in


20 its bottomless online news hole, is it covering an issue in a way that best serves its audience? (). Organizational constrains mi ght make that much content impossible to manage, but research has shown a vari ety of trends in online newspapers. Trends in online newspapers Several studies have suggested that onlin e newspapers are complementary to print editions, not competitive. People who go online to seek out information about politics, sports, business, science, inte rnational affairs, local news, and entertainment news are likely to continue to read about those niche interests in traditional media (Dutta-Bergman, 2004). Dutta-Bergman used survey data from more than 3,000 individuals and concluded that new media news is complementary to tr aditional media and that content is the critical ingredient in media choice (p. 58) Chyi and Lasorsa (2002) and Zaharopoulos (2003) also found that readers us ed both print and online editions of the same newspaper, making them complementary. However, the State of the News Media 2005 survey by the Project for Excellence in Journalism indicated that online news is beginning to cannibalize traditional media forms and that online news is likely to draw even more users away from newspapers in the future because Web users are far younger than newspaper readers. The same study shows that the percentage of regular user s of online news was up to 29% in 2004 from 23% in 2000. This growth would be enhanced more dramatically if newspapers would add more original content to the Web produc t, said Peng et al. (1999), adding that the result would be cost-effective operations. In many cases, though, newspapers have posted nothing more than what appears in their pr int editions to their Web sites (Gubman &


21 Greer, 1997; Martin, 1998; Neuberger, Tonnema cher, Biebl, & Duck, 1998; Peng et al., 1999; Tankard & Ban, 1998; Weiss, 2004). Shovelware. The derisive term shovelware de scribes print content shoveled up to dump online wholesale, without alterati on. Studies have found that many newspapers post about one-quarter to just over half of th e content they produce fo r their print editions (Arant & Anderson, 2000; Neuberger et al., 1998; Saksena & Hollifield, 2002). Neuberger et al. said a strong orientation to ward print content c ould be both a good and bad thing: if readers base their expectations of the online product on the print edition they are used to, then shovelware and similar or ganization can benefit a site. However, too much faithfulness to the print original m eans that unique Web opportunities go untapped. Content may receive a variety of treatm ents online, but Weiss (2004) found that 65% of stories added to news papers home pages had no contextual features added. Her content analysis of 20 newspaper Web sites l ooked at stories to see whether contextual features such as photos, related stories, section additions, hyperlinks, polls, forums, blogs, slideshows, video, audio, or maps had been uploaded to the site as well. Many stories are placed online with few modifications from the print version (Martin, 1998). A case study of two newspape rs showed that online staff members sometimes wrote new headlines to fit their space requirements and often had to modify photographs from the print versions in orde r to fit the online templates, similar to Singers (2001) findings. Martin said online staff members identified their primary job function as selecting and reformatting exis ting content, not creating new content or enhancing it.


22 Online staff members of other sites have said their sites have original content, but it is not in daily news sections (Neuberg er et al., 1998; Tank ard & Ban, 1998). About three-quarters of 135 online newspapers report ed posting content not in the print edition, but the content usually included evergreen features and community services, such as dining guides, tourism information, regiona l information, and some special project stories. In general, research shows that many newspaper Web sites are posting stories without significant ed iting or additions to their Web sites, and many have questioned whether that hurts the newspaper in the l ong run by creating a competitor and whether it ignores some of the Internets technical possibilities. Not all online editors thought any kind of changes had to occur: Good print copy makes for good online copy, one said in a survey by Arant and Anderson (2000). This kind of edition-blind evaluation to justify shovelware overlooks some of the documented differences about the way users process information on a monitor versus on paper (Poynter Institute, 2004). Interactivity and multimedia use. Interactivity has been called the Internets outstanding characteristic (Mor ris, 2001), but researchers have found mixed results about its use on newspaper Web sites. Lowrey (2003) performed a census of Mississippi newspapers online and created a degree of site interactivity variable by totaling occurrences of interactive features on the 48 sites. Lowrey found a mean 3.67 occurrences of e-mail links from stories, list of staff e-mails, e-mail links to Webmasters, other contact information, reader comments pos ted from stories, chat rooms, bulletin boards, ability to e-mail stories to others polls, search functions, and hyperlinks to supplemental information. The data suggested that larger newspapers sites are more


23 likely to be interactive, a finding echoed about many features unique to the Web (Greer & Mensing, 2004; Tankard & Ban, 1998; Weiss, 2004; Zaharopoulos, 2003). Some scholars also consider multimedia applications to have interactive properties, and its use also has been ex amined on newspaper sites. A multimedia presentation may use audio, video, graphics, or other methods to give users a more complete picture of a story. From a different perspective, Sundar (2000) called multimedia a misleading term because it doesnt refer to multiple media but multiple senses involved in processing a message or multiple channels used to transmit it. The multiple channels might include animati on, audio, video (Gubman & Greer, 1997), or pictures (Sundar, 2000). Multimedia gives read ers the sense that they can control how they experience a story, and it might also cause them to rate a site as more professional (Sundar, 2000). The sensory experience provided by the Internet is similar to the television, which provides visual and audito ry stimulation (Welch, 2004). The findings of Sundars experiment, though, suggest that multimedia applications actually hinder memory for story content. Gubman and Greer (1997) defined multimedia as animation, audio, and video, and their content analysis found only 12 sites us ing any of those in news sections of 83 online newspapers, and Massey (2 000) found it even more rare He dropped analysis of multimedia in analysis of Asian news sites because he found no occurrence of any applications. Kamerer and Bressers (1998) found few uses of audio and video and suggested that the technological limitations of Web access still made these features impractical because of long loading times and inconsistent buffering.


24 Advances in time and technology may have shifted the picture about multimedia use. Dibean and Garrison (2001) found 30 perc ent of site pages for six newspapers offering video in 1999 and 27 percent offering audio. The sample distributiontwo small circulation, two mid-sized, and two large circulation newspapersprobably skewed these findings, though, since Schultz (1999) found that larger newspapers were more likely to offer multimedia applications. Greer and Mensing (2004) performed a longitudinal analysis of more than 80 online newspapers beginning with the data collected in the Gubman and Greer (1997) study. Only a handful were using audio and video content in 1997, but nearly half were doing so by 2003. A great deal of this increase was in the amount of multimediaaudio, video, or animationused in news stories; in the early years of the study, most multimedia use was in advertisements. Schultz (1999) described multimedia and interactivity as separate spheres, but the presence of one on a news site often means the site is likely to use the other to tell stories. Schultz found 16% of newspaper sites using multimedia applications, as well as a correlation between that group and the us e of interactive functions, which he operationalized as the presence of discussion forums. Of sites that used multimedia applications, 69% had forums, and only 26% who lacked multimedia on their sites hosted forums. Schultz also created an index of in teractive options by counting general e-mail addresses, e-mail links from particular storie s, e-mail links to politicians and officials, discussion forums, chat rooms, polls, and online letters to the editor. Online newspapers with multimedia had a mean score of 5.88 on th e index, whereas others had an average of 3.74. The difference was statistically significant.


25 Another study counted streaming vide o, streaming audio, photos, animated graphics, static graphics, slideshows, and photo galleries as multimedia components, and Welch (2004) found news Web sites using all of them except animated graphics. Photos were by far the most frequently used mu ltimedia component. Welch also found sites using interactive elements including polls and surveys, discussion forums, quizzes, search tools, e-mail capability, related Web links, fee dback forms, and several other tools. The ability to e-mail articles to others and the pr ovision of related Web links were the most frequently used interactive tools. Local news online Newspapers divide their page s into sections to comply with press constraints, and these divisions often fall by types of conten t; the front section contains national and international news, another section contains local news, and still others might include sports, business, and features. All sections are likely to include a mix of staff-written and wire stories. The papers Web site might cont ain these types of stor ies in a structure and organization that mimics the newspaper, but several studies indicate that local news is handled differently on the Web than it is in print editions. Local news is one of the most common types of information on newspaper sites, Greer and Mensing (2004) found; 95% of 81 online newspapers sampled included it in 2003, up slightly from about 90% when the longitudinal study was first begun in 1997. Zaharopoulos (2003) found most of the stories posted on newspaper sites home pages to be local news stories, and several editors repor ted that they publish all or most of their papers locally written stor ies online (Arant & Anderson, 2000). Martins (1998) case study found that the Raleigh News & Observer could post only stories written by its


26 reporters about local or state news. This preference for local stories also emerged in other studies, such as one by Chyi and Sylvie (2001). Their survey revealed that more users of online newspapers are local rather than l ong-distance and that the local audience was precisely the one newspapers aimed for. Blake (2004) examined gatekeeping be tween print and online editions of newspapers and concluded that t he gate from the print to the online format is primarily a geographic filter that allows much local and state content to appear while serving as a barrier to less-local news be ing redistributed online (Discu ssion section, ). More state and metropolitan stories were available onlin e than initially expected at the larger newspaper, and this paper also showed a st rong preference for staffwritten content over stories from other sources. Blake found the smaller paper had more global and national stories than expected, and a lthough it selected more non-staff articles for its Web site than locally written articles, the site relied on non-staff content less heavily than the print edition. The study showed that newspaper si tes consistently ch ose content based on geography whether locally or non-locally oriented. Singer (2001) found a significant differe nce between the percentage of metro stories in online editions a nd print editions. In the samp le of six Colorado papers representing a variety of circulation sizes, about 45% of all stories online were metro stories, and just 31% percent of all print it ems were metro stories. The differences were particularly sharp between the two mid-sized papers represented; the Colorado Springs Gazette had about 34% metro stories in pr int and about 55% online, and the Pueblo Chieftain had about 28% metro storie s in print and nearly 68% online. Singer pointed out that nearly 79% of stories th at appeared in print only were from outside each papers


27 circulation area. The affiliation of reporters for the stories in both print and online editions shows a split even more dramatic than the local percentages: staff writers wrote or contributed to 59% of stories on the Web s ites, and they wrote just under 36% of the stories in the print sample. The inclusion of photography and other ar twork was another part of the Colorado study because Singer (2001) said a photograph or infographic tells a story in its own right and is worth inclusion in any discussion of the relative emphasis given to particular types of newspaper content (p. 76). Broadening the discussion about content to the components that make up a message allows cons ideration of gatekeeping from more than the selection point of view. Singer found a significant difference between the art used in print and online. The term art had broad a pplication in Singers study; this category included not only photographs, but infographics and logos, as well. Out of more than 3,400 stories in the print sample, 48% had some form of art, whether it was a simple identifying graphic or multi-photo package. The online sample had nearly 1,400 stories, and just 18% of those had an art element e ither in the story, on the home page menu, or as standalone art. If location is indeed a gate for stories between print and online editions, it seems worthwhile to examine whether it applie s for more than story selection. Donohue, Tichenor, and Olien (1972) included shapi ng, display, withholding, or repetition of message components in their definition of in formation gatekeeping, and these elements can be compared between local and non-local stories in print and online editions. The contextual elements that support and run alongs ide the text of a stor y in print and online formats are the focus of this study. Several studies (Blake, 2004; Singer, 2001) have


28 compared print and online newspapers, but th ey focused on story selection. This study focuses not on selection, but on content presentation. RQ1: Does story content differ between pr int and Web versions of daily stories? RQ2: Do contextual elements differ between print and Web versions of daily stories? RQ3: Is there a difference in the amo unt of contextual el ements published with stories by staff reporters and st ories from other sources? RQ4: Is there a difference in the amount of contextual elem ents published with stories focused on local issues compared with stories about state, national, or international issues?


29 Chapter 3: Method This study compared print and Web versions of stories through a content analysis of stories appearing on the front pages and me tro section front pages of five newspapers. The first two research questionDo content and contextual elements differ between print and Web versions of daily stor ies?required the comparison of stories that appeared in both editions. The researcher chose to analy ze stories on section front pages to combat previous findings that newspapers post to the Web only some of their stories published in print (Peng et al., 1999; Regan, 1995; Singer, 2001). Using both the front page, which carries international and national news as well as important local stories, and the metro section front page, which focuses on local news by definition, allowed examination of the final two research questions: Is there a diffe rence in the amount of contextual elements published with staff-written stories and storie s about the metropolitan area compared with stories by other sources and about other geographic areas? Studyi ng both the print and online editions not only allowed comparison of content stan dards between the two types of media but also whether each medium alone treated local stories differently than national and international ones. The papers selected for this study have the highest circulation numbersall greater than 425,000 on Sundaysof papers in th e southern United States. These papers are the Houston Chronicle the Dallas Morning News the Atlanta Journal-Constitution the St. Petersburg Times and the Miami Herald All have circulation numbers placing them in the top 25 largest newspapers in the nation, according the Audit Bureau of


30 Circulations (2005a), except for the Dallas Morning News which is under censure by the agency (Audit Bureau of Circulations, 2005c). The bureau censured the Dallas paper in Ju ly 2004 for misreporting its circulation figures (Audit Bureau of Circulations 2005c), and as a consequence, the Morning News was excluded from such lists while it underw ent audits every six months for two years (Audit Bureau of Circulations, 2004, 2005b). Ho wever, the bureaus most recent audit placed the Sunday circulation of the Dallas Morning News at 655,809, a figure second in the list of southern publications only to the circulation of the Houston Chronicle (Audit Bureau of Circulations, 2005a, 2005c). The Houston Chronicle is a Hearst-owned newspaper in Texas with Sunday circulation of 720,711, according to the Audit Bur eau of Circulations (2005a). It has a partnership agreement with KHOU that allows the Chronicle to use the television stations video on its Web site (American Press Institute, 2005), and its Web address is Like the Chronicle, each newspaper site was unique to the publication and not shared with any pa rtners in a portal format. Belo owns the Dallas Morning News in Texas, and the paper has a partnership with a television station, We b site, cable television network, Spanish-language newspaper, and a youth-oriented newspape r (American Press Institute, 2005). Its Web address is The Atlanta Journal-Constitution in Georgia has a S unday circulation of 610,338 and is owned by Cox Newspapers. The newspaper does not have any convergence partnerships, and its Web address is


31 The St. Petersburg Times in Florida has a Sunday circulation of 432,231 and is owned by the Times Publishing Co. The paper has a cross-promotional relationship with television station WTSP (American Press Institute, 2005), and its Web address is Knight-Ridder owns the Miami Herald in Florida; the paper has a Sunday circulation of 429,697. The Herald s partnership with a television station, radio station, and cable television station provide the newspaper with additional video and audio components for its Web site (American Pre ss Institute, 2005). The papers Web address is Several studies (Schultz, 1999; Greer & Mensing, 2004) have found that larger newspapers have more sophisticated Web s ites that include more photos and multimedia elements than papers with smaller circulati ons. Looking only at large newspapers in this study limits potential differences between site s of papers that have varying resources based on their circulation sizes. The papers are all major metropolitan dailies from one quadrant of the country. Constructed week sampling was used for the content analysis because this method has produced more reliable results than ra ndom sampling and consecutive days (Jones & Carter, 1959; Riffe, Aust, & Lacy, 1993; Stempe l, 1952). This type of sampling creates a composite week over a period of time so that each day of the week is represented, but the selection of that day is random. This me thod prevents oversampling of days with unusually large newspapers, such as Sundays which can happen with random sampling. Constructed weeks are superior to consecuti ve weeks because the results are easier to generalize over time (Riffe et al., 1993; Wimmer & Dominick, 2003). For this study, the


32 researcher created two composite weeks fr om the days between September 18, 2005, and October 15, 2005, a period devoid of major ho lidays or other pla nned and nationally significant events. The research er took the four Sundays that occurred in the month and randomly selected two of them. The process wa s repeated for the other six days of the week, for a total of 14 days that included each day of the week twice. Both Stempel (1952) and Riffe et al. (1993) sa id precision increases only sli ghtly with sample size in newspaper content analysis, and the faint marg in of improvement is often not enough to outweigh the extra resources require d to double or triple a sample. Content analysis online has special challe nges because of the fleeting nature of content on the Web and because the structure of Web pages can defy analysis strategies used with traditional media (McMillan, 2000). McMillan concluded that content analysis is perfectly valid to apply to the moving target of the Web (p. 93). Weiss (2004) suggested that researchers c ontinue to explore the use a nd standards of online content analysis. Singer (2001) warned that the Web can be a bear for researchers using content analysis, a method whose reliability and thus cr edibility rest primarily on the fact that the content is stable and the classification is reproducible (p. 70). This cautionary note is particularly true for online newspapers; their news content is updated at least daily, if not more often. Even if researchers take screen snapshots of Web pages, they often have a few hoursor, at best, a few daysto examine the content in the c ontext of the site before it disappears forever. These factor s make careful coding of online content a priority because it ofte n cannot be replicated. Individual stories were the unit of analys is in this study. The content analysis was carried out based on the stories on the print ed itions front pages and metro section front


33 pages of the five newspapers in the sample. The bias toward print content was necessary in this study from a gatekeeping point of view. Blake (2004) made a similar decision because online-only content never passed through a print gate before its publication to the Web. This study looked for differences between print and online versions of stories, which meant that stories had to be in both edi tions; the researcher used the print editions to form a directory of stories to find online because the print paper almost always serves as the content provider for the online edition (Chyi & Lasorsa, 2002). Newspapers have judged stories on the front page and front of the metro sectio n to be the most relevant, newsworthy items for their readers that day. St ories inside the two s ections and in other sections of the print edition were excluded in both the print and online analyses. For each newspaper each day, the researcher examined stories on the front and metro front pages and the corresponding stor ies in real time online (see Appendix A). The final step was to compare the two versi ons. Print stories were coded for newspaper; reporters affiliation, which could include sta ff, wire or news service, contributor, unknown, or some combination of those that migh t be noted in a storys byline or footer; geographic emphasisinterna tional, national, state, or metro; and elements accompanying the story. The elements c ould include photos, sidebars, infoboxes, infographics, artwork, refers to additional content in print or online, pull quotes, and pull outs of portions of the story (see Appendix B). The researcher began online data collecti on each day at 11 a.m. beginning with the Houston Chronicle and moving to the smallest paper, usually completed by 2 p.m. Collecting the stories online at the same ti me each day avoided potential problems with any updating deadlines each site met each day. The researcher visually scanned home


34 pages for each newspaper and the top pages of major news sections to look for the days stories online. If the visual search turned up nothing, the rese archer used the online search function to look for the headline. If both search methods failed, the story was classified as print only. Once the story was found online, th e researcher coded whether it was the same version of the story that had run in print, whether it had the same headline, and what elements accompanied the story. Like the print edition, this could include photos and infoboxes, but it also might include features unique to the Web: video, audio, source documents, polls, quizzes, animated graphics, st atic graphics, slideshows, photo galleries, related story links, relate d Web links, and live chats (Welch, 2004; Weiss, 2004). Discussion forums and blogs were counted onl y if they focused on the story in question and did not serve as an intera ctive tool for the whole secti on or site (see Appendix C). After looking at the print and online ve rsions of the stories separately, the researcher put them side-by-side to co mpare leads and contextual elements. The researcher noted whether the first paragraph of each story was the same or different in its two versions and compared whether any contextual elements were repeated in both editions. Elements that could overlap editions included photos, si debars, infographics, infoboxes, artwork, refers, pull quotes, and pull outs. The researcher coded all 14 days of cont ent for analysis. However, two additional coders examined two days worth of data, or about 14% of the total sample, in order to determine the reliability of the data. The additi onal coders analyzed the print editions of the newspapers and screen shots of the on line versions of stories collected by the researcher because the stories pages were no longer current when the additional coders evaluated the material.


35 The statistical program SPSS, version 13.0, was used for data analysis. Most of the data in this study was collected at the nomin al level, so a great deal of the analysis consisted of frequency calculations and chi-square measures of significance. However, some questions about data collected at the ra tio level were best answered by parametric statistics, including t -tests and ANOVA. To establish a measure of reliability for the content analysis, two additional coders evaluated a portion of the data. The researcher code d all 14 days of newspapers, and the additional coders both evaluated the fi rst two days of the sample. The additional coders used regular print editions of each newspaper coded, but they had to use screen shots of each story onlin e rather than seeing it on a live We b site. They were able to see what each story looked like on the day it was posted, but they could not follow links or make any Web features active. The first two days of the sample included 88 stories, of which all coders entered data on 83 stories, or 13% of the entire sample of 635 stories. Intercoder reliability was measured on the print element variable because of the large amount of data included in it. Holstis formula produced a reliability coeffici ent of .85 with Coder 1 and .90 with Coder 2. Holstis formula is two times the number of cases on which coders agree divided by total number of coding decisions. Although the fo rmula has been criticized for failing to account for the role of chance in code r agreement (Wimmer & Dominick, 2003), Neuendorf (2002) said a Holstis reliability coefficient of .90 or greate r is almost always acceptable, and that a coefficient of .80 or greater is acceptable in most situations.


36 Chapter 4: Results The 14-day sample yielded 635 stories on the front and metro section front pages of the five newspapers, and versions of 612 of the stories, or 96.4% of the total sample, also appeared on the newspapers Web sites. The number of print stories was distributed evenly across the newspapers, as shown in Table 1, and nearly equal numbers of stories appeared on the front pages of both sections : 49.1% (n=312) on the front page and 50.9% (n=323) on the metro section front page. Table 1. Distribution of stories by newspaper Newspaper Frequency Percent Houston Chronicle 121 19.1% Dallas Morning News 133 20.9% Atlanta Journal-Constitution 124 19.5% St. Petersburg Times 125 19.7% Miami Herald 132 20.8% Total 635 100% RQ1 The first research question asked whether story content differed between print and online versions of stories. The operational measure for this question was a comparison of a storys lead be tween its print and online versions. Of the 612 stories that appeared both in print and online, 96.1% had the same lead word-for-word in both media, a significant finding, according to a chi-square goodness-of-fit test, X 2 (1, N=612)= 519.77, p < .001. Chi-square goodness-of -fit test compares observed frequencies to expected frequencies, determining whether distribution among categor ies is significant, or likely to have occurred by ch ance (Wimmer & Dominick, 2003).


37 The remaining stories had differences in the first paragraph that could have included rewriting, updating, or us ing a different version of a story. A small percentage of stories differed in content between media. Although the text of stories was substantially the same in all but a few cas es, a larger percentage ran under different headlines in print and online. Stories onlin e had different headlines than their print versions 24% (n=147) of the time. A chi-s quare goodness-of-fit test also indicated this finding to be significant, X 2 (1, N=612) = 165.24, p < .001. RQ2 The second research question asked whether contextual elements differed between the two media. Of the 612 stories that appeared both in print and online, 335 stories had contextual elements in both media. In 203 of those cases, or 32% of the entire sample, the stories shared at least one elemen t between media, whether it was a picture, infobox, infographic, or other contextual element. The remaining 132 stories (20.8%) had unique elements in each medium, whether it wa s a different photo or other traditional element or the use of online-only elements th at newspapers could not reproduce in print. A chi-square goodness-of-fit te st indicated that the re sults were significant, X 2 (2, N=612) = 51.54, p < .001. Table 2. Comparison of contextual elem ents between print and online stories Element comparison Frequency Percentage* At least one version lacki ng contextual elements 277 43.6% At least one contextual element in common 203 32.0% All contextual elements different 132 20.8% Story not online 23 3.6% Total 635 100.0% X 2 (2, N=612) = 51.54, p < .001


38 Of the 32% of stories that had elemen ts in common between versions, five of eight types of elements were repeated betw een versions. Table 3 breaks down the number of contextual elements that stories had in common between print and online editions. Because some stories had more than one element in common, percentages do not add up to 100% across elements. Table 3. Elements in common between print and online stories Element Number per story Frequency Percentage X 2 value 0 516 81.3% 1 84 13.2% 2 8 1.3% 3 1 0.2% 4 1 0.2% 5 2 0.3% Story not online 23 3.6% Photo Total 635 100.0% 2068.22* df=5 0 510 80.3% 1 90 14.2% 2 12 1.9% Story not online 23 3.6% Infobox Total 635 100% 703.41* df=2 0 597 94.0% 1 15 2.4% Story not online 23 3.6% Infographic Total 635 100.0% 553.47* df=2 0 604 95.1% 1 8 1.3% Story not online 23 3.6% Refer Total 635 100.0% 580.42* df=1 0 609 95.9% 1 3 0.5% Story not online 23 3.6% Sidebar Total 635 100.0% 600.06* df=1 p < .001


39 Three of eight print elements never were in common with online elements: artwork, pull quotes, and pull outs. These el ements also were not among the most frequently used in print editions; photos and infoboxes were the most common contextual elements with print stories. More than 85% (n =545) of stories in th e sample had at least one contextual element in print, which is a significant finding, according to a chi-square goodness-of-fit test, X 2 (1, N=635) = 326.02, p > .001. Table 4 shows the eight categories of c ontextual elements for print and shows how often one or more examples of each ki nd of element appeared in print stories. Table 4. Number of stories with type s of print elements present (n=635) Element Frequency Percentage* Photo 433 68.2% Infobox 164 25.8% Refer 120 18.9% Infographic 93 14.6% Sidebar 53 8.3% Pull quote 53 8.3% Pull out 7 1.1% Note Some stories had multiple types of el ements, so table does not total 100%. p > .001 Fewer stories online had contextual el ements; 58% (n=355) of the 612 stories online had one or more contextual elements. Table 4 shows the 17 categories of contextual elements for online stories and s hows how often one or more examples of a particular kind of element appeared with a story.


40 Table 5. Number of stories with type s of online elements present (n=612) Element Frequency Percentage* Photo 142 23.2% Related Web links 120 19.6% Infobox 119 19.4% Related story links 102 16.7% Video 76 12.4% Photo gallery 73 11.9% Static graphic 69 11.2% Interactive graphic 44 7.2% Document 34 5.6% Poll 35 5.6% Audio 23 3.8% Blog 22 3.4% Discussion forum 17 2.8% Other 15 2.5% Live chat 8 1.3% Quiz 2 0.3% Slideshow 2 0.3% Note Some stories had multiple kinds of elements, so table does not total 100%. p > .001 The number of contextual elements with each story in print was related to the storys section. An independent groups t -test, which compares the means of two unrelated groups (Weaver, 1989), revealed that a greater number of contextual elements appeared with front-page stories (M=2.82, SD=2.23) th an with metro section stories (M=1.60, SD=1.39), t (df)=633, p < .001. The same trend held true for elements that ran in both print and online versions of a story. A great er number of common contextual elements ran in print and online with stories that ran on the prin t front page (M=0.48, SD=0.73) than with stories that ran on the print metro section (M=0.36, SD=0.68), t (df)=610, p = .034. Front-page placement made it more likely for a story to have a greater number of contextual elements.


41 More print stories have contextual elemen ts than the same stories online. Slightly more than half (58%) of the stories had elem ents in both media, and about a third of the stories repeated the same elements in both print and online ve rsions. Although the response to RQ2 does not provide overwhelming support, the results showed that a fair portion of stories offered unique features in both media. RQ3 The third research question asked whether the amount of contextual elements published with stories by staff writers differed from stories by other sources. Overall, data analysis did not reveal a ny significant relationships between reporter affiliation and the amount of contextual el ements published with the reporters story. Stories written by staff member s alone far outnumbered stories from any other source, as Tables 6 and 7 show. When stories where st aff writers collaborated with other sources are also considered, more than 90% of storie s on the front and metr o section front pages carried the reporting of a staff writer, but th e distinctions were necessary because in a number of cases, a staff contribution to a st ory seemed to amount to a few sentences or paragraphs inserted to locali ze a story from another source. One-way analysis of variance compares three or more means at a time involving one independent variable (Weaver, 1989), and the researcher used this te st to evaluate the mean frequency of contextual elements by seve n categories of reporter affiliation in print stories, which are shown in Table 5. ANOVA showed no signi ficant relationship between print reporter affiliation and frequenc y of print elements, F(6, 628) = 1.54, p = 0.162.


42 Table 6. Print reporter affiliation and mean print elements Reporter affiliation Print story frequency Print story percentage* Mean elements per story Staff 545 85.8% 2.16 Wire 49 7.7% 2.06 Staff & wire 33 5.2% 2.88 Contributor 3 0.5% 3.00 Unknown 1 0.2% 6.00 Staff & contributor 3 0.5% 2.33 Staff & unknown 1 0.2% 1.00 Total 635 100.0% 2.20 X 2 (6, N=635) = 2677.98, p < .001 ANOVA also showed no significant re lationship between online reporter affiliation and frequency of online photos, F(6, 605) = 1.04, p = 0.396, or frequency of online infoboxes, F(6, 605) = 0.69, p = 0.657, shown in Table 6. Table 7. Online reporter affiliation and mean frequency of photos and infoboxes Reporter affiliation Online story frequency Online story percentage* Mean photos per story Mean infoboxes per story Staff 526 85.9% 0.29 0.24 Wire 46 7.5% 0.24 0.15 Staff & wire 31 5.1% 0.10 0.16 Contributor 4 0.7% 0.75 0.50 Unknown 1 0.2% 0.00 0.00 Staff & contributor 3 0.5% 0.33 0.00 Staff & unknown 1 0.2% 0.00 0.00 Total 612 100.0% 0.28 0.23 X 2 (6, N=612) = 2588.10, p < .001 The data also allowed analysis of the types of contextual elements with each story; no print story had more than six of ei ght types of elements, and no online story had more than nine of 17 types of elements. The mean number of types of elements per story


43 (see Table 8) also had no significant relationship with print reporter affiliation, F(6, 628) = 0.970, p = 0.444, or online reporter affiliation, F(6, 605) = 0.576, p = 0.750. Table 8. Online reporter affiliation and mean types of elements represented Reporter affiliation Print story frequency Mean types of print elements Online story frequency Mean types of online elements Staff 545 1.45 526 1.54 Wire 49 1.51 46 1.15 Staff & wire 33 1.79 31 1.10 Contributor 3 1.67 4 1.00 Unknown 1 1.00 1 1.00 Staff & contributor 3 2.33 3 1.00 Staff & unknown 1 1.00 1 0.00 Total 635 1.48 612 1.48 RQ4 The fourth research question aske d whether the amount of contextual elements published with stories about metroarea news differed from stories about other geographic areas. Stories about each newspapers metro area had the lowest mean number of contextual elements in print, as Table 9 shows, and ANOVA revealed a significant difference in the means, F(3, 631) = 2.86, p = .036.


44 Table 9. Story geographic emphasis and mean frequency of print elements Geographic focal area Print story frequency Percentage* of print stories Mean elements per story International 47 7.4% 2.66 National 115 18.1% 2.54 State 126 19.8% 2.15 Metro 347 54.6% 2.04 Total 635 100.0% 2.20 X 2 (3, N=635) = 320.71, p < .001 Although newspapers published more stor ies about the metro area than all other geographic emphases combined, those stories had fewer contextual elements with them than any of the other categories. Stories about other geographic areas also tended to incorporate a slightly more diverse mix of contextual elemen ts, as displayed in Table 9. ANOVA showed that metro stories in print ha ve a mean of 1.38 types of contextual elements out of eight, below a total mean of 1.48, F(3, 631) = 5.812, p = .001. As with the mean in print features, the mean of types of online features that ran with metro stories fell in third place among the four geographi c areas. The ANOVA was significant, F(3, 608) = 3.666, p = .012. Table 10. Story geographic emphasis and m ean types of elements represented Geographic focal area Print story frequency Mean types of print elements Online story frequency Mean types of online elements International 47 1.85 44 1.18 National 115 1.73 106 1.66 State 126 1.37 125 1.94 Metro 347 1.38 337 1.28 Total 635 1.48 612 1.48


45 Data indicated that stories with a ge ographic emphasis on the metro area have fewer contextual elements than stories with other geographic emphases in print. Although the mean number of photos with online stories about the me tro area is slig htly higher (0.29) than the mean number of photos on line for all geographi c emphases (0.28), the difference is not significant, ANOVA revealed, F(3, 608) = 1.555, p = .199. The mean number of infoboxes with online stories about the metro area (0.20) was lower than the mean for any other geographic emphasis a nd therefore below the overall mean (0.23). However, ANOVA showed that differences in the mean number of infoboxes in each geographical category was not si gnificant, F(3, 608) = 1.191, p = .312. International stories have a higher mean number of contextual elements in print and online than stories with other geographic emphases, but th e differences were not significant.


46 Chapter 5: Conclusion This study set out to examine the ways ne wspapers and their Web sites distribute and display news, and the results showed different standards for print content and contextual elements. First of all, nearly all of the stories that app eared on the two section front pages examined also appeared online. Previous studies (Blake, 2004; Singer, 2001) that have examined all news stories in a ne wspaper and on its Web site reported half to three-quarters of stories in pr int also appearing on the news papers Web site. This study has no basis for reporting such a percentage for all stories in an issue, but it did reveal that more than 96% of storie s on the front page and metro se ction front page also appear on the Web. This comparison with past findings woul d seem to indicate that placement on one of these section front pages in print coul d create a positive force for a story, making it easier for it to move past online gatekeepers and gatekeeping routines. Competition for front-page story slots is among the fiercest fi ghts for space at any given newspaper, and when a story makes it through the gate, it means it is more likely to receive attractive packaging, good placement, or repetition (Shoemaker, 1991). The finding about the consistent inclusion of stories from the two news sections opens the door for future study about story placement and its selection across media. Moving beyond story selection, the data suggested that story text may not face additional gatekeeping processes between public ation and posting of editions; only 4% of stories in the sample had different leads be tween media. In traditional inverted-pyramid


47 style news writing, the lead is the place in th e story that contains the most important information, which is often the most forward-lo oking. Therefore, it is a likely place to look for the most updated information in a story. Only a handful of stories offered different content across media as measured by the lead. Once newspapers publish stories, that is the way the stories stay in most cas es, regardless of the limitless opportunities for updated information online. However, this finding also may not mean that newspaper Web sites offer only old news. Some stories that were posted online the same way th ey ran in print had updated versions elsewhere on the site. For instance, several newspapers in the sample posted online the print version of a story about a to ur boat capsizing in New York, almost as if the Web site were running the print version for the sake of record. However, if users went to site sections that provided continuously updated wire news, they would find different versions of the story that mi ght give fresh death tolls or details about the disaster. The inclusion of contextual elements al so provides the opportunity to examine the presentation aspect of gatek eeping. About a third of the entire sample had at least one contextual element in common across prin t and online media, and Shoemaker (1991) pointed to repetition as one si gn that a story has a strong pos itive force to make it through a gate. The 32% figure is, by itself, neither good nor bad. On one hand, the figure could encourage those who have conc luded that online newspapers do little but post the text of any given story and skip elements that might have to be resized or reformatted for digital display in order to get the el ements, such as photos or inf ographics, ready to post online. On the other hand, some might be discouraged by the thought that so many stories are repeating elements from one medium to the next, possibly indicating that the newspapers


48 do not value the differences between platfo rms. Until journalists and users view Web sites as more than carriers for modified print content, they have little room to grow, some have argued (Lowrey, 1999). Finding no relationship between reporter affiliation and contextual elements shakes the notion that staff-written stories are guaranteed better play or display simply because they are a unique commodity to the news paper. It would be easy to assume that a staff reporters access could translate into more photos or other elements to give context to a story in any medium, but the average numbe r of print contextual elements and online photos and infoboxes that ran with staff-wr itten stories showed no significant variance from the mean. A related thought about the use of contextual elements might assume that stories about the metro area would have more elements than stories about issues and events from greater distances away. In fact, stories about international news had the highest mean of contextual elements published with them both in print and online. Stories about the metro areas actually had fewer contextual elements in print on average than international, national, or state stories, and this difference was signifi cant. Although th e number of contextual elements with these stories was lower, more than half of the entire sample was made of stories focusing on the geographic metro area.6% in print and 52.8% online. This finding is disappointing. Many sources offer international and national news; a person in Florida can read different perspec tives and reports of the same global news produced in Washington, New York, London, and Riyadh. The metropolitan newspaper may be one of only a few sources that offers local news, though. Enhanced international and national coverage make newspapers competitors with hundreds of other sources


49 which may have more resources or more di rect connectionsrather than allowing them to serve a local niche. In ignoring their local content, newspape rs are missing the point online, and they are missing a powerful opportunity for compe tition. The growth of the online medium depends on competition because only by pr oviding a desirable commodity that will attract an audience will online newspapers be able to attract advertisers. Advertising revenue provides the resources to pay staff to create enhanced news content online. Newspapers are caught in an online Catch22: until they put some money into Web operations, things cannot change, and until newspapers change some of their content and delivery, hopes of making a profit on the medium seem dim. All the while, local content languishes. Limitations The period of study began about three week s after Hurricane Katrinas landfall in New Orleans, and hurricane clea n-up and recovery stories st ill were receiving prominent play in newspapers across the country. However, Houston and Dallas were particularly large centers for evacuees, and their special c overage of the disaster could have affected the results of this study. For in stance, both of these newspape rs created a standing box of online features and links that was posted with each story about Hu rricane Katrina on the Web sites. These standing boxes provided cont extual content for the stories, but it was not unique; readers saw the same content with story after story, day after day. The Houston Chronicle and Dallas Morning News did the same thing for Hurricane Rita, which directly a ffected three of the cities involved in this study during its first week. The storm struck a glancing blow to the Florida Keys, which are inside


50 Miamis coverage area, and it came ashore near Houston on September 23, sweeping up to cities near Dallas. Houston evacuated befo re the hurricane, which affected newspaper production; during one day of the study, the Chronicle did not produce a traditional metro section, choosing instead to focus all of its local content on hurricane preparation and put that in the front section. Th e special coverage leading up to and following Hurricane Rita seemed to affect the number of front-pag e stories produced and the amount of online elements available. The purposive sample of newspapers for this study assembled a more homogenous group of papers in terms of size and location than would be expected in a random sample of American newspapers and th eir Web sites. The results of this study cannot be generalized to the en tire population but create a sn apshot of these newspapers. Another limitation includes story duplication on the Web site of the Atlanta Journal-Constitution The newspaper provides a prominent link to its print edition, where most of the stories that ran in the paper that day ar e posted by section. The stories include some contextual elements like photos and infoboxes. However, visitors to other sections of the site might find links to the same storiesthat have different contextual features than the first online story. Rather than having one version of a story linked to from several places, the site architecture calls for at least two vers ions of many stories, which could have created a coding nightmare. However, this study consistently sampled only one set of stories, the ones under the p rint edition link. This assured consistent sampling, but it could have been at the expe nse of unique Web features for that site. That limitation points to another difficulty of online research: pa rticular pages can be difficult to find within a site. The design of this study called for extensive searching


51 for a story online before determining that it a ppeared in print only. Th ere is the possibility that search efforts just did not turn up stories that did exist on the site. But if a researcher with determined intent could not find a stor y, the utility of the story to a casual browser must be questioned. In many cases, online ne wspaper sites are re ally quite a mess. Several coding decisions complicated anal ysis of the data. With the exception of online photos and infoboxes, which were c oded by frequency, other online elements simply were coded for presence or absence. Because of the great variety of types of online elements and the many ways they could be displayed or linke d from stories, the researcher noted whether a type of elem ent was present but not how many times it occurred with a story. This data could not be analyzed at the same level as information about contextual elements in print or in common between editions because these variables were measured at the ratio level. Mo st of the online elements were measured at the nominal level. Raising the coding level for the online elements variables would allow more specific analysis and direct comparison with print elements and common elements. Measuring the elements that were the same between versions of a story was a fairly simple comparison that yielded exact in formation about the kinds of elements that are repeated between print and online editions. However, just because a story was counted as having elements in common di d not mean the story didnt have other contextual elements that were unique between media. Future research This study points to a number of ques tions for future research. Whether frontpage story placement in print acts as a ga te for online story selection would be a worthwhile question. Although it intuitively make s sense that the stories chosen as the


52 most important in print also would be importa nt online, the high per centage of print and online overlap in this study al so simply could indicate a gr eater percentage of stories online overall. It also would be interesting to examine whether the affiliations of contextual elements producersphotographers, graphic artists, videographershave any effect on the frequency and use of these elem ents in print and online media. Examining the content differences betw een print and online newspapers answers the question of what is going on but not why or how. Further study of newspapers online gatekeepers and gatekeeping processes is warranted. Scholars also should examine newspapers commitment of resources to producing contextual el ements for either medium. The ability of newspaper Web sites to offer readers something new with their daily news than they would receive in the pa per has advanced in the dozen or so years that newspapers have been online. This study indicates that newspapers still have more potential for growth than substance to ma ny claims about content differences between print and online newspapers. Readers are able to find a mirror of stories they see in print, and they are more likely to find contextual el ements that provide visual storytelling in print than online. The most common online-only contextual elements are links to related sites externally or stories internally, and even those appear in less than one-fifth of stories online. Even though use of contextual elements with online stories is sparse compared to print editions, it should be noted, however, that more than half of the stories online had at least one contextual element posted with th em. Even though newspaper Web sites have been slow to embrace and use interactive a nd unique online features, the presence of


53 contextual elements online that do not appear in print is evidence that the print edition of a newspaper is not always a gate to online posting. Newspapers may be making progress, but they still have a long way to go before claiming widespread content differences between print and online editions.


54 References Abbott, E.A., & Brassfield, L.T. (1989). Co mparing decisions on releases by TV and newspaper gatekeepers. Journalism Quarterly, 66, 851-856. Adhikari, D., Everbach, T., & Fahmy, S. (2002). Watching the watchdogs: An ethnomethodological study of news deci sion making at a small midwestern newspaper. Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Miami Beach, FL. American Press Institute. (2005). U.S. Convergence Tracker Retrieved August 28, 2005, at http://www.mediacenter .org/convergencetracker/. Audit Bureau of Circulations (2004, July 12). ABC board acts to reinforce confidence in ABC audit. Retrieved August 27, 2005, at Audit Bureau of Circulations (2005a, March 31). Reader Profile: Top 150 newspapers by largest reported circulation Retrieved August 16, 2005, at Audit Bureau of Circulations (2005b, May 13). ABC information bulletin #05R91: Commencement of censure audit cycle. Retrieved August 26, 2005, at


55 Audit Bureau of Circulations (2005c, August). Reader profile: The Dallas Morning News Retrieved August 26, 2005, at Bass, A.Z. (1969). Refining the gatekeep er concept: a U.N. radio case study. Journalism Quarterly, 46, 69-72. Beard, F., & Olsen, R.L. (1999). Webmasters as mass media gatekeepers: a qualitative exploratory study. Internet Research: Electroni c Networking Applications and Policy, 9, 200-211. Berkowitz, D. (1990). Refining the gatekeeping metaphor for local television news. Journal of Broadcasting and Electronic Media, 34, 55-68. Berkowitz, D. (1991). Assessing forces in th e selection of local television news. Journal of Broadcasting and Electronic Media, 35, 245-251. Blake, M. (2004, August). The transparent gate: Online and pr int editions at two central Florida newspapers. Paper presented at the meeting of the Association for Education in Journalism and Mass Co mmunication, Toronto, Ontario, Canada. Breed, W. (1955). Social control in the newsroom: a functional analysis. Social Forces, 33, 326-335. Brown, R.M. (1979). The gatekeeper reassessed: a return to Lewin. Journalism Quarterly, 56 595-601, 679. Cassidy, W.P. (2002, August). Offering help creating the da ily me: a content analysis of the credibility and ed itorial policy statements of online news sites. Paper presented at the meeting of the Associat ion for Education in Journalism and Mass Communication, Miami Beach, FL.


56 Chyi, H.I. (2005). Willingness to pay for onlin e news: an empirical study on the viability of the subscription model. Journal of Media Economics, 18, 131-142. Chyi, H.I., & Lasorsa, D.L. (2002). An expl orative study on the market relation between online and print newspapers. Journal of Media Economics, 15, 91-106. Chyi, H.I. & Sylvie, G. (1998). Competing wi th whom? Where? And How? A structural analysis of the electronic newspaper market. The Journal of Media Economics, 11, 1-18. Chyi, H.I., & Sylvie, G. (2001). The medium is global, the content is not: the role of geography in online newspaper markets. The Journal of Media Economics, 14 231-248. DeJean, D. (1995). The on-line version: how it might be done. Nieman Reports, 49, 1016. Dibean, W., & Garrison, B. (2001). How six online newspapers use Web technologies. Newspaper Research Journal, 22(2), 79-93. Retrieved November 1, 2004, from Expanded Academic ASAP database. Dimmick, J. (1974). The gate-keeper: an uncertainty theory. Journalism Monographs, 37 Donohue, G.A., Olien, C.N., & Tichenor, P.J. (1989). Structure and constraints on community newspaper gatekeepers. Journalism Quarterly, 66 807-812, 845. Donohue, G.A., Tichenor, P.J., & Olien, C.N. (1972). Gatekeeping: mass media systems and information control. In F.G. Kline & P.J. Tichenor (Eds.), Current perspectives in mass communication research (pp. 41-69). Beverly Hills, CA: Sage.


57 Dutta-Bergman, M.J. (2004). Complementarity in consumption of news types across traditional and new media. Journal of Broadcasting & Electronic Media, 48, 4160. Retrieved November 1, 2004, from E xpanded Academic ASAP database. Epstein, E.J. (2000). News from nowhere: television and the news (2nd ed.). Chicago: Ivan R. Dee. Fortunati, L. (2005). Mediatization of the Ne t and Internetization of the mass media. Gazette: The International Jour nal for Communication Studies, 67, 27-44. Gieber, W. (1956). Across the desk : a study of 16 telegraph editors. Journalism Quarterly, 37 423-432. Gieber, W. (1964). News is what newspapermen make it. In L.A. Dexter & D.M. White (Eds.), People, Society, and Mass Communications (pp. 173-182). New York: The Free Press of Glencoe. Greer, J., & Mensing, D. (2004). U.S. news Web sites better, but small papers still lag. Newspaper Research Journal, 25, 98-112. Gubman, J., & Greer, J. (1997, August). An analysis of online sites produced by U.S. newspapers: Are the critics right? Paper presented at the meeting of the Association for Education in Journa lism and Mass Communication, Chicago. Harmon, M.D. (1989). Mr. Gates goes electron ic: the what and why questions in local TV news. Journalism Quarterly, 66, 857-863. Hirsch, P.M. (1977). Occupational, organiza tional, and institutional models in mass media research: toward an integrated framework. In P.M. Hirsch, P.V. Miller, & F.G. Kline (Eds.), Strategies for communication research Beverly Hills, CA: Sage.


58 Huang, E., Davison, K., Shreve, S., Davis, T., Bettendorf, E., & Nair, A. (2003, August). Facing the challenges of convergence: Media professionals concerns of working across media platforms. Paper presented at the mee ting of the Association for Education in Journalism and Mass Communication, Kansas City, MO. Jones, R.L., & Carter, R.E. (1959). Some procedures for estimating news hole in content analysis. Public Opinion Quarterly, 23 399-403. Jones, R.L., Troldahl, V.C., & Hvistendahl, J.K. (1961). News selection patterns from a state TTS-wire. Journalism Quarterly, 38, 303-312. Kamerer, D., & Bressers, B. (1998, August). Online newspapers: A trend study of news content and technical features. Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Baltimore, MD. Killebrew, K.C. (2004). Managing media convergence. Ames, Iowa: Blackwell. Lewin, K. (1947). Frontiers in group dynamics II: channels of group life; social planning and action research. Human Relations, 1,143-153. Livingston, S., & Bennett, W.L. (2003). Gate keeping, indexing, and live-event news: Is technology altering the construction of news? Political Communication, 20, 363380. Lowrey, W. (1999). From map to machine: Conceptualizing and designing news on the Internet. Newspaper Research Journal, 20 (4), 14-27. Retrieved November 1, 2004, from Expanded Academic ASAP database. Lowrey, W. (2003). What influences small ne wspapers to decide to publish online news? Newspaper Research Journal, 24, 83-90. Retrieved November 1, 2004, from Expanded Academic ASAP database.


59 Martin, S.E. (1998). How news gets fr om paper to its online counterpart. Newspaper Research Journal, 19, 64-73. Massey, B.L. (2000). Market-based predictors of interactivity at Southeast Asian online newspapers. Internet Research: Electronic Networking Applications and Policy, 10, 227-237. McCombs, M.E., & Shaw, D.L. (1976). Sett ing the political age nda: structuring the unseen environment. Journal of Communication, 26, 18-22. McCombs, M.E., & Shaw, D.L. (1993). Th e evolution of agenda-setting research: twenty-five years in th e marketplace of ideas. Journal of Communication, 43, 5867. McMillan, S.J. (2000). The microscope and th e moving target: The ch allenge of applying content analysis to the World Wide Web. Journalism and Mass Communication Quarterly, 77 80-98. Morris, J. (2001, August). Newspapers in the age of the Inte rnet: adding interactivity to objectivity Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Washington, DC. Neuberger, C., Tonnemacher, J., Biebl, M., & Duck, A. (1998). Online: The future of newspapers? Germanys dail ies on the World Wide Web. Journal of Computer Mediated Communication, 4. Accessed June 8, 2005, at Pavlik, J.V. (1997, July/August). Th e future of online journalism. Columbia Journalism Review Accessed June 13, 2005, at http://arc /97/4/online.asp


60 Pavlik, J.V. (2001). News frami ng and new media: digital tool s to re-engage an alienated citizenry. In S.D. Reese, O.H. Gandy, & A.E. Grant (Eds.), Framing public life: perspectives on media and our understanding of the social world (pp. 311-322). Matwah, NJ: Lawrence Erlbaum. Peng, F.Y., Tham, N.I., & Xiaoming, H. (1999). Trends in online newspapers: a look at the U.S. Web. Newspaper Research Journal, 20, 52-63. Project for Excellence in Journalism. (2005, March). State of the news media 2005. Washington, DC. Accessed August 22, 2005, at Poynter Institute. (2004, July). Eyetrack III: Online news consumer behavior in the age of multimedia Retrieved March 18, 2005, from eyetrack2004/index.htm. Reese, S.D. (2001). PrologueFraming public li fe: a bridging model for media research. In S.D. Reese, O.H. Gandy, & A.E. Grant (Eds.), Framing public life: perspectives on media and our understanding of the social world (pp. 7-31). Matwah, NJ: Lawrence Erlbaum. Reese, S.D., & Ballinger, J. (2001). The r oots of a sociology of news: Remembering Mr. Gates and social cont rol in the newsroom. Journalism and Mass Communication Quarterly, 78, 641-658. Regan, T. (1995). News alone is not enough. Nieman Reports, 49, 78. Riffe, D., Aust, C.F., & Lacy, S.R. (1993). Th e effectiveness of random, consecutive day and constructed week sampling in newspaper content analysis. Journalism and Mass Communication Quarterly, 70 133-139.


61 Saksena, S., & Hollifield, C.A. (2002). U.S. newspaper and the development of online editions. International Journal on Media Management, 4, 75-84. Sasser, E.L., & Russell, J.T. (1972) The fallacy of news judgment. Journalism Quarterly, 49, 280-284. Schaffer, J. (2001, March 31). Interactive journalism Keynote speech presented at the Pittsburgh SPJ Region 4 Conference, Pittsburgh, PA. Retrieved November 8, 2004, from ngcj/speeches/s_pi ttsburghspj.html. Schultz, T. (1999). Interactive options in online journalism: A content analysis of 100 U.S. newspapers. Journal of Computer Mediated Communication, 5(1 ) Available online at vol5/issue1/schul tz.html. Accessed November 8, 2004. Shoemaker, P.J. (1991). Communication Concepts 3: Gatekeeping Newbury Park, CA: Sage. Shoemaker, P.J., & Reese, S.D. (1996). Mediating the message: Theories of influences on mass media content (2nd ed.). White Plains, NY: Longman. Shoemaker, P.J., Eichholz, M., Kim, E., & Wrigley, B. (2001). Individual and routine forces in gatekeeping. Journalism and Mass Communi cation Quarterly, 78, 233246. Singer, J.B. (2001). The metro wide web: Changes in newspapers gatekeeping role online. Journalism and Mass Communication Quarterly, 78, 65-80. Singer, J.B. (2002). Information trumps inte raction in local pape rs online caucus coverage. Newspaper Research Journal, 23 (4) 91-96. Retrieved November 1, 2004, from Expanded Academic ASAP database.


62 Singer, J.B. (2005, August). Stepping back from the gate: Online newspaper editors and the co-production of content in Campaign 2004. Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, San Antonio, TX. Snider, P.B. (1967). Mr. Gates revis ited: a 1966 version of the 1949 case study. Journalism Quarterly, 44, 419-427. Stempel, G.H., III. (1952). Sample size for classifying subject matter in dailies. Journalism Quarterly, 29 333-334. Stempel, G.H., III. (1985). Gatekeeping: the mix of topics and the selection of stories. Journalism Quarterly, 62 791-796, 815. Sundar, S.S. (2000). Multimedia effects on processing and perception of online news: A study of picture, audio, and video downloads. Journalism and Mass Communication Quarterly, 77, 480-499. Tankard, J.W., & Ban, H. (1998, August). Online newspapers: Living up to their potential? Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Baltimore, MD. Tewksbury, D., & Althaus, S.L. (2000). Di fferences in knowledge acquisition among readers of the paper and online ve rsions of a national newspaper. Journalism and Mass Communication Quarterly, 77, 457-479. Tuchman, G. (1978). Making news: a study in th e construction of reality. New York: The Free Press.


63 Weaver, D.H. (1989). Basic statistical tools. In G.H. Stempel III & B.H. Westley (Eds.), Research Methods in Mass Communication (pp. 49-89). Englewood Cliffs, N.J.: Prentice-Hall. Weinberg, S. (1996, Nov./Dec.). Can content-providers be investigative journalists? Columbia Journalism Review, 36-38. Weiss, A.S. (2004, August). The now what factor: the l evel of innovativeness among online newspapers Paper presented at the meet ing of the Association for Education in Journalism and Mass Co mmunication, Toronto, Ontario, Canada. Welch, A.L. (2004). Multimedia and intera ctive components in converged media Unpublished masters thesis, University of South Florida, Tampa. Available online at Westley, B.H., & MacLean, M.S., Jr. (1957). A conceptual model for communications research. Journalism Quarterly, 34, 31-38. White, D.M. (1950). The gatekeeper: a case study in the selection of news. Journalism Quarterly, 27, 383-390. Wigley, S. (2004, August). Whos setting the media agenda? A look at non-traditional interactive media and the gatekeeping process Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Toronto, Ontario, Canada. Williams, B.A., & Delli Carpini, M.X. (2004, May). Monica and Bill all the time and everywhere: The collapse of gatekeeping and agenda setting in the new media environment. American Behavioral Scientist, 47, 1208-1230.


64 Wimmer, R.D., & Dominick, J.R. (2003). Content analysis. In Mass media research: an introduction (7th ed.) (pp. 139-165). Belmont, CA: Thomson-Wadsworth. Zaharopoulos, T. (2003, August). Online versions of U.S. daily newspapers: Does size matter? Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Kansas City, MO.


65 Appendices


Appendix A: Print and Online Sample Articles Print Houston Chronicle, September 19, 2005, Page A1 66


Appendix A (Continued) 67


Appendix A (Continued) Online Houston Chronicle, September 19, 2005 68


Appendix A (Continued) 69


Appendix A (Continued) 70


Appendix A (Continued) Dallas Morning News, September 19, 2005, Page B1 71


Appendix A (Continued) 72


Appendix A (Continued) Online Dallas Morning News, September 19, 2005 73


Appendix A (Continued) Print Atlanta Journal-Constitution, September 19, 2005, Page A1 74


Appendix A (Continued) 75


Appendix A (Continued) Online Atlanta Journal-Constitution, September 19, 2005 76


Appendix A (Continued) 77


Appendix A (Continued) Print St. Petersburg Times, September 19, 2005, Page A1 78


Appendix A (Continued) 79


Appendix A (Continued) Online St. Petersburg Times, September 19, 2005 80


Appendix A (Continued) Print Miami Herald, September 19, 2005, Page A1 81


Appendix A (Continued) 82


Appendix A (Continued) Online Miami Herald, September 19, 2005 83


84 Appendix B: Coding Sheet NEWS: HC DMN AJC SPT MH DATE: PRINT ONLINE SECT. Front Metro P. ONLY Y N HED HED Same Different AFFIL. Staff Wire Staff & Wire Contributor Unknown AFFIL. Staff Wire Staff & Wire Contributor Unknown GEOG. Intl Natl State Metro GEOG. Intl Natl State Metro ELEM. Y N photos sidebars __ infographics infoboxes __ artwork other AFFIL.: Staff Wire Staff & Wire Contributor Unknown ELEM. Y N ___ photos infoboxes ___ video audio docs ___ poll quiz i. graphic ___ blog disc. forum ___ s. graphic slideshow ___ photo gallery rel. Web links ___ rel. story links live chat ___ other AFFIL.: Staff Wire Staff & Wire Contributor Unknown SAME: Y N LEAD LEAD Same Different


85 Appendix C: Coding Instructions NEWS Select newspaper and date. HC =Houston Chronicle, DMN =Dallas Morning News, AJC =Atlanta Journa l-Constitution, SPT =St. Petersburg Times, MH =Miami Herald PRINT SECT Select front or metro (which may be called City & State, Metro, Tampa & State, or Metro & State) HED Write headline that appears with stor y. In some cases, a story may have more than one headline, so here are some guidelines. If a big headline acts as an umbrella over two or more stories, use the smaller headlines that appear directly over each story. If only one story falls under the big h eadline, use the big one in that case, even if there is a smaller headline directly over the story. If a small headline (like a label, such as Disaster Relief, or a feature headline that particularly addresses th e action in the photo) is stacked on top of a picture that is stacked on top of a story that has anothe r headline directly over it, use the headline directly ove r the story, not the small overline. AFFIL Select the affiliation of the reporter. Staff Wire =Associated Press, Reuters, Bloomberg News, or any chain corres pondent (identified with Cox, KnightRidder, etc.), Staff & Wire =combination of any of the above (staff writer might be listed alone in byline, but story has a footer that says wire material was used), Contributor =might be identified as a co rrespondent or special to the Chronicle, etc., Unknown GEOG Location the story focuses on: International National State, Metro Determine this by skimming first few se ntences of story. A story may have more than one geographic focus. ELEM Identify the presence of any elements besides story text with yes or no and place the number of each kind of element next to it. Photos =photographs or artists renderings that have a box around the outside. Sidebars=related stories that appear on the same page as the story or its jump. Story topics should show a strong relationship, not just two stories about the military or a hurricane. Infographics =information delivered gr aphically. Includes maps. Infoboxes =facts, figures, or details that suppor t a story and appear in a box or module with the story. Artwork =includes sketches or any kind of illustration floating without a border. Other =anything else. If you choose this option, please describe the element to the right. Examples include: refer (a note directing readers to other storie s in the paper or on the Web), pull quote (a


86 Appendix C (Continued) quote from the story that is inset into the text in big type), or pull out (a sentence or two from the story that is inset into the text in big type). AFFIL Select the affiliation of the photographer or artist. Staff Wire =Associated Press, Reuters, Bloomberg News, or a ny chain correspondent (identified with Cox, Knight-Ridder, etc.), Staff & Wire =combination of any of the above (staff writer might be listed alone in bylin e, but story has a footer that says wire material was used), Contributor =might be identified as a correspondent or special to the Chronicle, etc., Unknown If an element, particularly an in fobox, does not name the person responsible for creating it, mark it as staff if th e element appears in a staff-written story or wire if it appears in a wire-written story. If the element appears in a staff & wire story, then mark the element as unknown affiliation. ONLINE P. ONLY Does the story appear in the print edition only? If a scre enshot of the story does not appear in that newspapers folder for that date, the answer is yes. HED Note whether the primary headline on the online story is the same or different than the print edition. Only write in the headline again if it is different. AFFIL Select the affiliation of the reporter. Staff Wire =Associated Press, Reuters, Bloomberg News, or any chain corres pondent (identified with Cox, KnightRidder, etc.), Staff & Wire =combination of any of the above (staff writer might be listed alone in byline, but story has a footer that says wire material was used), Contributor =might be identified as a co rrespondent or special to the Chronicle, etc., Unknown GEOG Location the story focuses on: International National State, Metro Determine this by skimming first few se ntences of story. A story may have more than one geographic focus. ELEM Identify the presence of any elements besides story text with yes or no and place the number of each kind of elem ent next to it if it is a photo or infobox. Otherwise, simply put an X next to any element present, or use an S next to any element that appear s in a standing box. A standing box would be one that appears with several relate d stories (such as a box labeled DeLay indicted or Hurricane Rita). Elemen ts should appear with intent; i.e., it doesnt count if every page of a site li nks to a general discussion forum in its site architecture. Photos =photographs or artists renderings that have a box around the outside.


87 Appendix C (Continued) Infoboxes = facts, figures, or details that support a story and appear in a box with the story. Video=links to video clips. Audio =links to audio clips. Docs=links to documents that we re sources for the article. Poll =allows readers to vote on a question presented. Quiz =a series of questions for readers to complete. I. graphic =interactive graphic. Can the r eader click or manipulate the presentation of the da ta? Can include maps. Blog=either a reporters Webl og or a blog that readers can participate in. Disc. Forum =discussion forum or bulletin board. S. graphic =static graphic. The graphic does not update or respond to user input. Can include maps. Slideshow =series of photos that automa tically runs when clicked. Photo gallery =series of photos that must be manually run when clicked. Rel. Web links =links to related Web sites. Rel. story links =links to other related stories or coverage on that newspaper site. Live chat =link for readers to participate in a real-time online conversation with one another, experts and sources, or journalists. May also be a link to a transcript of a chat that is already past. Other =anything else; please describe it to the right. Might include refer (a note but no link directing readers to othe r stories in the paper or on the Web) or submit photos (a feature allowing readers to send in their own photos). AFFIL Select the affiliation of the journalist. Staff Wire =Associated Press, Reuters, Bloomberg News, or any chain corres pondent (identified with Cox, KnightRidder, etc.), Staff & Wire =combination of any of the above (staff writer might be listed alone in byline, but story has a footer that says wire material was used), Contributor =might be identified as a co rrespondent or special to the Chronicle, etc., Unknown If an element, particularly an in fobox, does not name the person responsible for creating it, mark it as staff if th e element appears in a staff-written story or wire if it appears in a wire-written story. If the element appears in a staff & wire story, then mark the element as unknown affiliation. SAME Circle yes or no to note whether any of the elem ents that appeared in the Web version of the story was the same as any element that appeared in print. If yes, please write to the right wh ich element was in both media, such as photo or infobox. LEAD Circle same if the first paragraph of the Web story says exactly the same thing as the print story. Circle di fferent if the leads vary at all.