Is INTERNET CONTENT DIFFERENT AFTER ALL? A CONTENT ANALYSIS OF MOBILIZING INFORMATION IN ONLINE AND PRINT NEWSPAPERS By Lindsay H. Hoffman The present study answers calls of previous research to analyze the content of online news, assessing differences between print and online newspapers in frequency of "mobilizing information" (MI)—information aiding people to act on pre-existing attitudes. Conceptual and operational definitions of locational, identificational, and tactical MI are included. Results revealed that online newspapers did not have significantly more MI than their print counterparts. This finding counters the assumption that online newspapers have more mobilizing content than print.

Research shows a decline in voter turnout since the 1950s, while other forms of political participation also have declined over the past four decades,' Reduced newspaper readership has been named as one precursor to this falloff in participation,^ with newspaper readership decreasing from 47% in 2000 to 42% in 2004,^ Some scholars hail the Internet as a possible route for reinvigorating the democratic process and encouraging citizens to become more involved.'' Others have argued the Internet could negatively affect the electorate, much like other media such as the telephone and television, which at first were thought to enhance participation,' However, a study by Kraut et al, suggests that such negative effects of the Internet have dissipated,^ "Changes in the Internet environment itself," they noted, "might be more important to understanding the observed effects than maturation or differences between samples,"' Those seeking news online increased from 23% in 2000 to 29% in 2004,* Reading polifical news is also popular: 40% of Internet users reported this activity in 2002,' This suggests that news and politics online could have substantial effects and that systematic analysis of fhis content could be fruitful. Communication scholars have long advocated examination of content'" in addition to time spent with a medium." Content, more than just the diverse materials presented on various media, changes from day to day,'^ Analysis of content is important in understanding what inforI&MC Quarterly Vol. S3, No. 1 Spring 2006 58-76 ©2006AE]MC

58

Lindsay H. Hoffman is a Ph.D. student in the School of Communication at Ohio State University.

JOURNALISM & MASS COMMUNICATION QUARTERLY

mation people are exposed to, above and beyond measures of time spent with media. Content analyses of fhe Internet, though, have generally focused on how messages are presented rather than what is communicated,'^ Thus, in the absence of a true theoretical framework, many Infernef studies fail to come to replicable conclusions, Bucy asserted, "Web content is offen measured and described in a fheory-free confexf, adding liftle to our conceptual understanding of the online medium."'^ Although effects cannot be ascertained from a content analysis, fhe absence of confenf research in fhe Infernet-effecfs debate ignores fhe importance of studying confenf before effecfs. As Riffe, Lacy, and Fico argued, "in fhe long run, one cannot study mass communication without studying content."'^* One type of confenf that has been sfudied in analyses of news and fhat is the focus of fhis study is mobilizing information (MI), defined simply as information found in news fhaf aids people to act on attitudes they already hold."* Theoretical Framework. The introduction of a new communicafion technology, such as fhe Infernef, allows scholars to "rethink, rather than abandon, definifions and categories," including conceptualizations like MI."" This study seeks to answer the question: Does the presence of MI differ in online versus prinf newspaper contexts? Eveland suggesfed applying uniform analyses to media, assessing "mixed attributes" to differenfiafe media,"* This sfudy assesses prinf and online confent in as uniform a way as possible fo evaluate whether new technology has changed the content of newspapers. Importantly, new technology alters the sociology of news production, which suggests that content differences may be found. For insfance, Pavlik predicts that new media will encourage journalists to move beyond norms of objectivity." Some cases of what Pavlik calls "contextualized journalism" have already occurred. For example, reader feedback via e-mails fo journalisfs has changed the way future stories are written, readers can create unique content, and journalists have relinquished some control to consumers who can customize their news experience. Newspapers, in particular, are being forced to make their Web sites "commercially viable" to compete with other sources.^" Boczkowski found fhaf online newsrooms differ from print newsrooms in the amount of autonomy they have, policies, and definitions of Web site users,^' He suggesfed that online, news "moves from being mostly journalisf-centered.,,fo also being increasingly audience-cenfered."^^ The Internet could also have advantages over other media because reduced material cost can provide more incentive for consumers fo search for political informafion^^ as well as decrease production costs for news producers. The Internet as a Participatory Medium. Much debate surrounding the Internet as a political and participatory medium relates to its ability to create new relationships;^" function as a virtual forum;^' or enhance social networks,^*^ Findings generally show that this social aspect of the Internet is ineffective in producing social trust or engagement,^' Some research has found, however, that online users who search for informafion and is INTERNET CONTENT DIFFERENT AFTER ALL?

59

exchange e-mail, rather than those who chat or play games, are more active in civic life.^* Thus far, no empirical research directly links the Internet to political participation,^' Indirect effects have been monitored, and "polifical learning and mofivafion for patficipation are enhanced through various types of communicafion, which in turn increases the level of political participation,"'"' In this way, the informafion fhat so many users seek online could encourage participation. Mobilizing Information. MI can be found online via hyperlinks, contact informafion, and linked confenf on ofher Web sites. Although it does not inherently motivate citizens, MI enables them to act on existing motivations by providing informafion on how to follow fhrough with an acfion,^' Research on MI has declined since the 1980s, yet the concept continues to be cited in the context of media's role in mobilizing political participation and civic engagement,'^ as well as in health communication,^' MI is relevant to the discussion of media effecfs and polifical parficipation because MI has the potential for direcfly encouraging citizens to become involved in politics, Lemert and his colleagues, credited with the term, acknowledged that MI originated in publications on environmental and media studies,** Two articles, published mere months apart, are cited as the concept's inspiration, Murch's content analysis of newspapers revealed fhaf local media focused more on nafional rafher fhan local pollufion issues, excluding information that could provide information on which to act,^' Rubin and Sachs's report correlated increased public interest in environmental issues and increased media discourse on the environment during the 1970s,^* While "mobilizing informafion" was nof referenced explicifly, Rubin and Sachs referred to "specific informafion, and the day-to-day alerting which is necessary for any citizen to participate effecfively in the legislative process," which "the public at large still expects the news media to provide,"^' They asserted that the public does not have access fo such informafion outside fhe media and "one cannot become involved without af least knowing when and where hearings are being held, and who if is that should be contacted in order fo voice an opinion,"'" Three Types of Mobilizing Information. Locational MI provides infor-

mation about the time and place of an acfivity; identificational MI provides names and contact information for people or groups; and tactical MI is the explicit and implicit instructions for certain behaviors, such as facfics used in a sfrike. Because some MI could be controversial or even harmful, Lemert argued that mainstream media are reluctant to provide coverage of such evenfs, lef alone publish information fhaf might encourage participation,'^ Lemert also differenfiated between "political MI," (locational, identificational, and tactical) and nonpolidcal MI (such as advertisements, cooking recipes, or garden dps). He admitted, however, that MI "needs to be described in more detail" and that it is "a relatively new concept, and our thinking about it is still changing,"*"* How does one move forward, fhen, in a meficulous sfudy of MI in news content? Lemerf provided calls for research, mosf of which cen60

JOURNALISM & MASS COMMUNICATION

QUARTERLY

tered on reasons MI is excluded from media content. He claimed, however, that there is so little MI in news that it becomes difficult to test for it."' Previous work on MI has focused on environmental or specialized contexts. The present study seeks to analyze the concept in a more general news context. Where Is Mobilizing Information? MI is essential in advertising; the advertiser wants potential customers to know where and how to purchase a product or service (e.g., "in your grocer's freezer"). In news, however, MI varies depending on the type of news presented. Positive and noncontroversial items are most likely to include MI because they are "harmless" or "good for the community."''^ Local items are more likely than nonlocal and wire items to include MI."^ Editorial pages, including letters to the editor, are extremely unlikely to contain MI because of a journalistic emphasis on an opinion forum rather than a "bulletin board" of community or political causes." A contemporary content analysis of MI must acknowledge that there could be differences between Lemert's conceptualizations and the reality of today's news content. Online mobilizing information can take the form of e-mail links, discussion boards, or links to government agencies, interest groups, political candidates, and other sites. Mobilizing Information Online. There are multiple reasons for revisiting MI, particularly in light of new media technology. Hyperlinks lead Web users to new information, either within a site or outside it. Some online newspaper sites provide audio/video links or interactive links to discussions, polls, and chat rooms, which may also provide further mobilizing information. While much online newspaper content is loaded by "shovelware" straight from print content, the possibilities for additional MI arguably abound in this new technology."^ Because online newspapers are nonlinear, they may provide the user with more control and make information more readily available. Hyperlinks are common in online news."* Tewksbury and Althaus claim that hyperlinks in online newspapers provide access to information that is often buried in print newspapers."^ Hyperlinks represent a shift away from the traditional gatekeeping role for mainstream media, by providing the user with more control and interactivity, enhancing comprehension."^ Moreover, online newspapers can cross-link to related sites and potentially provide more in-depth articles in the nearly limitless space available. In addition, online newspapers might present different content because, "to attract and serve a younger demographic, [they] must live by different rules.""' For example, SFGate.com (the portal for the San Francisco Chronicle) published a photo of barebacked runners; the print Chronicle took a more conventional approach, printing photos of clothed runners.^ Lawrence.com (Kansas) published an entertainment column online with numerous profanities that, in the paper version, were printed with asterisks.^' Although these are by no means examples of MI, they provide anecdotal evidence that online newspapers do not always publish exactly the same content or follow the same rules as print papers. If online /s /NTERNET CONTENT DiFFERENT /IFTER

ALL?

newspapers need fo make themselves "commercially viable," they may differentiate themselves from their prinf counterparts^^ through inclusion of MI, which could be considered unsuitable for prinf. Scholars have already assessed some differences between online and "offline" confenf. For example, fraditional print and broadcast campaign informafion has been found to differ from campaign Web sites, which actually conformed to look like each other by including similar porfals of contenf.^' This suggesfs that if content tends fo become homogenized, as Bagdikian claimed, online newspapers may resemble ofher Web sifes more fhan their print counterparts.^ Therefore: HI: Online newspapers will have more MI fhan print newspapers, Addifionally, a question thaf has nof been clarified in previous research is whether one type of MI is more prevalent in news content, A first step toward understanding the types of MI is assessing which appears most often, RQl: Does one type of MI (locafional, idenfificational, or facfical) exist more than another in print news content? RQ2: Does one type of MI (locational, identificational, or tactical) exist more than another in online news confenf? Method

S2

Sample. This sfudy sampled a broad selecfion of American newspapers, both prinf and online (see Table 1), selected fo include a variety of communifies, ownership statuses, circulation rankings, content, and regions. News media are subject to publication cycles, so this analysis used the "constructed week" method, which randomly selects an issue for each day of the week to account for systematic variation due fo fhe day of fhe week,'^ Riffe, Lacy, and Fico recommended two constructed weeks from a year for daily newspapers (randomly selecting two Mondays, two Tuesdays, etc,),^'' For convenience purposes, and fo obtain the mosf current sample, this study analyzed one constructed week (seven days) during fhe seven-week period from 19 Ocfober fo 3 December 2003,^^ The days within each week were selected using a random numbers fable without replacement. Online confent can vary from hour fo hour, so online news was downloaded and archived at fhe same time for each day sampled,^* Measurement. The unif of analysis was fhe individual news sfory. Prinf newspapers were analyzed for presence and type of MI in texf. However, online newspapers require a differenf set of indicafors to idenfify MI, Web-specific elements such as discussion boards and hyperlinks cannot be ignored. Therefore, nonlinear elements—audio/video links, JOURNALISM & MASS COMMUNICATION QUARTERLY

TABLE 1 Newspapers Sampled for the Content Analysis New York Times www ,nytimes ,com

City Population'

8,008,278

Newspaper Circulation^

1,109,371 subscribers

Ownership'

Owned by the New York Times Company

Notes

National, agenda-setting paper

Washington San Columbus Denver Post Philadelphia Post Francisco Dispatch www Inquirer www Chronicle www,dispatch ,denverpost www,philly ,washington www,sfgate ,com ,com ,com/mld post,com ,com /inquirer /chronicle 572,059

776,733

711,470

759,864 subscribers

512,042 subscribers

244,204 subscribers

554,636

1,517,550

305,929 365,154 subscribers subscribers

Owned by Owned by Independently Jointly Owned by the the Hearst owned by the owned by Knight Washington Corporation Dispatch the Media Ridder Post Media Group News Company Group and Media General National paper

Regional Regional There is no paper that paper, and Saturday has done the only edition civic independently for the journalism owned paper Denver projects, in this sample Post, so with history the Rocky of regional Mountain activism News was substituted. It is owned by E,W, Scripps,

Regional paper

' Obtained from United States Census, (2000), 2 Obtained from Editor & Publisher Yearbook (New York: Editor & Publisher Co,, 2002), ' Obtained from Columbia Journalism Review, "Who Owns What?" (2003), inferactive links, and hyperlinks—were coded, Definifions were created to differentiate between types of pages and types of links. Conceptual and Operational Definitions. Because explicit coding rules and criferia for analyzing MI have nof been published previously, both conceptual and operational definitions of locafional, idenfificational, and tactical MI were constructed for this analysis,^' Although MI can be found within differenf contexts, this study examines only political news, as derived from Perloff's definifion of politics: "a process whereby a group of people, whose opinions or interests are initially divergent, reach collecfive decisions which are generally Is INTERNET

CONTENT

DIFFERENT AFTER

ALL?

63

regarded as binding on the group, and enforced as common policy."'''' Therefore, political MI contained only those news stories about decisions that are regarded as binding on the group and enforced as common policy. Specifically included were stories in National, International, Political, and Local/Metro sections.*' Only MI that fit this definition of "politics" was coded. Operationally, this means that to be included, each instance of MI had to represent a decision to be made regarding public policy or government. Because this is still quite broad, an example may help demonstrate criteria for inclusion. An announcement for a neighborhood social where community members are welcome to share food and play games would not be included, while an announcement for a neighborhood meeting to discuss new regulations enforced by city council would. The question to be answered in deciding what content should be included was, "Does this represent the fact that multiple opinions exist and a common policy should or will ultimately be enforced by political actors?" The three types of MI are not mutually exclusive and more than one type may exist in a unit of analysis, particularly with tactical MI (i.e., a unit may describe specifically how to accomplish an activity, but also provide the time and place for that activity). Locational MI provides information about a time and place for an activity. Excluded from examples of locational MI were program listings for television and radio, as well as obituaries. The fundamental attribute of locational MI is that it provides a time and place for a political activity or event. An example of locational MI from this study's sample provided information about a Philadelphia City Council committee: "The committee meeting will be at 3 p.m. tomorrow in Room 696 at City Hall."*^ Online, there are multiple opportunities for locational MI that may not exist in print. A meeting announcement might be hyperlinked from a story about a new residential development, or a hyperlink could link to details on a weekend rally. For both print and online, the question to ask when assessing locational MI was, "Does this give me enough information to attend an event in which I might be interested?" Identificational MI provides names and contact information for a political candidate or representative, interest group, political party, corporation, or government agency. Examples include names and addresses or phone numbers; names and positions in relatively stable, easily locatable organizations (i.e.. Vice President Dick Cheney); and company or group names. Again, the emphasis is on strictly political or publicpolicy contexts and the key element is contact information. Online, a hyperlink is inherently identificational if it links to an e-mail address or a Web site of a person or organization. Reporters often provide a link to their e-mail address in their byline. Because this study examined political news, these links were included, as they allow readers to act on information they read. Not all newspapers provided this information. However, the Philadelphia Inquirer, for example, provided both a phone number and e-mail address for each article." Tactical MI makes available explicit and implicit behavioral models for action."^ In this study, only explicit tactical information was 64

JOURNAUSM & MASS COIMAUNICATION QUARTERLY

coded. Explicit tactical MI tells a person exactly how to accomplish a political activity without necessarily providing a time, place, or contact information. The chief component of tactical MI is that it tells a reader how to do something, such as how to participate in a rally or how to affect the legislative process. An example comes from the San Francisco Chronicle, which encouraged citizens to contact officials when things went wrong in their neighborhoods.*"' The Chronicle provided specific information on how to make a complaint to be published in the newspaper and, once complaints were published, contact information for officials in charge was given. This represents tactical MI because explicit directions are provided for how to accomplish the activity. The question asked in ascertaining tactical MI was, "Does this give me enough information to become involved in some way, based on the attitudes I may already hold?" Procedure. Print articles were collected through subscriptions to each newspaper during the sampled period. Online articles were collected using WebZIP software (version 5.0.5.896) from Spidersoft and were downloaded only at the first level—the news article page and subsequent pages for that same article (i.e., pages 2, 3, etc. if the article did not fit on one page).^ The content beyond the news article page(s) was not downloaded for analysis. Online and print articles were selected based on whether they fit the coding criteria, then assigned numbers. Links were assessed for mobilizing information based on the information they provided on the news article page(s) (i.e., if information provided on the page indicated that MI was accessible by clicking on that link). An a priori power analysis was conducted using GPOWER software (version 2.0) in order to minimize the sample size while maintaining sufficient power.*-' The total sample before this analysis was 1,418 articles. It was concluded that 65 articles per version of each newspaper (online and print) were adequate to maintain power, and the sample was reduced to 780 articles. Because some issues did not include enough articles to reach this total, the final sample was 763.*''* Reliability. Two coders assessed reliability after being trained with coding rules and definitions. Reliability was analyzed for locational, identificational, and tactical MI. Eighty-seven articles were sampled from the same seven-week period from days on which sampling did not take place. This meets Wimmer and Dominick's recommendation of coding 10% to 20% of the total sample for reliability.*^' Krippendorf's alpha was used because it takes into account chance agreement.''" All reliability coeffiecients met standards set by Krippendorf and Riffe, Lacy, and Fico (above .80), except tactical MI in online text (0.76), tactical MI in print text (0.72), and locational MI in online text (0.66).^' Latent content is admittedly more difficult to assess; therefore, those coefficients above .65 were deemed acceptable.'^ Disagreements were addressed by the coders until understanding was reached before actual coding began. In order to prevent biases due to coder fatigue or repetitive stress, care was taken not to code arti/s (NTEKNET CONTENT DIFFERENT AFTER ALL?

65

cles in any systematic order and coding was completed in one- to twohour increments. Results

66

MI in Newspapers by Type. Of all MI found in the sampled articles, identificational was the most plentiful. Lemert's definition, which stated simply that this type of MI "usually requires" both identification (such as a name) with locational information, did not provide specific coding rules or operationalizations." Therefore, coding rules were created. This may explain the differences in MI found here when compared to Lemert's and colleagues' findings.'" Hungerford and Lemert found that 20% of environmental stories had some form of MI, while Lemert, et al. reported that 22.8% of main news stories had some form of MI.'= Similar results were gleaned from Stanfield and Lemert who found that, in two mainstream dailies, between 13% and 17% of all stories had some form of MI.™ Perhaps because of a more inclusive operationalization of identificational MI, 99.5% of articles in the present study contained some form of identificational MI, while 9% of all articles contained locational MI, and 5.1% contained tactical MI. MI by Medium: Online versus Print. HI predicted more MI overall in online newspapers than in print because of added space, technological advances, and relative freedom. But even with the inclusion of hyperlinks, video/audio links (of which there were zero occurrences with MI), interactive links, and online artwork, this hypothesis was not supported. There were indeed more occurrences of MI online than in print, but this relationship was not significant, t(761) = 1.60, p = .11. When all the online-specific links were excluded, and only MI in text was included (i.e., the body of the story without any links), print newspapers had slightly more MI in text than online newspapers. Again, results were below the standards for marginal significance, f(761) = -1.59, p = .11. RQl asked whether one type of MI was more prevalent in print and online newspapers. There was significantly more identificational MI than locational MI online, f(378) = -23.24, p <.0005. Identificational MI also occurred more frequently in print than locational, ^383) = -36.23, p < .0005. Online, identificational MI had significantly more occurrences than tactical MI, f(378) = 36.00, p <.0005. In print, identificational MI occurred significantly more frequently than tactical MI, f(383) = 35.77, p < .0005. The comparison between locational and tactical MI yields the only nonsignificant difference, and this is specific to print. There were no more occurrences of locational MI than tactical MI in print, f(383) = -.38, p = .701. The difference online, however, was significant, with more locational than tactical MI, f(378) = 2.85, p < .005. Thus, identificational MI occurred with more frequency in both print and online contexts than either locational or tactical MI, and locational MI occurred more frequently online than in print. However, this last difference is most likely due to the enormity of the San Francisco Chronicle's archival links in its "ChronicleWatch." Table 2 lists the differences in the mean frequencies of MI by both media type and newspaper.

louRNALisM & MASS COMMUNICATION QUARTERLY

TABLE 2 Three Types of Mobilizing Information by Media Type and Newspaper Online

n Type:

CD 58

LMI Mean .09 s.d. .28

Print

NYT 64

WP 64

PI 65

SFC 65

DP 65

CD 62

NYT 65

WP 64

PI 65

SFC 64

DP 64

Total

.30 .46

.23 .66

.05 .21

3.34 10.4

.08 .33

.10 .30

.02 .12

.13 .55

.02 .12

.03 .18

.09 .34

.38 3.17

7.63 3.57

8.91 4.86

8.17 3.90

8.85 4.26

7.62 4.96

8.94 5.00

7.60 4.10

8.19 3.25

9.00 4.71

8.23 4.37

7.00 4.63

8.16 4.39

.08 .27

.03 .18

.06 .30

.12 .38

.08 .27

.03 .18

.09 .63

0

.06 .24

.09 .29

.16 .80

.07 .37

Dvn Mean 7.79 4.58 s.d.

TMI Mean s.d.

0

Note: CD = Columbus Dispatch, NYT = New York Times, WP = Washington Post, PI = Philadelphia Inquirer, SFC = San Francisco Chronicle, DP = Denver Post. LMI = locational mobilizing information, IMI = identificational mobilizing information, TMI = tactical mobilizing information. Unit of analysis = article.

Online newspapers do not differ significantly from print newspapers in terms of presence of mobilizing information, and look much like the "shovelware" versions Singer suggested.'' The relative similarity between print and online versions of the newspaper—even accounting for value-added items such as links to archived content, audio/video links, and discussion boards—is startling. This was not a print-to-print or text-only comparison; on the contrary, this analysis took into account those elements that make online newspapers different from print newspapers. This would suggest that the same fears Lemert cited as reasons journalists exclude MI in print newspapers might also apply to those publishing online.'" That said, however, online newspapers are certainly not the only political and news content available online. There are countless government and interest-group Web sites, discussion boards, and "blogs," as well as nontraditional news sources unavailable in other media formats. Moreover, non-newspaper sites may create "runoff" to newspaper sites." Therefore, although this study was limited to print and online newspapers, it provides a first step in examining MI online. In addition to journalistic fears of endorsement, partisanship, or libel, there might be other factors acting to keep online newspapers similar to print versions.™ It has been argued that powerful media corporations establish norms and routines that keep journalists deferential to the media structure and that normalize content."' Journalists who adhere to these routines are considered professional and rewarded, while those who oppose them might be reprimanded."^ The most widely discussed Is INTERNET CONTENT

DIFFERENT AFTER

ALL?

Discussion

67

of these norms is objectivity or neutrality—the desire to be fair and present "both sides" of an issue—a norm particularly widespread in American journalism."^ The adherence to norms could deter journalists from publishing MI online that threatens the status qiio."" However, online newspapers might simply be seeking consistency with their print counterparts in order to maintain a common voice. Interviews with online journalists could provide useful insights into these dynamics. This study operationally distinguished among the three types of MI—locational, identificational, and tactical. Past research has focused on the reasons MI is excluded from news, and recent work on political participation that alludes to, or sometimes assumes, the presence of MI in news often fails to differentiate among the three types. The near exclusion of tactical and locational MI in this study indicates that explicit MI is rare in both print and online newspapers. This raises questions about the validity of differentiating among these types that occur so infrequently. However, the volume of identificational MI begs important new questions: Is identification of a political individual enough to mobilize action on the part of the reader? Moreover, in this era—when to find more information, all a person has to do is "Google" someone— are mobilizing possibilities greater than they were before these technologies were available? Surveys and experiments with online users could provide us with answers. Future studies could expand on this analysis by conducting similar research in health, environmental, and other contexts. Health communication researchers have already adopted the MI concept, although the conceptual and operational definitions used here have diverged substantially from those in the political and environmental contexts."^ A useful study would apply the same operationalizations to each context. Additionally, Lemert and Ashman, as well as Dunwoody and Griffin, examined mobilizing content in controversial or negative contexts, suggesting that tone also influences inclusion of MI."^ A second, and laudable, addition to this content analysis would be an expanded examination of all content in online newspapers, including that found on discussion boards, chat rooms, hyperliriked sites, and audio/video clips. This information is inherently different from what is in print newspapers. An expanded sample of newspapers including more regional variation might also find more differences. Examining news content on blogs, portal sites, and search engines could also yield new and interesting findings. Additionally, as Tremayne found. Web sites affiliated with broadcast networks are more "linked" than those affiliated with newspapers. An interesting direction for research would be to compare these two in terms of their mobilizing information."' Beyond examining MI in content, researchers could assess the cognitive processes involved in reading and understanding MI. Eveland and Scheufele argued that those in a community with higher SES are more likely to see MI and, perhaps more important, "they possess the information processing strategies and cognitive skills necessary to identify mobilizing information, even if it is less prominently placed or men-

68

JOURNALISM & MASS

COMMUNICATION

QUARTERLY

tioned only implicitly."** Individual and socioeconomic differences could play a key role in the usefulness of MI. Finally, a reliability assessment for inclusion of political news in the sample would make the interpretation of these results more powerful.

Print newspaper readership has consistently been correlated with higher rates of political participation, voter turnout, and civic engagement. If Internet news use is on the rise while print news use declines, what does this mean for the link between political participation and media use? This study suggests that online newspapers provide content that simply reinforces print content. But even if mobilizing information in print and online newspapers is not significantly different, the effects may differ across readers. Because this study examines only content, effects cannot be ascertained, but this is clearly an area for further research. Indeed, the Internet could draw new readers who would not have otherwise been exposed to mobilizing content. In this case, more people may be exposed to more mobilizing information, despite declines in print newspaper readership. Some researchers have noted that higher ievels of Internet use are associated with higher levels of political participation."' Johnson and Kaye asserted that Internet users represent "model citizens" and are more likely to be politically interested, efficacious, and seek information from the media.'" Moreover, evidence suggests that both gender and race gaps in Internet use are closing or closed, and children increasingly use the Internet at school and home.'' Improved understanding of how individuals use the Internet, as well as what content they view, will clearly add meaning to the present study's findings. The changing nature of the Internet environment demands that researchers take into account how both its audience and its content must continually be monitored.

Conclusion

NOTES 1. Ruy A. Teixeira, Why Americans Don't Vote: Turnout Decline in the United States, 1960-1984 (New York: Greenwood Press, 1987); Sidney Verba, Kay. L. Schlozman, and Henry E. Brady, Voice and Equality: Civic Voluntarism in American Politics (Cambridge: Harvard University Press, 1995); Dietram A. Scheufele and Matthew C. Nisbet, "Being a Citizen Online: New Opportunities and Dead Ends," Harvard International Journal of Press-Politics 7 (June 2002): 55-75; Robert D. Putnam, "Bowling Alone: America's Declining Social Capital," Journal of Democracy 6 Qanuary 1995): 65-78. 2. Robert D. Putnam, Bowling Alone: The Collapse and Revival of American Community (New York: Touchstone, 2000). 3. Pew Research Center for the People and the Press, "News Audiences Increasingly Politicized: Online News Audience Larger, More Diverse," (8 June 2004). Is INTERNET CONTENT DIFFERENT AFTER ALL?

69

4. See Bruce Bimber, "The Internet and Political Transformation: Populism, Community, and Accelerated Pluralism," Polity 31 (1998): 133-60; Thomas J. Johnson and Barbara K. Kaye, "A Vehicle for Engagement or a Haven for the Disaffected? Internet Use, Political Alienation, and Voter Participation," in Engaging the Public: How Government and the Media Can Reinvigorate American Democracy, ed. Thomas J. Johnson, Carol E. Hays, and Scott P. Hays (Lanham: Rowman & Littlefield, 1998), 123-35. 5. Robert Kraut, Michael Patterson, Vicki Lundmark, Sara Kiesler, Tridas Mukopadhyay, and William Scherlis, "Internet Paradox: A Social Technology That Reduces Social Involvement and Psychological WellBeing?" American Psychologist 53 (September 1998): 1017-31; Putnam, "Bowling Alone: America's Declining Social Capital." 6. Robert Kraut, Sara Kiesler, Bonka Boneva, Jonathan Cummings, Vicki Helgeson, and Anne Crawford, "Internet Paradox Revisited," journal of Social Issues 58 (January 2002): 49-74. 7. Kraut et al., "Internet Paradox Revisited," 68. 8. Pew Research Center, "News Audiences Increasingly Politicized." Figure is based on those who reported going online for news at least three days per week. 9. Pew Internet & American Life Project, "America's Online Pursuits: The Changing Nature of Who's Online and What They Do," (22 December 2003); see also Pew Internet & American Life Project, "Modest Increase in Internet Use for Campaign 2002: Political Sites Cain, but Major News Sites Still Dominate," (5 January 2003); and Bruce Bimber, Information and American Democracy: Technology in the Evolution of Political Power (Cambridge: Cambridge University Press, 2003). 10. See Herbert Blumer, "Suggestions for the Study of Mass-Media Effects," in American Voting Behavior, ed. Eugene Burdick and Arthur A. Brodbeck (New York: Free Press-Macmillan 1959), 183-94; Steven H. Chaffee, "Mass Media Effects: New Research Perspectives," in Communication Research: A Half-Century Appraisal, ed. Daniel Lerner and Lyle M. Nelson (Honolulu: University Press of Hawaii, 1977), 210-41; Pippa Norris, "Does Television Erode Social Capital? A Reply to Putnam," PS: Political Science and Politics 29 (September 1996): 474-80. 11. See Putnam, "Bowling Alone: America's Declining Social Capital"; Putnam, Bowling Alone. 12. Blumer, "Suggestions." 13. Christopher Weare and Wan-Ying Lin, "Content Analysis of the World Wide Web: Opportunities and Challenges," Social Science Computer Review 18 (August 2000): 272-92. 14. Erik P. Bucy, "Second-Generation Net News: Interactivity and Information Accessibility in the Online Environment" (unpublished manuscript). 15. Daniel Riffe, Stephen Lacy, and Frederick G. Fico, Analyzing Media Messages: Using Quantitative Content Analysis in Research (Mahwah, NJ: Lawrence Erlbaum, 1998), 32. 70

JOURNALISM b MASS COMMUNICATION QUARTERLY

16. James B. Lemert, Does Mass Communication Change Public Opinion After All?: A New Approach to Effects Analysis (Chicago: Nelson-Hall, 1981); James B. Lemert and J. Larkin, "Some Reasons Why Mobilizing Information Fails to Be in Letters to the Editor," Journalism Quarterly 56 (autumn 1979): 504-12; James B. Lemert, Barry N. Mitzman, Michael A. Seither, Roxana H. Cook, and Regina Hackett, "Journalists and Mobilizing Information," journalism Quarterly 54 (winter 1977): 721-26. 17. Merrill Morris and Christine Ogan, "The Internet as Mass Medium," Journal of Communication 46 (winter 1996): 42. 18. William P. Eveland, Jr., "A 'Mix of Attributes' Approach to the Study of Media Effects and New Communication Technologies," Journal of Communication 53 (September 2003): 395-410. 19. John V. Pavlik, Journalism and New Media (New York: Columbia University Press, 2001). 20. Morris and Ogan, "The Internet as Mass Medium," 43. 21. Pablo J. Boczkowski, Digitizing the News: Innovations in Online Newspapers (Cambridge: The MIT Press, 2004). 22. Boczkowski, Digitizing the News, 183. 23. Bruce Bimber, "Information and Political Engagement in America: The Search for Effects of Information Technology at the Individual Level," Political Research Quarterly 54 (March 2001): 53-67. 24. Kraut et al., "Internet Paradox" and "Internet Paradox Revisited"; Paul Rich, "American Voluntarism, Social Capital, and Political Culture," The Annals of the American Academy of Political and Social Science 565 (1999): 15-34. 25. Richard Davis, The Web of Politics: The Internet's Impact on the American Political System (New York: Oxford University Press, 1999). 26. Putnam, Bowling Alone; Barry Wellman, "Computer Networks as Social Networks," Science 293 (September 2001): 2031-34. 27. Johnson and Kaye, "A Vehicle"; Dhavan V. Shah, Nojin Kwak, and R. Lance Holbert, "'Connecting' and 'Disconnecting' with Civic Life: Patterns of Internet Use and the Production of Social Capital," Political Communication 18 (April 2001): 141-62; Dhavan V. Shah, Jack M. McLeod, and So-Hyang Yoon, "Communication, Context, and Community: An Exploration of Print, Broadcast, and Internet Influences," Communication Research 28 (August 2001): 464-506; Mira Sotirovic and Jack M. McLeod, "Values, Communication Behavior, and Political Participation," Political Communication 18 Quly 2001): 273-300. 28. Shah, McLeod, and Yoon, "Communication, Context, and Community." 29. Scheufele and Nisbet, "Being a Citizen Online." See also Davis, The Web of Politics, and Johnson and Kaye, "A Vehicle." 30. Scheufele and Nisbet, "Being a Citizen Online," 57. 31. Lemert, Does Mass Communication Change. 32. William P. Eveland, Jr. and Dietram A. Scheufele, "Connecting News Media Use with Caps in Knowledge and Participation," Political Communication 17 (July 2000): 215-37; Bruce Hardy, "Mobilizing Information Online: The Effects of Primary-Source and SecondarySource Website Use on Political Engagement" (paper presented at the Is INTERNET CONTENT DiffERENT AFTER ALL?

71

annual meeting of the Midwest Association for Pubiic Opinion Research, Chicago, IL, November 2004); Jack M. McLeod, Dietram A. Scheufele, and Patricia Moy, "Community, Communication, and Participation: The Role of Mass Media and Interpersonal Discussion in Local Political Participation," Political Communication 16 (July 1999): 31536; Zhondang Pan and Jack M. McLeod, "Multilevel Anaiysis in Mass Communication Research," Communication Research 18 (April 1991): 14073; Michael Schudson, The Power of News (Cambridge: Harvard University Press, 1995); Shah, Kwak, and Holbert, "'Connecting' and 'Disconnecting.'" 33. Aian M. Frager, and Alice Khan, "How Useful Are Elementary School Health Textbooks for Teaching about Hearing Heaith and Protection?" Language, Speech, and Hearing Services in Schools 19 (Aprii

1988): 175-81; Laurie Hoffman-Goetz, Charlene Shannon, and Juanne N. Clarke, "Chronic Disease Coverage in Canadian Aboriginal Newspapers," Journal of Health Communication 8 (September 2003): 47588; Megan MacDonald and Laurie Hoffman-Goetz, "Cancer Coverage in Newspapers Serving Large and Small Communities in Ontario," Canadian Journal of Public Health 92 (September/October 2001): 372-75. 34. Lemert, Does Mass Communication Change; Lemert et al., "Journalists"; Lemert and Larkin, "Some Reasons"; S. Hungerford and James B. Lemert, "Covering the Environment: A New "Afghanistanism'?" Journalism Quarterly 50 (autumn 1973): 475-81, 508; James B. Lemert, "News Context and the Elimination of Mobilizing Information: An Experiment," Journalism Quarterly 61 (summer 1984): 243-49, 259; James B. Lemert and Marguerite Gemson Ashman, "Extent of Mobilizing Information in Opinion and News Magazines," Journalism Quarterly 60 (winter 1983): 657-62; James B. Lemert and Roxana Cook, "Mobilizing Information in Broadcast Editorials and 'Free Speech' Messages," Journal of Broadcasting 26 (winter 1982): 493-96; Douglas W. Stanfield and James B. Lemert, "Alternative Newspapers and Mobilizing Information," Journalism Quarterly 64 (summer/autumn 1987): 604-7. 35. Arvin W. Murch, "Public Concern for Environmentai Pollution," Public Opinion Quarterly 35 (spring 1971): 100-6. 36. David M. Rubin and David P. Sachs, Mass Media and the Environment, Volume II. The Environmental Information Explosion: The Press

Discovers the Environment, Stanford University (ERIC Document Reproduction Service No. ED058710,1971). 37. Rubin and Sachs, Mass Media, 9. 38. Rubin and Sachs, Mass Media, 32. 39. Lemert, Does Mass Communication Change. 40. Lemert, Does Mass Communication Change, 118. 41. This difficulty is apparently what led Lemert and his coiieagues to study alternative newspapers (Stanfield and Lemert, "Aiternative Newspapers"); environmentai news (Hungerford and Lemert, "Covering the Environment"); pubiic service announcements (Lemert and Cook, "Mobilizing Information"); liberal, conservative, and news magazines (Lemert and Ashman, "Extent of Mobilizing Information"); 72

JOURNALISM & MASS COMMUNICATION

QUARTERLY

and to interview journalists to discover their reasons for excluding MI (Lemert, Does Mass Communication Change, and Lemert, "News Context"). 42. Lemert, Does Mass Communication Change, 134; see also Sharon Dunwoody and Robert J. Criffin, "Structural Pluralism and Media Accounts of Risk," in Mass Media, Social Control & Social Change, ed. David Demers and K. Viswanath (Ames, IA: Iowa State University Press, 1999), 139-58. 43. Lemert, Does Mass Communication Change, and Lemert, "News Context." 44. Lemert and Larkin, "Some Reasons"; see also Ralph R. Thrift, Jr., "How Chain Ownership Affects Editorial Vigor of Newspapers," Journalism Quarterly 54 (summer 1977): 327-31. 45. Jane B. Singer, "The Metro Wide Web: Changes in Newspapers' Gatekeeping Role Online," Journalism & Mass Communication Quarterly 78 (spring 2001): 65-80. 46. Pavlik, Journalism and New Media; Eric S. Fredin and Prabu David, "Browsing and the Hypermedia Interaction Cycle: A Model of SelfEfficacy and Coal Dynamics," Journalism & Mass Communication Quarterly 75 (spring 1998): 35-54. 47. David Tewksbury and Scott L. Althaus, "Differences in Knowledge Acquisition Among Readers of the Paper and Online Versions of a National Newspaper," Journalism & Mass Communication Quarterly 77 (autumn 2000): 457-79. 48. Daniela V. Dimitrova, Colleen Connolly-Ahern, Andrew Paul Williams, Lynda Lee Kaid, and Amanda Reid, "Hyperlinking as Catekeeping: Online Newspaper Coverage of the Execution of an American Terrorist," Journalism Studies 4 (August 2003): 401-14. 49. Steve Oudng, "Newspapers' Taste Standards Get Loose Online," Editor and Publisher, (28 May 2003). 50. Outing, "Newspapers' Taste." 51. Outing, "Newspapers' Taste." 52. Morris and Ogan, "The Internet as Mass Medium." 53. Michael A. Xenos and Kirsten A. Eoot, "Politics As Usual, or Politics Unusual? Position Taking and Dialogue on Campaign Websites in the 2002 U.S. Elections," Journal of Communication 55 (March 2005): 16985. 54. Ben H. Bagdikian, "The U.S. Media: Supermarket or Assembly Line?" Journal of Communication 35 (summer 1985): 97-109. 55. Daniel Riffe, Charles F. Aust, and Stephen R. Lacy, "The Effectiveness of Random, Consecutive Day and Constructed Week Sampling in Newspaper Content Analysis," Journalism Quarterly 70 (spring 1993): 133-39; Riffe, Lacy, and Fico, Analyzing Media Messages. 56. Riffe, Lacy, and Fico, Analyzing Media Messages. 57. This time period was selected because it was seen as likely to have more occurrences of MI related to politics and voting, as regional elections were taking place. Is /NTERNET CONTENT DIFFERENT AFTER ALL?

73

58.1 selected 1:30 a.m. (EST) as the download time, just after press time for most print newspapers. This allows for maximum comparability between print newspapers and their online counterparts (William P. Eveland, Jr. and A. Timiraos, "Variations in Story Frequency and Prominence in Web and Print Newspapers" [unpublished manuscript, 2001]). 59. Exhaustive research and an attempt to contact Lemert yielded no published or unpublished coding criteria for MI. These criteria are based on conceptual and operational explications in previous research. 60. Richard M. Perloff, Political Communication: Politics, Press, and Public in America (Mahwah, NJ: Lawrence Erlbaum Associates, 1998), 7. 61. A coding instrument identified the section for the coder, as some newspaper sections differed in name in online and print versions. If the story did not indicate a section, the dateline and/or byline were used to identify story type. For example, a dateline of "CHICAGO" in the Columbus Dispatch would be considered National, and a byline of "John Wildermuth, Chronicle Political Writer" would be considered Politics. If the section was split (e.g., "National/International"), datelines and bylines were utilized to assess the nature of the article origin. 62. A. S. Twyman, "Council Eyes Campaign Donations," The Philadelphia Inquirer, 18 November 2003, sec. B, p. 6. 63. These took the form of "Contact staff writer Carrie Budoff at 610313-8211 or [email protected]," Carrie Budoff, "Debt Pushes Pittsburgh to the Brink" The Philadelphia Inquirer, 7 November 2003, sec. A, pp. 1,18. 64. Lemert, Does Mass Communication Change. 65. Chronicle Watch, The San Francisco Chronicle, (2004). 66. Spidersoft, "WebZIP (Version 5.0.5.896)" [Computer software]. (2002). 67. Franz Faul and Edgar Erdfelder, "GPOWER: A Priori, Post-Hoc, and Compromise Power Analyses for MS-DOS [Computer software]" (Bonn, FRG: Bonn University, Dep. Of Psychology, 1992). 68. With an effect size at .20 (anticipating smail effects), alpha at .05, and power set at .80, a t-test group comparison would require a combined (both online and print) sample size of approximately 620 articles. In order to reach this sampie size overall, I estimated that approximately 65 articles per newspaper (each for print and oniine versions) would be sufficient. However, not every issue had 65 articies. The Columbus Dispatch, for example, had fewer than 65 articles in the original sample that actually fit the coding criteria. Additionally, there were a few cases (seven total) in which articles were excluded during the coding period because they were shown not to fit the coding criteria (e.g., a wireauthored article discovered only after coding had begun). After assessing power, a set of random numbers was created for each newspaper, and the first 65 numbers (without replacement) were entered into the sample. 69. In this case, the number of articles coded for reliability was 87, or 11.4% of the total sample. Roger D. Wimmer and Joseph R. Dominick, Mass Media Research: An Introduction, 5th ed. (Belmont, CA: Wadsworth, 74

JOURNALISM & MASS COMMUNICATION QUARTERLY

1997). 70. Klaus Krippendorf, "Krippendorf's Reliability (Version 3.12a) [Computer software]" (Philadelphia: Annenburg School for Communication, 1996); Klaus Krippendorf, Content Analysis: An Introduction to Its Methodology (Beverly Hills: Sage, 1980). 71. Krippendorf, Content Analysis; Riffe, Lacy, and Fico, Analyzing Media Messages. 72. Kim A. Neuendorf, The Content Analysis Guidebook (Thousand Oaks, CA: Sage, 2002). 73. Lemert and Cook, "Mobilizing Information," 495. 74. Stanfield and Lemert, "Alternative Newspapers"; Lemert, "News Context"; Lemert and Ashman, "Extent of Mobilizing Information"; Lemert and Cook, "Mobilizing Information"; Lemert et al., "Journalists"; Hungerford and Lemert, "Covering the Environment." 75. Hungerford and Lemert, "Covering the Environment;" Lemert et al., "Journalists." 76. Stanfield and Lemert, "Alternative Newspapers." 77. Singer, "The Metro Wide Web." 78. Lemert, Does Mass Communication Change. 79. See, for example, Andrew Paul Williams, Kaye D. Trammeil, Monica Postelnicu, Kristen D. Landreville, and Justin D. Martin, "Blogging and Hyperlinking: Use of the Web to Enhance Viability During the 2004 U.S. Campaign," Journalism Studies 6 (May 2005): 177-86. 80. Lemert, Does Mass Communication Change. 81. Davis, The Web of Politics; Michad Margolis and David Resnick, Politics as Usual: The Cyberspace "Revolution" (Thousand Oaks, CA: Sage, 2000); Pippa Norris, Digital Divide: Civic Engagement, Information Poverty, and the Internet Worldwide (Cambridge: Cambridge University Press, 2001). 82. W. Lance Bennett, News: The Politics of Illusion, 4th ed. (New York: Longman, 2001); Warren Breed, "Social Control in the Newsroom: A Functional Analysis," Social Forces 33 (May 1955): 326-55; Pamela J. Shoemaker and Stephen D. Reese, Mediating the Message: Theories of Influence on Mass Media Content, 2d ed. (White Plains, NY: Longman, 1996); David H. Weaver and G. Cleveland Wilhoit, The American Journalist in the 1990s: U.S. Newspeople at the End of an Era (Mahwah, NJ: Lawrence Erlbaum, 1996). 83. Thomas E. Patterson and Wolfgang Donsbach, "News Decisions: Journalists as Partisan Actors," Political Communication 13 (October 1996): 455-68. 84. John Soloski, "News Reporting and Professionalism: Some Constraints on the Reporting of the News," Media, Culture & Society 11 (April 1989): 207-28. 85. Randal A. Beam and James Bernstein, "The Impact of Channel Reliance on Tactical Mobiiizing Information about HIV Transmission" (paper presented at the arinual meeting of AEJMC, Washington, DC, August 1989); Frager and Kahn, "How Useful"; Daniela B. Friedman and Laurie Hoffman-Goetz, "Cancer Coverage in North American Publications Targeting Seniors," Journal of Cancer Education 18 (March ;s INTERNET CONTENT DIFFERENT AFTER ALL?

75

2003): 43-7; Hoffman-Goetz, Shannon, and Clarke, "Chronic Disease Coverage"; Corvnie M. Kristiansen and Christina M. Harding, "Mobilization of Health Behavior by the Press in Britain," Journalism Quarterly 61 (summer 1984): 364-70, 398; Barbara Straus Reed, "The Link between Mobilizing Information and Service Journalism as Applied to Women's Magazine Coverage of Eating Disorders" (paper presented at the annual meeting of AEJMC, Minneapolis, MN, August 1990). 86. Lemert, Does Mass Communication Change; "News Context"; Lemert and Ashman, "Extent of Mobilizing Information"; Dunwoody and Griffin, "Structural Pluralism." 87. Mark Tremayne, "The Web of Context: Applying Network Theory to the Use of Hyperlinks in Journalism on the Web," Journalism & Mass Communication Quarterly 81 (summer 2004): 237-53. 88. Eveland and Scheufele, "Connecting News Media Use," 221. 89. M. Kent Jennings and Vicki Zeitner, "Internet Use and Civic Engagement: A Longitudinal Analysis," Public Opinion Quarterly 67 (fall 2003): 311-34; Lori M. Weber, Alysha Loumakis, and James Bergman, "Who Participates and Why? An Analysis of Citizens on the Internet and the Mass Public," Social Science Computer Review 21 (February2003): 26-42. 90. Johnson and Kaye, "A Vehicle." 91. Pew Research Center, "News Audiences Increasingly Politicized."

76

JOURNALISM & MASS COMMUNICATION QUARTERLY