Persuasive Essay
lovlyl
media_credibility_reconsidered.pdf
MEDIA CREDIBILITY RECONSIDERED:
SYNERGY EFFECTS BETWEEN ON-AIR
AND ONLINE NEWS
By Erik P. Bucy
This experimental investigation of media credibility examined the combined, or synergistic, effects of on-air and online network news exposure, placing student and adult news consumers in broadcast news, online news, and teleivebbing conditions. Results indicate that perceptions of network news credibility are affected by channel used. Perceptions of credibility were enhanced when the channel used was consistent with the news source being evaluated, suggesting a channel congruence effect. In addition, evidence is offered for the existence of a synergy effect between on-air and online news.
Faced with unprecedented competitive pressures from new challengers and new technology, network news organizations are adjusting to the fact that a growing audience for news relies primarily on the Internet rather than television.' As the viewership for the "big three" network news broadcasts continues to slide to roughly a third of the overall evening television audience, down from 72% in the early 1980s, the audience for Net news is expanding; about one in three Americans is now going online for news at least once a week.^ As more people log on for news, fewer are tuning in.̂ The decline in nightly network news viewing is reportedly greater for Internet users than non- users; only 26% of Internet users report regularly watching a nightly network news broadcast, compared to 35% of nonusers who tune in regularly.''
At the same time, however, an increasing number of television viewers with Internet access are telewebbing, or surfing the Web while watching television (in so-called "two-screen" households). A recent online survey of some 2,000 Internet users by Scarborough Research found that half of all online respondents reported having a television in the same room as their computer.' Within these two-screen households, 91% of respondents said they watched television and surfed the Web simultaneously. Estimates by the industry research firm Dataquest placed the American telewebbing population at 52 million as of 2001, up from 27 million in 1999.' Younger adults aged 18-34 are the most enthusiastic telewebbers. In the Scarborough survey, more than a
Erik P. Bucy is an assistant professor in the Department of Telecommunications and adjunct professor in the Schoo! oflnforttiaiics, Indiana Universitii, Btoomington.
Introduction
Vol. SO, No. 2
Sl>nifiier2IMa
247
quarter (27%) reported going online and "always or often" watching television at the same time.' Younger audiences are also more likely to turn to the Internet for news. Almost half of those under age 30 (46%) go online for news at least once a week, compared to just 20% of those age 50 and over.* Older Americans are far more likely to say they get their news from television or the newspaper.
Network news operations were among the first electronic media to realize the importance of establishing an online presence and in the mid-1990s committed to developing their sites relatively early in the Web's development.' Major broadcast news sites are now among the most trusted and heavily visited information sources online. In fact, many Internet users consider the networks' online news to be more believable than the parent news organization.^" Given that assess- ments of media credibility are strongly, and have been historically, associated with use of and reliance on a particular media channel," the believability accorded to Net news by people who are already online is not surprising. But the surge in online news credibility among Internet users does not necessarily imply a corresponding decrease in television news credibility among iheoverali news audience. Credibility perceptions do not behave in a zero-sum fashion.'^ Instead, to the extent that online media are being used as a complement to broadcast media, network news organizations may stand to benefit from their online operations.
Nevertheless, in recent years other coverage-related factors have called the credibility of network news into question. During the past decade, amidst accusations of media bias, misplaced news priorities, and sensationalism, credibility has resurfaced as an issue of central importance to news division operations. The miscalled 2000 presidential election and overplayed Chandra Levy and Monica Lewinsky stories brought this point home in pronounced fashion. When events this high profile are mishandled, news media come under intensepublic scrutiny, raising questions about their authority to broadcast the news as they see fit. Yet while the networks were roundly criticized for getting the election story wrong,'' and then for manufacturing lurid interest in an alleged affair between California Congressman Gary Condit and his missing intern, they were generally praised for coverage of the World Trade Center attacks." In either scenario, the depth and interactivity afforded by the networks' online operations, which allow users to mine the news for more information at their own pace and according to individualized interests, might have the effect of mitigating negative perceptions of media credibility while enhancing already positive evaluations of news veracity and network professionalism.
Defining Media Credibility. For conceptual clarity, media credibility canbedefined as perceptions of anewschannel'sbelievability, as distinct from individual sources, media organizations, or the content of the news itself.̂ ^ Media credibility differs from source credibility, which focuses on characteristics of message senders or individual speakers, such as trustworthiness and expertise. Each tradition has its own history of research, with the former stemming from studies on newspaper reporting accuracy and the latter emerging from the attitude change studies of Carl Hovland and associates." Since the introduction
248
of the Roper questions about relative media credibility in the late 1950s, the measurement of media credibility has been vigorously debated." Subsequent research has shown that the way the concept is operationalized and measured influences credibility ratings among respondents.^' When measured as a single perceptual dimension, media credibility is most consistently operationalized as believability." Other dimensions used to measure the concept include accuracy, bias, fairness, and completeness of information, among others.^ A multiple dimension approach to measuring credibility is now the norm in academic research.
Given the ongoing traffic-building efforts by broadcasters to promote "enhanced television" and coax viewers into visiting their Web sites to avail themselves of online information, the question arises as to whether cross-platform media use is in fact influential in cultivating perceptions of broadcast news credibility and, by extension, in maintaining positive audience evaluations of network news. While survey results vary, studies by the Pew Research Center and Online News Association^' are finding that many online users judge the Internet as a news medium to be as or more credible than traditional media, including both print and broadcast. To assess whether there are enhanced benefits from cross-platform media use, this study compares perceptions of TV and Net news credibility after exposure to online news, broadcast news, and a combined, or telewebbing, condition.
Investigating Synergy Effects, Synergy, a marketing term, refers to the benefits derived from selling two or more compatible products simultaneously (e.g., a blockbuster movie, the book on which the movie is based, the movie soundtrack, and action figures derived from the story), as well as the harnessing of relations between two or more elements of a media production and distribution process, including the repurposing of content created for one medium for distribution by another.^ In the context of this study, synergy refers to the possibility of two contiguous media behaviors—on air and online news consumption—having a greater impact on perceptions of media credibility than exposure to either medium by itself. Although teiewebbing emerged as a mass phenomenon in the late 1990s,^ the effects of cross-platform media use remain underexamined and little understood.
In addition to examining synergy effects in a real-time setting, this study adds to a small but growing body of research that has applied the experimental method to questions of television news credibility.^^ The preponderance of studies on media credibility has been survey based,^ and most charmel comparison studies have examined the credibility of television news in relation to newspapers.^ The survey approach to studying credibility allows researchers to track broad trends in public opinion about the media but the results of such surveys are limited in their ability to explain cause-and-effect relationships.
Survey research has established that use of a particular media channel correlates with high credibility ratings for that channel.^ But does media use lead to enhanced perceptions of credibility, or do people tend to use media they already consider credible? Survey studies, because they lack control and generally do not address the issue
249
of temporal sequence, carmot effectively answer this question. Surveys, moreover, are hampered in a data-gathering sense because they rely on reflections of past media use, an estimation process prone to distortion and social desirability biasing. Placing participants directly in different media settings (e.g., telewebbing, television only, or Net news only) offers the opportunity to assess perceptions of credibility as they occur rather than relying on reflections of past media behavior. The experimental situa tion also enables the researcher to control the sequence of presumed effects.
Thedocumentation of synergy effects on percep tions of credibility could have important implications for the broadcast news industry, and network news organizations in particular, as audiences tend to pay more attention to and become more reliant on media they consider credible.^* Perceptions of credibility have also been found to influence the processing of issue importance. When media are perceived as credible, acceptance of information veracity increases and agenda- setting effects—accepting issues given prominent media coverage as valid and important—are likely to occur.^' In an era of convergent media and twenty-four-hour news channels, when the future of the nightly newscast itself is being called into question,* the networks are challenged with maximizing their believability, accuracy, fairness, and informational completeness across delivery platforms.
Media Credibility and Age Cohort. Studies of media credibility have consistently fo\md an association between age and education with credibility assessments. In general, older, more educated audiences tend to be the most critical of media, while younger, less educated news consumers are more likely to be accepting of news coverage and to evaluate the media as credible.'^ The youngest adults, 18-24 years old, are the most likely to rate news media highly believable. And it is network news, more than most other forms of media, that has been likely to elicit these demographic response patterns.'^ Sophistication, life experience, and knowledge of the press—a type of news literacy— combine to make more seasoned audiences skeptical of the nightly
Although displacement theory would suggest that a traditional medium such as television, besieged by new challengers and new communication technologies, would directly lose itsaudience to popular forms of new media, some studies conclude that the Internet supplements rather than replaces television viewing.*" These findings zie consistent with the idea that online sources provide what industry players Uke ABC-TV refer to as an "enhanced television" experience. Robinson and colleagues even found a positive relationship between computer use and television viewing.^'Other researchers haveargued, however, that television viewing is negatively affected by computer or internet use.* None of the aforementioned studies, however, explicitly considers telewebbing as a media behavior that might influence the interpretation of these findings.
In a time of rapid technological change and format experi- mentation, credibility remains central to understanding public perceptions of network news as well as encouraging acceptance of the
250 /OUKNAUSM & MASS CoMMlpacxnoN
Internet as a trusted source of news and information. Documenting the potential contributions of online news operations to perceptions of network news credibility through experimental investigation may enhance industry and scholarly awareness of the nonmonetary value of the networks' online news operations beyond boftom-line issues of profitability and audience share. If such effects are found, this will suggest to the industry that investments in online news operations may be justified not on the basis of economic considerations alone but on the grounds of enhancing public perceptions of network news through multiple media channels.
To assess whether cross-platform media use of online and on-air Research news has a greater impact on audience perceptions of media credibility Questions than exposure to either medium in isolation, this study of synergy effects is guided by the following research questions:
RQl: Do different audiences for news, specifically younger and older age groups, rate the credibility of on-air and online news differently?
RQ2; What influence does media channel have on audience evaluations of media credibility?
RQ3: Is there a combined, or synergy, effect between on-air and online news sources that enhances perceptions of credibility over use of either medium in isolation?
RQ4: Is there an interaction between media channel and age group resulting in different synergy effects for younger and older news audiences?
Operationally, this study took the form of a 4 (media channel) x 2 (age group) between subjects, factorial experiment. The first factor, media channel, had four levels: television news. Net news, both TV and Net news, and no exposure (control group). The second factor, age group, had two levels: imdergraduate students and older adults. Student subjects consisted of 84 undergraduates enrolled in communications courses at a large Midwestern university and ranged in age from 18 to 25 (M=20). Adult subjects consisted of 83 members of a local community group and ranged in age from 26 to 80 (M = 49). A total of 167 subjects participated in the study, populating each cell of the design with a minimum of 20 subjects. Student subjects received extra credit for their participation, while adults received a small monetary incentive. Slightly more than half of the study participants (52.1%, n = 87) were female, while just under half (47.9%, n = 80) were male.
Procedure. The study was conducted in the weeks immediately following the September 11 terrodstattacks.The news focus on terrorism remained prominent for the duration of the study, although the early
R E D / B I U T Y R£CONSff)EB£D
Method
251
emphasis on domestic security issues gradually gave way to a focus on the war in Afghanistan. Within each age group, subjects were randomly assigned to one of the four media channel conditions. To guard against primacy and recency effects, each condition had four randomized orders. At the beginning of the experiment, subjects completed a questionnaire asking about their media use, political attitudes and knowledge, and demographic information.
Subjects assigned to the television condition were shown four, randomly ordered network news stories (from ABC, NBC, CBS, or MSNBC) about the World Trade Center attacks and were asked to rate their responses. The stories were recorded off-air from network news coverage beginning on the day of the attacks and for several days thereafter. Each story consisted of an anchor lead-in, followed by a news report about the attacks or Bush administration response to the crisis and a direct presidential sound bite in reaction to the unfolding events. Altogether, there were eight unique stories in the stimulus material, two from each network. The average length of each story was just over 2 minutes (M = 125.75 seconds). After viewing each story, subjects completed a series of evaluative measures and open-ended questions designed to elicit any specific impressions prompted by the stories.
Subjects assigned to the Net news condition were instructed to visit two unaltered network news sites (among ABC, CBS, NBC with Tom Brokaw, or MSNBC with Brian Williams) and perform either an interactive or reading task for approximately 5 minutes. The specific pages visited featured news stories, photos, graphics, and interactive features about the terrorist attacks, making the online content comparable with the broadcast news stories. The decision to not replicate the exaci verbal content of the broadcast news stories and leave the Net news sites unaltered was made with the intention to maximize ecological validity and to deliberately test the effect of medium as opposed tc modality. Consistent with previous channel studies, "news items were selected so that the same substantive information was included in each medium's version of the story, but the journalistic presentation variei as it actually does by medium."^' After visiting each site, subject completed a series of evaluative measures designed to elicit credibility assessments of the individual sites.
Subjects assigned to the cross-platform (synergy) condition, ii randomly alternating order, first viewed four of the broadcast new stories and then visited two Net news sites. To guard against orde effects, some subjects first viewed the broadcast news stories and thei went online, while others went orJine first then viewed the storie Again, after viewing each story and visiting each site, subjects complete a series of evaluative measures and open-ended questions.
At the end of each media condition, subjects completed a series < overall credibility measures separately for television and Net news. Media credibility, the focal concept of this study, was operationalize as the degree to which news consumers judge network newscasts an Web sites to be believable, fair, accurate, informative, and in-depth. Each item comprising the credibility construct was measured using a'
252
TV News Credibility
Fair
Accurate
Believable
Informative
ln-Depth
Credibility Index
Note: For each item,
TABLE 1 Mean TV News Credibility Ratings by Age Group
Younger (Student)
5.08 (1.09)
5.30 (1.02)
5.54 (1.21)
5.82 (.98)
4.89 (1.45)
5.33 (.90)
1 = Not at All, 7 =
Older (Adult)
4.24 (1.34)
4.46 (1.22)
4.54 (1.38)
4.51 (1.54)
3.09 (1.33)
4.17 (1.12)
d.f.
164
164
164
164
164
164
t-value
4.44
4.80
4.97
6.54
8.38
7.36
Very. Standard deviations appear in parentheses.
P
.0001
.0001
.0001
.0001
.0001
.0001
point scale (1 = not at all; 7 = very). To determine the suitability of using the five items together in combined credibility indexes, a reliability analysis was performed for both TV news and Net news credibility. For both indexes the Cronbach's alpha coefficient was .89.
Subjects assigned to the control condition were not exposed to any media, but simply completed the experimental questionnaire and overall media credibility measures.
Abetween-subjectsanalysisofvariance{ANOVA) was performed Results for the media channel and age group factors on the combined credibility indexes. For the age group factor, there were significant main effects for both the TV news index, F(l, 160) = 54.28, p < .0001, eta-squared = .26, as well as the Net news index, F(l, 160) = 9.26, p < .01, eta-squared = .06. To reveal the contribution of individual items, a series of t-tests were run for each index separately.
Table 1 shows age-group differences for evaluations of TV news credibility, broker down by individual item. Across all items, the younger age group rated TV news credibility significantly higher than the older age group, reserving its highest ratings for informativeness and believability. Interestingly, both age groups rated TV news lowest on its depth. On the bottom row of the table the combined index scores show that younger audience members evaluated TV news to be significantly more credible overall than older audience members.
253
TABLE 2 Mean Net Nm/s Credibitity Ratings by Age Group
Net News Credibility
Fair
Accurate
Believable
Informative
In-Depth
Credibility Index
Note: For each item.
Younger (Student)
4.73 (1.20)
4.60 (1.22)
4.61 (1.27)
5.73 (.97)
5.54 (1.2S)
5.04 (.91)
1 = Not at All, 7
Older (Adult)
4.31 {1.47)
4.26 (1.44)
4.43 (1.45)
4.94 (1.50)
4.57 (1.76)
4.50 (1.34)
A.J.
159
159
159
159
159
159
t-value
1.97
1.60
.83
4.00
3.99
2.99
V
.05
n.s.
n.s.
.0003
.0001
.01
= Very. Standard deviations appear in parentheses.
Table 2 shows a similar, though less marked, pattern for evalua- tions of Net news credibiiity. Students were much more inclined to rate online news as informative and in-depth than adult subjects. As with TV news, adults aJso considered Net news significantly less fair than students. Again on the bottom row of the table the combined index scores show that younger audience members evaluated Net news to be significantly more credible overall than older audience members.
Figure 1 shows the main effects for age group on the combined index scores in bar chart form. Although students' credibility ratings are higher overall than adults, within age groups adults rate Net news to be significantly more credible (M = 4.50, s.d. = 1.34) than TV news (M = 4.16, s.d. = 1.13), i(76) = -2.64, p < .01. Within the yoiinger age group, the reverse occurs; students rate TV news more credible overall (M = 5.33, s.d. = .90) than Net news (M = 5.(M, s.d. = .91), t(83) = 2.61, p < .01. Another difference between the TV and Net news credibility ratings is revealed by the fewer degrees of freedom in the t-tests for the older age group when compared to the younger group (76 versus 83). Several of the older adult participants said they had no online experience and declined to even rate Net news credibility for this reason.
In response to the first research question, then, different audiences do seem to rate the credibility of on-air and onlirie news differently. Younger audience members are much more inclined to assign both news channels higher credibility ratings than older audience members. But within groups, adults consider Net news to be more fair, informative.
.25*
FIGURE I Main Effects for Age Group
Mean Rating
4 . 0 -
3.5
TVNews Credibility
Net News Credibility
Students Adults
and in-depth, and therefore more credible overall, than TV news. Students, on the other hand, give TV news higher marks than Net news on every item except in-depth and regard television as the more credible news source. Care should be taken not to read too much into these results, since these age-group differences do not take into account the unique effect of channel exposure, which is analyzed next.
The second research question asked what influence media channel had on audience evaluations of credibility. Analysis of variance produced a significant main effect for media channel on the TV news credibility index, f(3, 160) = 3.94, p < .01, eta-squared = .07. Across all subjects, exposure to television news resulted in the highest TV news index scores (M = 5.16, s.d. = 1.14) of any media channel. These scores were significantly higher than the control group (M = 4.54, s.d. = 1.12), which was exposed to no media, and the Net news-only group (M = 4.47, s.d. = 1.31), as determined by Tukey post-hoc comparisons (see Table 3). Media exposure did not have a significant effect on ratings of Net news credibility, which varied comparatively little regardless of channel used.
Figure 2 shows the main effects for media channel on the credibility indexes in bar chart form. The columns representing exposure to TV news suggest an effect of charmel congruence on credibility evaluations, where perceptions of credibility are enhanced when the channel used is consistent with the news source being evaluated. Simply put, television news viewing results in the highest evaluations of TV news credibility of any condition. Going online has the same effect on evcduations of Net news credibility for students (M=5.26, s.d. = .69), at a level significantly higher than the control group ratings (M = 4.65, s.d. = .79), t(39) = 2.66, p < .01, but not for adults (see Table 3). In three out of four cases, then, channel congruence enhances credibility evaluations. In response to the second research question, direct media exposure mostly enhances
255
Index
TV News Credibility
Net News Credibility
TABLE 3 Mean Credibiiity Ratings, Main Effect for Media Channel
Students
Adults
O\eraU
Students
Adults
Overall
TV
5.54 (1.08)
4.69 (1.07)
5.16, (1.14)
5.15 (1.02)
4.27 (1.28)
4.76 (1.21)
Web
5.30 (.93)
3.59 (1.05)
(1.31)'"
5.26 (.69)
4.18 (1.24)
4.74 (1.13)
Both
5.38 (.83)
4.40 (.95)
(1.01)'""
5.06 (1.04)
4.86 (1.26)
4.96 (1.14)
Control d.f.
5.06 (.70)
4.01 (1.23)
4.54^ 3 (1.12)
4.65 (.79)
4.65 (1.54)
4.65 3 (1.20)
f p Eta Squared
3.94 .01 .07
.77 n.s. .01
Note: For each item, 1 = Not at All, 7 = Very. Standard deviations appear in parentheses. Overall mearw for TV News Credibility that do not share subscripts differ at the p<.05 level in Tukey post- hoc comparisons.
credibility evaluations when the charmel is consistent with the news source being evaluated, resulting in what might be called a congruence effect. When the channel is inconsistent, direct exposure still results in higher evaluations than no exposure for students but lower evaluations for adults.
The third research question asked whether there is a synergy effect between on air and online news sources that enhances perceptions of credibility over use of either medium in isolation. Figure 2 illustrates the potential existence of this phenomenon. Subjects exposed to both broadcast and online news in the same experimental session (telewebbing) rated TV and Net news credibility higher overall than any other condition except for the TV congruency case (TV exposure producing the highest TV credibility evaluations). For both students and adults, teiewebbing has a small positive impact on TV news credibility (M = 4.90, s.d. = 1.01) beyond Net news use (M = 4.47, s.d. = 1.31) and no exposure (M = 4.54, s.d. = 1.12), but these individual comparisons were not significant in Tukey post-hoc comparisons (see Table 3).
The fourth research question asked whether there is an inter- action between media channel and age group resulting in different synergy effects for younger and older news audiences. Figure 3 shows a near significant interaction for media channel and age group on evaluations of Net news credibility, F(3,160) = 2.15, p < .10, eta-squared = .04.* Interestingly, the control groups for both students and adults
256 & MASS COMMUNJOOION Q
FIGURE 2 Main Effects for Media Channel
4.75-
Mean Rating
• TVNews CredibiHty
D Net News Credibility
Control TV Web Both
Media Channel
(shown on the left side of the graph) rated Net news credibility the san\e (M = 4.65), although the standard deviation was higher for adults (s.d. = 1.54) than students (s.d. = .79). When exposed to either TV or Net news, however, students and adults reacted quite differently (as indicated by the divergent upward and downward lines in the center of the graph). For adults, exposure had a negative effect and evaluations of Net news credibility fell from the control group level. For students, exposure had a positive effect and evaluations rose. When adults were placed in the telewebhing condition (shown on the right side of the graph), with exposure to both TV and Net news, evaluations of Net news jumped to their highest levels. For students, telewebbing caused perceptions of Net news to dip slightly. The interaction for media channel and age group for evaluations of TV news credibility was not significant.
To summarize these findings, when adults were placed in the telewebbing condition, with exposure to both TV and Net news, evaluations of Net news credibility jumped to their highest levels. Evaluations of TV news credibility aiso benefited from telewebbing, compared to the control group and to Web use alone. For students, telewebbing caused perceptions of TV and Net news credibility to increase in relation to the control group, but dip slightly or show no difference compared to other forms of media exposure. In answer to RQ4, then, telewebbing potentially results in greater synergy effects of Net news credibility for adults than students, but these results only approached significance.
This study of synergy effects, which adds to the experimental literature of media credibility, found significant main effects for media channel on evaluations of TV news credibility and for age group on
Discussion
257
Interaction for Media Ciiannel and Age Group
Net News Credibiiity
Mean Rating
Students
Adults
Conlro! TV Web Both
Media Channel
evaluations of both TV and Net news credibility. The ANOVA model also produced a near-significant interaction for the two factors on evaluations of Net news credibility, which may have achieved significance had the group sizes been larger. Consistent with survey studies of media credibility, younger audience members evaluated both TV and Net news to be significantly more credible overall than older audience members. In particular, students regarded both forms of network news as more informative than adults, while adults considered both forms of news to be significantly less fair than students. Both age groups rated TV news lowest for its depth. Within age groups, adults rated Net news to be significantly more credible than TV news, while students rated TV news more credible. These results are somewhat surprising, given that the audience for broadcast news tends to be older, while the orviine audience tends to be younger. Each age group evaluated the less familiar medium more generously.
Media charmel also influenced credibility evaluations. The ratings for television news suggested a channel congruence effect, where perceptions of credibility are enhanced when the channel used is consistent with the source being evaluated. This was the case for both students and adults. The finding of a congruence effect in an experimental setting differs from the common survey finding that media audiences regard the medium on which they are most reliant as the most credible because the evaluations here were generated in response to direct exposure to a particular medium and not based on reflections of past media use. Nevertheless, similar to the association of media reliance with high credibility evaluations revealed by surveys, audiences seem to regard media they have/«sf used more favorably than media to which they are not exposed.
258 /oURNALiSM & MASS COMMUNKATION QUAXTEKLY
Going online had the same congruence effect on evaluations of Net news credibility for students, but not adults. When the channel used is inconsistent, adults tend to evaluate the source less favorably compared to the control group. Adults were noticeably less inclined to view TV news as credible after exposure to tKe Web. For studenfa, any media exposure had a positive impact and evaluations rose relative to the control group.
As for cross-platform media use, the results provided some initial evidence for the existence of a synergy effect. The magnitude of these effects, though relatively small, was greater for adults than for stu- dents, perhaps reflecting students' propensity to multitask. A recent study of the gratifications obtained from new media found that students spend on average a total of 10.5 hours per day with a range of different media technologies.*' The telewebbing condition was arguably more novel for adults, and thus had a greater impact on their credibility evaluations. Overall, subjects exposed to the telewebbing condition rated TV and Net news credibility higher than the control group or subjects exposed orJy to Net news. The only exception was that TV exposure by itself elicited the highest evaluations of TV credibility.
The fact that several of the mean credibility ratings for telewebbing pointed to synergy effects but were not significantly different at the p < .05 level suggests that a larger number of subjects in this between- subjects experiment would have produced more significant results, Between-subjects experiments with 20 or fewer subjects per group, as was the case with this study, are notoriously underpowered.*^ To overcome this classic problem without having to involve hundreds of subjects in a costly and time-consuming data coUecfion process, future experimental research should take advantage of within-subjects, repeated measures designs, which provide greater statistical power with fewer subjects since each subject responds multiple times and serves as his or her own control.*'
When studying network news, the demographics of the audience are an essential consideration. Young adults who are wired, and therefore on a trajectory for the higher socioeconomic status that characterizes networknews viewers, represent the future of the on-air newsaudience. Whether members of this audience will ever completely migrate online for their news and information seems doubtful, at least at this stage of the Web's development. Students in this study reported almost two nights of network news viewing per week—even as members of a residential campus environment that provides "cost free," high- bandwidth orJine access. What seems almost certain, however, is a continued tendency by younger audiences to teleweb and, beyond that, multitask with a variety of information and communication technologies simultaneously.
The results of this study suggest that media exposure has differential effects on student and adult perceptions of news credibility- Future research should keep this finding in mind and not attempt to generalize about the news audience from a convenience sample of undergraduates. The experimental nature of this study, while desirable for the control it offered over the news channels subjects ware exposed
259
to, was limited in its ability to simulate actual viewing conditions. Subjects in the telewebbing condition, for instance, either saw a series of newscasts first and then went online, or spent some time online first and then watched a series of broadccists. At home (or in the office), news consumers who teleweb may use both media at the same time. Follow-up studies should attempt to recreate a real-time telewebbing or "enhanced television" experience in as naturalistic a way as possible. In addition, other genres of media content associated with telewebbing, such as live sports coverage or news specials, should be studied. Satisfying these two conditions may lead to more pronounced findings for synergy effects.
With the growth of telewebbing and multitasking as routine audience activities, it wiU become increasingly important to understand and measure perceptions of news credibility under conditions of simultaneous media use. Considering the potential contribution of synergy effects resulting from cross-platform media use could help recast the debate about media credibility so that the realities, rather than thehyperbole, of online developments are adequately taken into account. The networks have adapted to the online environment and are making gainful use of both delivery platforms. Moreover, most audience members do not stop using a source as heavily relied on as television just because a new medium arrives on the scene; instead, their media repertoire may become wider and more varied.
The rise of telewebbing and two-screen households presents media research with the challenge of understanding the impact of this new consumer behavior on evaluations of news content. With technological convergence, the independence of different delivery platforms is eroding. Even so, the stylistic differences between on-air and online news m.ay persist, withbroadcast news retaining its trademark chronological narrative and online (text-based) news embracing the inverted pyramid style of print reporting. Although both forms of news are similar in that they are delivered through electronic telecommunications terminals, they may be processed and evaluated with vastly different criteria, as is the case for assessmentsof newspaper and television news credibility.** From the perspective of the news audience, the ways in which television and the Web complement—and contravene—each other are only beginning to be vmderstood. Research should investigate the different processing styles required for on-air and online news, with the aim of revealing the psychological mechanisms underlying credibility judgments.
NOTES
1. This study was supported with a Research in Broadcasting Grant from the National Association of Broadcasters. The author would Uke to thank Christopher E. Beaudoin for his thoughtful reading of this manuscript and members of the Institute for Communication Research lab in the Department of Telecommunications at Indiana University, especially Seungwhan Lee, Yongkuk Chung, Byung-ho Park, and Leah
Haverhals, for their assistance with running subjects and inputting data. An earlier version of this paper was presented to the Broadcast Education Association NAB-sponsored Research Grant Panel, Las Vegas, NV, April 2002.
2. Pew Research Center, "TV News Viewership Declines" (13 May 1996), available at http://people-press.org/reports, 16 May 2002.
3. Deirdre McFariand, "First Scarborough National Internet Study Reveals Changes in How Online Consumers Use Traditional and Internet Media" (New York: Scarborough Research, 9 May 2001), available at h t t p : / / w w w .scarborough.com/scarb2002/press/pr_intemetstudyl. htm, 8 May 2002.
4. Pew Research Center, "Internet Sapping Broadcast News Audience" (11 June2000),availableathttp:/ /people-press.org/reports, 23 July 2000.
5. McFariand, "First Scarborough National Internet Study." 6. Jim Davis, "More People Surfing Web While Watching TV"
(CNET News.com, 15 February 2000), available at http:// news.com.com/2iaO-1040-236906.html?legacy=cnet&dtn.head, 8 May 2002; Bill Niemeyer and Steven Hoffman, "Interactive Television 4 Here Today," NATPE Media Trends 4 (winter 2002), available at http:/ /www.centrimedia.com/archives/2002_01_NATPE_Article.html, 8 May 2002,
7. Niemeyer and Hoffman, "Interactive Television Is Here Today." 8. Pew Research Center, "Internet Sapping Broadcast News
Audience." 9. Sylvia M. Chan-Olmsted and Jung Suk Park, "From On-Air to
Online World: Examining the Content and Structures of Broadcast TV Stations' Web Sites," Journalism & Mass Communication Quarterly 77 (summer 2000): 321-39.
10. Pew Research Center, "Internet Sapping Broadcast News Audience."
11. Bradley S. Greenberg, "Media Use and Believability: Some Multiple Correlates," journalism Quarterly 43 (winter 1966): 665-70,732; Wayne Wanta and Yu-Wei Hu, "The Effects of Credibility, Reliance, and Exposure on Media Agenda-Setting: A Path Analysis Model/' Journalism Quarterly 71 (spring 1994): 90-98; Bruce H. Westley and Werner J. Severin, "Some Correlates of Media Credibility," Journalism Quarterly 41 (summer 1964): 325-35.
12. Andrew J, Flanagin and Miriam J. Metzger, "Perceptions of Internet Information Credibility," Journalism & Mass Communication Quarterly 77 (autumn 2000): 515-40.
13. Steve McClellan, John M. Higgins, Dan Trigoboff, and Bill McConnell, "It's Gore! It's Bush! It's a Mess! Election Night Gaffes Mean Television Will Be Under Scrutiny for a Long Time," Broadcasting 6* Cable, 13 November 2000,6-10.
14. Andrew Kohut, "The Press Shines at a Dark Moment," Columbia Journalism Review, January/February 2002,54-55.
15. Spiro Kiousis, "Public Trust or Mistrust? Perceptions of Media Credibility in the Information Age," Mass Communication & Society 4 (fall 2001): 381-403.
M E D U CsiDinmrr RECONSIDHIED 261
16. Cecilie Gaziano and Kristin McGrath, "Measiirmg the Concept of Credibility," Journalism Quarterly 63 (autumn 1986): 451-62; for an overview, see CharlesC. Self, "Credibility," i n ^ n Integrated Approach to Communication Theory and Research, ed. Michael B. Salwen and Don W. Stacks (Mahwah, NJ: Lawrence Eribaum Associates, 1996), 421-41.
17. The Roper question asked: "If you got conOicting or different reports about fiie same news story from radio, television, the magazines, or the newspapers, which of the four versions would you be most inclined to believe—the one on radio or television or magazines or newspapers?" The special case of confhcting reports does not address the credibility of each medium apart from this scenario (Walter Gantz, "The Influence of Researcher Methods on Television and Newspaper News Credibility Evaluations," jourrml of Broadcasting 25 [spring 1981]; 155-69).
18. Gantz, "The Irifluence of Researcher Methods on Television and Newspaper News Credibility Evaluations"; Gaziano and McGrath, "Measuring the Concept of Credibility"; Tony Rimmer and David Weaver, "Different Questions, Different Answers? Media Use and Media Credibility," journalism Quarterly 64 (spring 19S7): 28-36, 44; Eugene F.Shaw, "Media Credibility: Taking theMeasure of a Measure," Journalism Quarterly 50 (summer 1973): 306-11.
19. Flanagin and Metzger, "Perceptions of Internet Information Credibility."
20. Thomas J. Johnson and Barbara K. Kaye, "Cruising Is Believing? Comparing Internet and Traditional Sources on Media Credibility Measures," Journalism &• Mass Communication Quarterly 75 (summer 1998): 325-40.
21. Pew Research Center, "TVNews Viewership Declines," "Internet Sapping Broadcast News Audience"; Howard I. Finberg, Martha L. Stone, and Dianne Lynch, Digital Journalism Credibility Study (New York: Online News Association, 31 January 2002), available at http:/ / wviav.joumalists.org/Programs/Research.htm, 17 May 2002.
22. Tim O'SuUivan, Brian Dutton, and Philip Rayner, Studying the Media: An Introduction, 26 ed. (London: Arnold, 1998).
23. Davis, "More People Surfing Web While Watching TV." 24. JohnE. Newhagen, "Effects of Televised Government Censorship
Disclaimers on Memory and Thought Elaboration During the Gulf War," Journal of Broadcasting & Electronic Media 38 (summer 1994): 339- 51; Maria Elizabeth Crabe, Shuhaa H. Zhou, Armie Lang, and Paul David Bolls, "Packaging Television News: The Effects of Tabloid on Information Processing and Evaluative Responses," Journal of Broadcasting & Electronic Media 44 (fall 2000): 581-98.
25. See, for example, Cecilie Gaziano, "How Credible Is theCredibility Crisis?" Journalism Quarterly 65 (summer 1988): 267-78, 375; Johnson and Kaye, "Cruising Is Believing?"; Self, "Credibility."
26. Richard F. Carter and Bradley S. Greenberg, "Newspapers or Television: Which Do You Believe?" Journalism Qturterly 42 (winter 1965): 29-34; John D. Abel and Michael O. Wirth, "Newspapers vs. TV Credibility for Local News," Journalism Quarterly 54 (summer 1977): 371-75; Raymond S. H. Lee, "Credibility of Newspaper and TV News,"
262
Journalism Quarterly 55 (summer 1978): 282-87; Gantz, "The Influence of Researcher Methods on Television juid Newspaper News Credibility Evaluations"; John Newhagen and Clifford Nass, "Differential Criteria for Evaluating Credibility of Newspapers and TV News," Journalism Quarterly 66 (summer 1989); 277-84.
27. Carter and Greenberg, "Newspapers or Television"; Rimmer and Weaver, "Different Questioris, Different Answers?"; Wanta and Hu, "The Effects of Credibility, Reliance, and Exposure on Media Agenda-Setting."
28. Gaziano, "How Credible Is the Credibility Crisis?"; Johnson and Kaye, "Cruising Is Believing?"; Wanta and Hu, "The Effects of Credibility, Reliance, and Exposure on Media Agenda-Setting."
29. Randy E. Miller and Wayne Wanta, "Sources of the Public Agenda: The President-Press-Public Relationship," Jnfernah'onai/ournflJ of Public Opinion Research 8 (winter 1996): 390-402.
30. Leonard Downie Jr. and Robert G. Kaiser, The News About the Nezvs: American Journalism inPeril{NewYoik:Knopi,2002);]eiiGialmck, "How Network News Outsmarted Itself," Columbia Journalism Review, May/June 2002, 54-55.
31. Ronald Mulder, "A Log-Linear Analysis of Media Credibility," Journalism Quarterly 58 (winter 1981): 635-38; Michael J. Robinson and Andrew Kohut, "Believability and the Press," Public Opinion Quarterly 52 (summer 1988): 174-89. Although it is not the focus of this study, gender is also strongly associated with news media credibility. As a rule, men evaluate news media as less credible than women. In terms of network credibility, Robinson and Kohut found a 5-percentage-point difference between genders, with women consistently more willing to report they believe the news media.
32. Robinson and Kohut, "Believability and the Press," 33. In their study, Robinson and Kohut operationalized knowledge
of the press with 11 close-ended items that included an xmderstanding of such basic concepts as a press release, newsroom roles, and ownership of major news organizations (Robinson and Kohut, "Believability and the Press").
34. Douglas A. Ferguson and Elizabeth M. Perse, "The World Wide Web as a Functional Alternative to Television, "journal of Broadcasting & Electronic Media 44 (spring 2000): 155-74; Barbara K. Kaye and Thomas ]. Johnson, "From Here to Obscurity: Media Substitution Theory and the Internet" (paper presented at the annual meeting of AEJMC, Phoenix, AZ, August 2000).
35. John P. Robinson, Kevin Barth, and Andrew Kohut, "Social Impact Research: Personal Computers, Mass Media, and Use of Time," Social Science Computer Review 15 (spring 1997): 65-82; John P. Robinson and Meyer Kestnbaum, "The Personal Computer, Culture, and Other Uses of Free Time," Social Science Computer Review 17 (summer 1999): 209-16.
36. John C. Schweitzer, "Personal Computers and Media Use," Journalism Quarterly 68 (winter 1991): 689-97; Joseph M. Kayany and Paul Yelsma, "Displacement Effects of Online Media in the Sociotechnical Context of Households," Journal of Broadcasting & Electronic Media 44
263
(spring 2000): 215-29; Norman H. Nie and Lutz Erbring, Internet and Society: A Preliminary Report (Stanford, CA: Stanford Institute for the Quantitative Study of Society, 2] April 2000), available at: h t t p : / / www.stanfoTd.edu/group/siqss/Press_Release/intemetS tudy.html, 17 May 2002.
37. W. Russell Neuman, Marion R. Just, and Ann N. Crigler, Common Knowledge: News and the Construction of Political Meaning (Chicago: University of Chicago Press, 1992), 34.
38. The exact wording of the credibility question was as foUows: "Please circle the number on each of the items below that best represents your overall evaluation of the following media as news sources." Subjects then completed 7-point scales rating television and the World Wide Web for fairness, accuracy, believability, informativeness, and depth. Although thecredibility question didnot mention networknews specifically, all the stimulus materials consisted of network news stories or Web sites. Subjects were therefore primed to evaluate the credibility of network news, not news media in general.
39. Flanagin and Metzger, "Perceptions of Internet Information Credibility"; Johnson and Kaye, "Cruising Is Believing?"
40. With the significance level set at p < .10, observed power = .68, indicating that a larger number of subjects may have produced more apparent group differences (see James P. Stevens, Applied Multivariate Statistics for the Social Sciences, 4th ed. [Mahwah, NJ: Lawrence Erlbaum Associates, 2002J).
41. Xiaomei Cai, "A Test of the Functional Equivalence Principle in the New Media Environment" (Ph.D. diss., Indiana University, Bloomington, 2001).
42. Stevens, Applied Multivariate Statistics for the Social Sciences. 43. Stevens, Applied Multivariate Statistics for the Social Sciences. 44. Newhagen and Nass, "Differential Criteria for Evaluating
Credibility of Newspapers and TV News."
264 IOimNAUsM& MASS COMMUNICATION
