Boomers versus Millennials: Online Media Influence on Media Performance and Candidate Evaluations

Facebook posts, YouTube videos, tweets and wooing political bloggers have become standard practice in marketing political campaigns. Research has demonstrated the effect of new media on a host of politically-related behavior, including political participation, knowledge acquisition, group formation and self-efficacy. Yet, issues related to media trust, media performance and candidate evaluations have not been fully explored. In addition, much of the political marketing research looks exclusively at the Millennial age cohort, ignoring other age groups, particularly Baby Boomers. This case study addresses whether attention to traditional (i. and political blogs) about the 2012 U.S. presidential campaign influences Millennials and Baby Boomers' media trust and performance ratings, as well as candidate evaluations. Panel surveys were completed by both age cohorts, Millennials (n = 431) and Baby Boomers (n = 360), during the last two weeks of the presidential election. Findings indicate that traditional sources, specifically television, rather than online sources are significantly linked to media trust and performance ratings among both Boomers and Millennials. Attention to traditional media for campaign information predicts Boomers' candidate evaluations, whereas Millennials' candidate evaluations are influenced by online sources, such as Facebook and candidate websites.


Introduction
Young adults' lives are both immersed and dependent on new media, which serves as a gateway to news consumption, socializing, professional networking, continuing education and political participation.As Millennials embraced social media, political campaigns integrated these new tools into their campaign strategies.As early as 2006, campaigns employed social media to communicate with and mobilize potential voters [1].By 2012, political campaigns were saturated with online technology.These online marketing efforts were not without merit, as a wealth of previous research shows that attention to online media can influence political attitudes and behaviors (e.g., [2][3][4][5][6][7][8][9][10]).Much of the latter research, however, examines new media usage among young adults or Millennials, focusing primarily on how media influences young adults' political participation and knowledge [2,11,12].We build on the prior research by examining the influence of attention to online media on political attitudes among two different age cohorts, namely Baby Boomers versus Millennials.
Millennials are indeed considered the generation of "digital natives," but recent survey data suggest that the digital generation gap may be closing [13].For example, a 2012 Pew Research Center study reported that 77% of Baby Boomers, ages 50-64, were going online [14].In fact, the number of Boomers using social media has significantly grown since 2005, with 51% of 50-64 year olds using social networking sites (SNS) in 2015 [15].As older age groups adopted these newer technologies, political and communication scholars began to examine the impact of news media attention and social media use for political reasons across different age groups (e.g., [16][17][18][19][20][21]).However, these studies often examine "younger versus older citizens", "first time voters versus experienced voters" and pre-defined age groups (e.g., 18-21 years, 22-35 years, 36-55 years, 56 and older) These samples of "older adults" cannot be treated as Baby Boomers, born between 1946 and 1964.Therefore, the political implications of Baby Boomers' increasing online media use remains relatively unstudied.To fill this gap, we conduct a case study, examining how reliance on certain online media types influences political attitudes and behaviors among Millennials and Baby Boomers during the 2012 U.S. presidential campaign.
This study draws on the media dependency theory to explore how Baby Boomers and young adults' use different media types to serve their political information needs.The media dependency theory explains that the more someone depends on a media source, the more often they use that media source [22].Then, in certain situations or constrained environments, the media source that an individual routinely uses becomes their primary information source, playing an influential role on their public opinion [23,24].Therefore, during campaign periods, we expect Boomers and Millennials to depend on different media types for campaign information, ultimately influencing their political attitudes in very different ways.By using survey data, we explore the similarities and differences between Baby Boomer and Millennials' level of attention to specific offline and online media sources (i.e., television, hard-copy newspapers, radio, political candidate websites, television network websites, online newspapers, Facebook, Twitter, YouTube, Tumblr and political blogs) about the 2012 U.S. presidential campaign.Most importantly, the media dependency theory would suggest that the media sources Boomers and Millennials pay attention to the most for campaign information will significantly predict their political attitudes, as well as their attitudes about the media.
We expand on previous media research by examining understudied political attitudes and perceptions among these age cohorts, particularly their levels of media trust and performance, as well as their candidate evaluations (for an exception, see [25]).The latter variables are important to examine because citizens rely heavily on the media to inform them about the political agenda, as well as the candidates.It remains unclear, however, what specific sources-blogs, social networks or YouTube-garner the highest trust and performance ratings among citizens, as previous findings are mixed [6][7][8]12].Moreover, previous reports suggest that age plays an important role in the way that the young and old evaluate a media outlet's digital and online content [15].Along with media trust and performance, we also investigate how media coverage about the campaign influences Baby Boomer and Millennials' evaluations of Romney and Obama.While it is well known that mass media coverage can influence a political candidate's electoral success [26,27], the impact of attention to online sources on candidate evaluations, particularly among different age cohorts, is inconclusive [5,7,8].
Last and most important, media consumption differs between these two groups.Previous research has demonstrated that older adults are heavier consumers of traditional media, such as television and hard-copy newspapers, compared to young adults [36,37].Indeed, Boomers are the only generation to have grown up with only three monolithic television sources and a local hard-copy newspaper.In contrast, Millennials were born in a media environment including more than three television channels and the dawn of the Internet.Therefore, Millennials adopted online technology more quickly than older citizens and use more social media than their elders [38][39][40].Specifically, Millennials spend 35 h per week on digital media, compared to 23 h for Boomers [41].In contrast, Boomers spend 37 h per week on traditional media versus 32 h for Millennials [41].In 2013, eMarketer [42] reported large, overall online adoption rates across all generations, but marked differences in platform usage: digital video viewers (Boomers, 47%; Millennials, 82%), Facebook (Boomers, 40%; Millennials, 75%) and Twitter (Boomers, 7%; Millennials, 19%).However, it not well understood how some age groups, specifically Baby Boomers, are using online media sources for political information during campaign periods.
As we have noted, very few scholars examine how dependency on certain media outlets for political content influences political attitudes and behaviors among Millennials and Baby Boomers.Previous scholarship largely examines attention to online sources by comparing the "young" versus the "old".For instance, during the 2008 U.S. presidential election, Haridakis and Hanson [17] find that "first time voters" were more likely than "older voters" to use the Internet, SNS and YouTube for political information.First time voters were also more motivated than older voters to seek out information about the political campaign on media sources.Studies conducted during the 2012 campaign confirm that older age groups (e.g., 50-64+ year olds) increased their adoption and use of social media for politics since the 2008 election (e.g., [43,44]).As older adults included online media into their political media diet, some scholars questioned how attention to online sources influenced seniors' political behavior.In one notable study, Bachmann et al. [16] find evidence that citizens both young (18-29 years) and old (30+ years) who prefer consuming news online have higher offline and online political participation.Yet, the link between preference for online news and actually casting a vote was much higher for young people than older people.Not surprisingly, offline news preference was more strongly linked to voting among older people rather than younger.

Generational Political News Consumption and Media Reliance
Recent presidential elections have been widely covered by newspaper reporters, television pundits and professional and amateur bloggers and posters alike (e.g., [45][46][47][48]).The 2012 presidential campaign was no different, as political messages were delivered through a mix of both traditional and online media with a record amount of money (approximately two billion dollars) spent across both the Obama and Romney campaigns [49].While the bulk of the money was spent on traditional marketing communication (i.e., television advertising), each party acknowledged the growing importance of digital, especially social media, with increased spending over 2008 numbers [49,50].Both presidential campaigns had an active presence online utilizing campaign websites, Facebook, Twitter, YouTube, blogs, Tumblr, Pinterest, Instagram and Spotify [51].However, it was the Obama campaign, spending ten-times more ($47 million) on digital media compared to Romney ($4.7 million), who fully embraced the social digital landscape [52].
Given the increasing amount of choices available to gather political news, Baby Boomers and Millennials have a variety of media sources on which they can rely.To explore their political news reliance, we build upon media dependency theory [22].Media dependency theory seeks to understand the effects of mass media and the relationship between media and its audience.Specifically, it focuses on the dependency that an individual has towards media to fulfill certain needs and goals, which is similar to the uses-and-gratifications theory (see [53][54][55]).According to the media dependency theory, citizens do not depend on all media equally.The level of dependency and the influence it wields are contingent on the social functions the media provides; the amount of influence increases with levels of dependence [22,56].Media dependency, also conceptualized as media reliance, has been used in a number of political communication studies that have explored channel effects [25,57].Thus, this study seeks to extend previous political communication research by determining which offline and online media sources Baby Boomers and Millennials depend on for campaign information.
It is known that Boomers and Millennials are relying on different news sources to varying degrees.In 2012, Boomers spent 77 minutes on average per day with news, compared to 46 minutes for Millennials [30].Typically, news consumption increases with age; however, Millennials' consumption has not significantly grown over an eight-year period ( [30,58]; see also [29]).For both cohorts, traditional media is losing ground to online news consumption.Although, Boomers continue to rely on traditional media outlets more so than Millennials.In the fall of 2012, among all citizens surveyed about where they received their news yesterday, 55% watched television, 50% received digital news, 39% got online/mobile news, 33% listened to the radio and 29% read a newspaper.Television, radio and newspapers have seen significant declines, whereas digital and online news have dramatically grown [59].Both cohorts were receiving news via digital news platforms in high rates: 49% of 40-49 year olds and 35% of 50-64 year olds compared to 41% of 18-24 year olds and 45% of 25-29 year olds received digital/online news [59].Despite some common ground, generational gaps can be seen in the number of Millennials who regularly see news/news headlines on SNS (i.e., 31% 18-24) compared to 10% of 50-64 year olds in 2012 [59].
Thus far, we have begun to illustrate cohort differences in general news consumption; however, do these similarities and differences remain when we examine attention to campaign coverage?Which offline and online media sources do Boomers and Millennials rely on the most during a campaign period?A national sample of adults revealed that television was the top source of campaign news (67%) during the 2012 presidential campaign, followed closely by the Internet (47%) [60].Internet news consumption grew by 12% between the 2008 and 2012 elections.In another survey of adults, 71% of online users relied most heavily on blogs, compared to 12% for social networks and less than 10% for YouTube and Twitter for political information [25].We do know that Boomers are more politically interested and invested.For example, a survey found that 28% of Boomers were "following election news very closely" compared to 17% of Millennials [31].Indeed, general news and topline political news consumption data are available, but we are unable to explore the differences in political news reliance between the age cohorts.
Considering the above data, it can be argued that Millennials, who outpace older age groups in all types of technology use [31], will depend on more online media sources for campaign information than Baby Boomers.Boomers, who still hold strongly to their traditional media habits, will likely depend more on offline, traditional media sources for campaign information.Of course, it is plausible that both Baby Boomers and Millennials will rely on a mixture of offline and online media sources during a campaign period.Boomers are increasingly using online sources [14,15] in addition to traditional media sources.Millennials, raised in a fragmented media environment offering a variety of offline and online sources, are unlikely to rely solely on online media outlets.At this point, it is unclear which offline and online media source Boomers and Millennials depend on the most during a campaign.Following the basic argument of the Media Dependency Theory, the media source that is used the most will play a more significant role on their political attitudes.This leads us to the following research question and hypothesis: Research Question #1: Between Millennials and Baby Boomers, what similarities and differences exist in the amount of attention given to traditional and online media sources (i.e., television, hard-copy newspapers, radio, political candidate websites, television network websites, online newspapers, Facebook, Twitter, YouTube, Tumblr and political blogs) about the 2012 presidential campaign?
Hypothesis #1: Millennials will be more likely to pay attention to online media sources for information about the 2012 presidential campaign than Boomers.Boomers will be more likely to pay attention to traditional, offline media sources for campaign information than Millennials.

Political News Consumption and Media Trust and Performance
Media trust and performance ratings are often thought to be indirect measures of how balanced, fair, objective and neutral media coverage is perceived to be (see [21]).Ample political communication research has addressed media neutrality and bias [61,62].Scholars have not only supported the notion of a media bias, but also asserted that mass media has a liberal bias [63]; whereas, other scholars believe that there is a conservative bent controlled by corporate interests [64].Yet, we know little about consumers' perceptions surrounding media trust and performance, particularly regarding political news, delivered on traditional outlets, such as television and radio, versus online outlets, such as SNS.This is important to examine, as traditional media is simultaneously being replaced and supplemented by online media (see [65]).
Overall, confidence in the mass media is declining [65].In fact, drops in media trust are particularly evident during national election years: 51% trusted traditional media a "great deal/fair amount" in 2000, dropping to 40% by 2012 [66].Consumers also perceive some media sources as more trustworthy than others.For example, among traditional sources, research has demonstrated that newspapers are perceived as more trustworthy than television news [65][66][67][68].Issues related to media trust, however, are more complex when considering online media.This can, in part, be attributed to the evolving nature and increasing number of Internet media source types and related consumer perceptions involving source credibility.According to Gallup trends, overall trust for news on the Internet is relatively low and consistent over a 15-year period (see [65,67]).Internet news may be evaluated more critically than traditional media [21], and the process to determine trustworthiness may be more difficult for online news [69].
Research on specific forms of new media has addressed media trust and performance in the form of blogs and numerous other online mediums [6,7,25,[70][71][72][73][74].A study examining online political information credibility during the 2004 presidential campaign found that only online versions of broadcast, cable television and candidate literature were perceived to be more credible than the traditional version [72].Comparisons of online sources have also found that blogs and political websites are deemed more credible than "traditional" online news websites (i.e., online newspapers and online broadcast news) by politically interested online users [70,71].Furthermore, in an experimental study, young adults exposed to campaign content on online broadcast news websites had higher media trust and performance ratings for Internet sources compared to respondents who were exposed to political content on YouTube ( [7]; see also [6]).Similarly, an additional study found that the more individuals relied on YouTube and blogs for political information, the lower their levels of media confidence; however, higher levels of confidence were found in individuals that relied more heavily on SNS and Twitter [25].Therefore, scholars are only beginning to understand the relationship between attention to various forms of media and citizens' media trust and performance levels.The latter is especially true for generational similarities and differences in attitudes toward the media.
Presently, overall confidence in social and political institutions is lower for both Baby Boomer and Millennial generations [74].Across media sources (not including SNS), Baby Boomers and Millennials have relatively equal levels of media trust (i.e., Millennials 41%; Baby Boomers 38%) and distrust (i.e., Millennials 21%; Baby Boomers 23%) regarding political news [75].While Boomers and Millennials may have similar levels of general media trust and performance, Young [21] reports that there are age differences regarding "relying on digital information" and the value placed on the "convenience of the online source".For example, young people use online sources more than older adults, placing more importance on online outlets' load time, ease of use on the mobile phone and amount of ad interference.Therefore, different age groups may vary in the way they trust and rate campaign content offered by online media.Indeed, previous academic research has supported the contention that information source matters when determining trust in a media source [76].
Given that the previous research is inconclusive on the differences between Boomer and Millennials' media trust and performance ratings, we investigate two research questions, RQ#2 and RQ#3, below, rather than offering hypotheses.We then offer two hypothesis, H#2 and H#3, to examine the empirical relationships between Boomer and Millennials' reliance on media sources and their levels of media trust and performance.The media dependency theory would likely suggest that the media sources that Boomers and Millennials depend on the most will be better predictors of their media trust and performance ratings.Again, this theory implies that the more an individual becomes dependent on a media source to fulfill their need for campaign information, the more influence and power that source will have on their attitudes.
Research Question #2: Between Baby Boomers and Millennials, what similarities and differences exist in their levels of trust in the media?
Hypothesis #2: Media source dependency for information about the campaign for president will predict Baby Boomers and Millennials' trust in the media.
Research Question #3: Between Baby Boomers and Millennials, what similarities and differences exist in their perceptions of the media's performance in covering politics?
Hypothesis #3: Media source dependency for information about the campaign for president will predict Baby Boomers and Millennials' perceived media performance.

Political News Consumption and Candidate Evaluations
Voters are exposed to a multitude of political messages sent over an increasing number of communication channels during the campaign season.Political messages vary in terms of the type of communication (e.g., text, video, audio), length and message content.Yet, these messages, contextualized by the type of communication channel in which they are delivered, work in concert to assist voters in formulating their candidate evaluation.Therefore, media (the type and amount of time spent on it) can influence the extent to which we "like" a political candidate.Candidate evaluations matter greatly in that how favorably citizens view a political candidate has direct implications on voting choice [77].
Previous research on traditional and online media has demonstrated that media influences a candidate's image or their likeability [78][79][80].In a comparison of television versus newspaper "reliant voters," McLeod, Glynn and McDonald [78] found that voting choice was influenced by candidate image characteristics more by individuals who heavily relied on television.More recent research, comparing television and Internet channel effects in the 2000 presidential campaign, found no effect for either channel or a difference between channels for candidate favorability evaluations [81].However, during the 2000 election, Kaid [82] did find significant differences in both Republican and Democratic candidate evaluations when political messages (ads, debates and news) were delivered over the Internet.In a follow-up study of the 2004 presidential election, Kaid and Postelnicu [79] found that exposure to television ads did not influence candidate likeability; however, Internet exposure, which included viewing ads on the candidate's campaign website and reading ad information on Factcheck.org,significantly changed candidate evaluations for both Democratic and Republican candidates.
The emergence and popularity of various online media sources has also brought into question their role in shaping candidate evaluations.Conducting research on the 2008 presidential election, Kaye [43] found that media reliance on SNS, online newspapers, online news magazines, partisan political blogs and candidate websites predicted candidate support for Obama, but not for McCain.The only media source that predicted support for McCain was cable television.Preference for a candidate also predicted their media reliance.Kaye [57] attributed these results, in part, to both Obama's 2008 online media campaign and younger voters' online media habits.An additional study that compared online news media and SNS (e.g., Facebook and Twitter) found that SNS did not influence candidate attitudes, such as likeability, but general online news was influential [83]; whereas, an experimental study that exposed young adults to 2008 presidential campaign information on YouTube and a broadcast news website found no effect on candidate evaluations [7].In contrast, more recent survey work reported that some online media sources can influence candidate evaluations, but their influence varied by online outlet, candidate and candidate trait [5][6][7][8][9]84].While some scholars, as discussed, are beginning to look at the effect of online media sources on candidate evaluations, there is still little consensus, and we know even less regarding age cohort effects.
One research question, RQ#4, is posed to compare Boomer and Millennials' evaluations of Romney and Obama.Given the age differences in voting preferences during recent elections [31], it is expected that Millennials will likely have higher evaluations of Obama, whereas Boomers will have higher evaluations of Romney (H#4a).Yet, it is unknown how attention to varying media sources for campaign content influence candidate evaluations among these age groups.H#4b is presented below to test the link between Boomer and Millennials' attention to media sources and their Romney and Obama evaluations.Building on the media dependency theory, the media sources that Boomers and Millennials depend on the most for campaign information will likely better predict their candidate evaluations.
Research Question #4: Between Baby Boomers and Millennials, what similarities and differences exist in their candidate evaluations?
Hypothesis #4a: Millennials will have higher evaluations of Barack Obama than Boomers.Boomers will have higher evaluations of Mitt Romney than Millennials.
Hypothesis #4b: Media source dependency for information about the campaign for president will predict Baby Boomers and Millennials' candidate evaluations.

Data
The sample is comprised of two segments: (1) a panel survey of college students born between 1983 and 1994 (n = 431) enrolled in a medium-sized, public university located in the Midwest; and (2) a non-student panel survey of Baby Boomers born between 1946 and 1964 (n = 360).Both panel surveys were distributed and completed online during the last two weeks of the 2012 presidential election campaign from 19 October 2012-5 November 2012.To survey young adults, an online survey was administered via university email to over 1000 students enrolled in introductory-level political science, sociology and psychology courses.This recruitment method elicited 476 responses with a response rate of 48%.In the college student panel, forty-three respondents who were not born between 1983 and 1994 were removed from the sample.In addition, two respondents who were not U.S. citizens were removed.To survey Baby Boomers, Survey Analytics (www.surveyanalytics.com)was commissioned to sample Baby Boomers from its BoomerOpinion panel.BoomerOpinion is a self-identified, opt-in panel of those individuals aged 46-65 years old who register on the BoomerOpinion website (http://Boomers.micropanel.com/)and provide an email address.Panel participants are invited via email to participate in online surveys at their discretion and are monetarily compensated.Individuals without Internet or email access are excluded from the panel.Eight hundred and eighty-four BoomerOpinion panelists viewed the online survey, and 502 panelists completed it.In the Boomer survey, fifty respondents who were not born between 1946 and 1964 were removed from the sample.An additional 92 respondents who were not U.S. citizens were also removed from the Baby Boomer sample.
The demographic and political characteristics of respondents are presented in Table 1.The Millennial respondents were relatively representative of the "typical" American young adult with a few exceptions: Caucasians and females were slightly overrepresented [85].Comparing our Boomer sample to the U.S. Census Bureau's [85] demographic description of "Baby Boomers", our participants are better educated and somewhat wealthier.Notes: a The question wording for education level differed for Boomers and Millennials.For Boomers, respondents were asked "What is your highest level of education?"(1 = less than high school; 2 = high school/GED; 3 = some college; 4 = Associate's degree; 5 = Bachelor's degree; 6 = Master's degree; 7 = Doctoral degree (PhD); and 8 = professional degree (MD, JD); for Millennials, respondents were asked: "What is your year in school?"(1 = first year; 2 = sophomore; 3 = junior; 4 = senior).
At this point, it is important to note some weaknesses regarding the Boomer and Millennial survey panels.First, these respondents represent two separate survey administrations, which are not randomly selected from a national population of Boomers and Millennials.Unfortunately, there is no one national survey that contains a robust oversample of both of these specific age groups, as well as the necessary variables about politics and media attention.We readily acknowledge that any comparisons between these two age groups are likely limited, and the empirical results discussed below cannot be generalized to the wider population of Boomers and Millennials.Second, the cross-sectional survey design does not allow an assessment of a cause and effect relationship between media attention and political attitudes.Given these limitations, the results reported below should be viewed as a case study, examining the relationship between media attention and political attitudes among these different age cohorts.

Dependent Variables
Media trust and performance: Adapted from the American National Election Studies (ANES) survey, we utilize two indicators to measure perceptions of political media coverage, particularly media trust and performance ratings.First, respondents agreed or disagreed with the following statement: "I trust the media to cover political events fairly and accurately."(1 = strongly disagree, 3 = neither agree nor disagree, 5 = strongly agree).Second, we asked respondents to rate the media's performance in covering politics: "Overall, how would you rate the media's performance in covering politics in America?" (1 = poor, 2 = only fair, 3 = good, 4 = excellent).This study examines both media trust and performance as dependent variables in order to gain a broader understanding of how citizens feel about political campaign coverage.
Candidate evaluations: This variable taps how much the respondent liked and disliked the presidential candidates Romney and Obama.To measure the latter, respondents were asked to indicate how much they liked or disliked each presidential candidate on a scale ranging from 0 (strongly dislike) to 10 (strongly like).The likes/dislikes question allows a direct assessment of citizens' overall evaluation of the candidates.Citizens' candidate evaluations are often significantly linked to their actual vote choice [63].

Independent Variables
Attention to media sources: The key independent variable was media attention to presidential campaign information.Specifically, we sought to measure how much respondents relied on specific offline and online media outlets during the campaign.In this research, traditional media includes three offline media outlets: television, hard-copy newspapers and radio.Online media is considered media content that is accessible on many different forms of digital devices.In this analysis, attention to eight online media sources are evaluated.In each survey, Boomers and Millennials were asked the following question: "How much attention did you pay to information on [television] about the campaign for president?" (1 = none, 2 = very little, 3 = some, 4 = quite a bit, 5 = a great deal).In subsequent questions, the phrase "television" was replaced with the words "hard-copy newspapers," "radio," "online newspapers," "Facebook," "Twitter," "Tumblr," "YouTube," "political blog," "television network websites" and "presidential candidate websites."Descriptive statistics for media attention variables are reported in Table 2. Control variables: Two categories of control measures, demographics and political predispositions, were included.Key demographic variables, such as age (in years), gender (1 = male, 0 = female) and race (1 = white, 0 = non-white), were asked of respondents.Both age cohorts were asked to indicate their education level, but with different question wording.For Boomers, respondents were asked "What is your highest level of education?"(1 = less than high school; 2 = high school/GED; 3 = some college; 4 = Associate's degree; 5 = Bachelor's degree; 6 = Master's degree; 7 = Doctoral degree (PhD); and 8 = professional degree (MD, JD).For Millennials, respondents were asked: "What is your year in school?"(1 = first year; 2 = sophomore; 3 = junior; 4 = senior).Adapted from the ANES survey, two political predispositional variables, political interest and party identification, were also included.
For political interest, respondents were asked "Some people don't pay much attention to political campaigns.How about you?Would you say that you have been very much interested, somewhat interested or not much interested in the political campaigns so far this year?"(1 = not interested at all, 3 = moderately interested, 5 = extremely interested).To measure partisan attachment, two questions were used: (1) "Generally speaking, do you usually think of yourself as a Republican, a Democrat, an Independent, or something else?" and ( 2) "[If respondent indicated Republican or Democrat]: Would you call yourself a strong Democrat/Republican or a not-very-strong Democrat/Republican?" Answers to both questions were combined into a five-point scale (1 = strong Republican, 2 = Republican, 3 = Independent, don't know, no preference, 4 = Democrat, 5 = strong Democrat).Descriptive statistics for demographics and political predispositions are reported in Table 1.

Results
To address Research Question #1, Boomers and Millennials were asked to indicate how much attention they paid to various offline and online media outlets for information about the presidential campaign.To test Hypothesis #1, frequency statistics, along with two-sample t-tests between proportions, were performed to determine if there is a significant difference between Boomers and Millennials with respect to their media attention (see Table 2).We find that Boomers devoted significantly more attention to hard-copy newspapers for campaign information than young people, t(777) = 8.70, p = 0.001.In contrast to H#1, there were no significant differences between the two age groups regarding the amount of attention to television and radio.In fact, both age cohorts heavily relied on television for presidential campaign information, as both Boomers and Millennials reported paying attention to it the most.The most significant media attention differences pertained to attention to online media outlets for presidential campaign information, with Millennials depending more on online sources than Boomers.Specifically, Millennials paid significantly more attention to television network websites, presidential candidate websites, Facebook, YouTube, Twitter and Tumblr than Boomers.There were no significant differences between the two age groups regarding the amount of attention to online newspapers and political blogs.
Overall, providing modest evidence for H#1, young adults more frequently paid attention to online sources for campaign information while older adults turned more attention to hard-copy newspapers.Both age cohorts paid attention to a mixture of traditional and online media sources for political information.The latter is consistent with research indicating that older adults supplement traditional media with online sources [31].Similar to Millennials, Boomers are supplementing their traditional (offline) political information diet with online newspapers and political blogs.
To tackle Research Questions #2 and #3, we asked both age groups how much they trusted the media to cover political events fairly and accurately as well as how they would rate the media's performance in covering politics.Our results indicate that media trust and performance ratings are low among both Boomers and Millennials.Around only 17%-18% of both seniors and young adults reported that they either strongly agreed and agreed that they trust the media (t(776) = 1.34, p = 0.179).Despite this low level of trust, Boomers and Millennials rated the media's performance at somewhat better levels; however, Millennials rated the media (39%) significantly higher than Boomers (34%) (t(789) = 3.14, p = 0.001).
Given the low level of media trust and performance ratings among Boomers and Millennials, it is important to test how Boomer and Millennials' attention to specific media is associated with their media trust and performance ratings (H#2 and H#3).As Table 2 shows, both age groups depended the most on television for campaign information.Therefore, drawing on the media dependency theory [22,[53][54][55], a significant link between Boomer and Millennials' television attention and their media trust and performance levels would be expected, as both age cohorts heavily relied on television.In addition, one may anticipate a significant link between Millennials' attention to online sources and their media trust and performance, as Millennials devoted more attention to online sources, particularly Facebook and television network websites, than Boomers.In contrast, no link is expected between Boomers' attention to online sources and their media attitudes, as Table 2 shows that Boomers did not rely on online media for campaign information.Instead, Boomers depended heavily on traditional media sources: television, hard-copy newspapers and radio.
To test these expectations, we conducted a series of ordered probit regression analyses, which indicate that attention to a majority of media sources, both offline and online, is not significantly associated with media trust or performance levels among Boomers and Millennials.See Table 3.There is one notable and consistent media link across age groups, however.Consistent with H#2 and H#3, Boomer and Millennials' attention to television, their number one source for campaign information, is significantly related to their media trust and performance.Young adults' attention to television network websites was also significantly linked to their media trust and performance levels (Columns 2 and 4).The latter relationship is consistent with H#2 and H#3, as Millennials depended more on television campaign websites for campaign information.Although Facebook was a top source for campaign information, attention to Facebook had no significant link to Millennials' media trust and performance ratings.As expected, there was no link between Boomers' attention to online sources and their media attitudes, with one exception.Interestingly, Boomers' attention to political blogs is associated with their media trust and performance (Columns 1 and 3).The latter is surprising as Boomers did not indicate depending heavily on blogs for political information (see Table 2).Notes: Cell entries are ordered probit regression coefficients with fully-standardized coefficients in parentheses.Fully-standardized coefficients were calculated using "listcoef, std help" post-estimation for the "oprobit" command in Stata 13; * p < 0.05; ** p < 0.01 (two-tailed); a The question wording for education level differed for Boomers and Millennials; for Boomers, respondents were asked "What is your highest level of education?"(1 = less than high school; 2 = high school/GED; 3 = some college; 4 = Associate's degree; 5 = Bachelor's degree; 6 = Master's degree; 7 = Doctoral degree (PhD); and 8 = professional degree (MD, JD); for Millennials, respondents were asked: "What is your year in school?"(1 = first year; 2 = sophomore; 3 = junior; 4 = senior).
Last, regarding the control variables, the results in Table 3 show a consistent, negative relationship between political interest and media trust and performance across both age cohorts, suggesting that higher political interest equals lower media trust and performance ratings; and vice versa: citizens who have higher media trust and performance ratings have lower political interest.
Next, addressing Research Question #4, we examine Boomer and Millennials' overall evaluations of the presidential candidates: Romney and Obama.The measures of candidate evaluations seek to capture respondents' considerations or views of the candidates' attributes that contribute to citizens' overall evaluations.Consistent with our expectations (H#4a), as well as prior reports [31], frequency statistics along with two-sample t-tests between proportions indicate that Millennials were significantly more positive about Obama (mean = 5.57) than Boomers (mean = 4.50) (t(738) = 4.32, p = 0.000).In contrast, Boomers were significantly more positive about Romney (mean = 5.04) than Millennials (mean = 4.59), but the difference between the age cohorts was just under the p = 0.10 threshold (t(704) = 1.76, p = 0.08).
As posed in Hypothesis #4b, we seek to assess how Boomer and Millennials' attention to certain media outlets influences their candidate evaluations.Again, the media dependency theory would suggest that the media outlet depended on the most, television for both age groups, for campaign information would be associated with their candidate evaluations.Since Millennials devoted more attention to online sources, particularly Facebook (see Table 2), a significant link between young adults' attention to certain online outlets and their candidate evaluations is expected.Boomers allocated less attention to online sources, so little to no link with their candidate evaluations is anticipated.Instead, since Boomers relied more on traditional sources, particularly hard-copy newspapers and radio, for campaign information, one may anticipate a link between Boomers' attention to more traditional outlets and their candidate evaluations.
To examine the aforementioned relationships (H#4b), Table 4 reports ordinary least squares estimations when the Obama and Romney evaluation scores were regressed against attention to offline and online media sources.As the media dependency theory suggests, Boomers' attention to traditional media sources-television, hard-copy newspapers and radio-for campaign information significantly predicts their candidate evaluations (Columns 1 and 2).Specifically, attention to television and radio were positively associated with Boomers' Romney evaluations.Boomers' attention to hard-copy newspapers and radio were significantly linked to their Obama evaluations.Interestingly, attention to radio was a consistent predictor of older adults' candidate evaluations, with a positive link to Romney and a negative link to Obama.Although Millennials paid the most attention to television for campaign information (see Table 2), there was surprisingly no link between television attention and Millennials' candidate evaluations (Columns 3 and 4).The latter relationship contrasts with H#4b.Instead, online sources, particularly Facebook and candidate websites, played a more important role in predicting Millennials' candidate evaluations.As anticipated, few online sources were associated with Boomers' candidate evaluations.For Boomers, only two online sources-online newspapers and Facebook-were significantly associated with their candidate evaluations, specifically for Romney.There were no online media sources linked to Boomers' Obama evaluations (Column 2).In general, it is notable that attention to many media sources-both offline and online-had little to no impact on candidate evaluations.Clearly, Boomers and Millennials' party identification was the strongest predictor of their candidate evaluations, with Democrats evaluating Obama at higher levels and Republicans evaluating Romney a higher levels.In addition, Columns 1 and 2 show that political interest was a consistent and significant predictor of Boomer candidate evaluations.

Discussion and Conclusions
It is well known that online media outlets play an important role in political news consumption.The reliance on news sources and media platforms, however, can vary by age, education, party identification and other factors.Indeed, young people are more likely to consume digital news while older adults prefer traditional media outlets [31].Thus, we examine Baby Boomer and Millennials' attention to 2012 presidential campaign information on specific offline and online media.Notably, we found that the top outlet for campaign information among both Boomers and Millennials was television.There were no significant differences between the age cohorts on attention given to television and radio.Our results continue to support the belief that a digital generation gap still exists when it comes to political media attention.Millennials are more dependent on online sources for presidential campaign information, such as Facebook, whereas older adults devote more attention to offline, traditional media, such as hard-copy newspapers.These findings do not suggest, however, that Millennials are only paying attention to digital sources and Boomers are only paying attention to traditional media sources.Boomers reported paying some attention to political content in online newspapers and Facebook.Millennials are indeed still relying heavily on television with a bit of attention to radio.Overall, we can confirm that television is the top source for political content among both Millennials and Boomers during a presidential campaign.Most importantly, we offer evidence that Boomers are now turning to some online sources for political information.Millennials have an eclectic news diet, which is mostly reliant on television with a heavy mixture of online sources.Boomer and Millennials' dependence on various media outlets indicates a need for campaigns to employ multiple media outlets to distribute political information effectively across age groups.
We also examine the similarities and differences regarding Boomers and Millennials' media trust and performance levels.Media trust and performance levels are important during a campaign period, as media is clearly the main source of political information for both age groups.We find that Boomers and Millennials have low media trust and performance levels, indicating that both age cohorts feel that media outlets are not fulfilling their role as political and civic educators.The latter findings should not be surprising, as citizens' trust in the mass media has dropped during recent election years, indicating that national elections produce lower trust in political media coverage [66].Millennials and Boomers only significantly differed on their media performance ratings, with Millennials rating media somewhat higher than Boomers.
Given that the media attention levels of Boomers and Millennials are different, we sought to examine how their attention to certain outlets are linked to their media trust and performance levels.In the framework of the media dependency theory: television is the most relied on source for campaign information among young and old age cohorts and, therefore, is significantly and positively related to media trust and performance both Boomers and Millennials.This finding provides evidence supporting the media dependency theory, suggesting that the more dependent an individual is on a specific media outlet, the more important and influential the media outlet will be to that individual.Here, more reliance on television for campaign information predicts media trust and performance, even among very different age groups.
As this study shows, Millennials depend more on online media sources than Boomers.Online media sources were indeed secondary to television for Millennials.Considering the media dependency theory, one could argue that young adults' reliance on online sources would be linked to their perceptions of media trust and performance.However, we find that attention to most online sources for campaign news did not translate into higher media trust or performance for Millennials.In fact, this was the case for both age groups.Despite increasing use of online sources, we find that attention to social media and micro-blogs, particularly Facebook, Twitter, Tumblr and YouTube, had no significant link to Boomers and Millennials' media trust and performance ratings.One may speculate that no relationship exists because online sources have no clear delineation of ownership, campaign information is largely user-generated news from outside controlled establishments and political information is not objective or balanced.This suggests that there is some skepticism among both Boomers and Millennials about gathering campaign information from social media or microblogs that mostly "share" information from friends, family and others followed or fanned.That is, social media sources may be simply entertainment rather than sources for political information.Of course, this study does not investigate the psychological motivations of Boomers and Millennials, and further research is necessary to find the reasons for their behavior.
Next, we find that Boomers and Millennials evaluated Obama and Romney on very different levels, with young adults evaluating Obama much more highly, whereas Boomers evaluated Romney more highly.Interestingly, we find that the variation in offline and online media attention between age cohorts predicts candidate evaluations in different ways.In line with the media dependency theory, Boomers' heavy reliance on television, hard-newspapers and radio for campaign information was linked to their candidate evaluations.Yet, we found no association between Millennials' attention to television and their candidate evaluations.This cuts against the framework of the media dependency theory, as Millennials' reported depending on television the most for campaign information.The latter findings are important because it was previously not clear if the increasing attention to online sources for campaign information influenced perceptions of candidates among different age groups.Our findings provide some evidence that online media sources played a more important role for Millennials regarding candidate perceptions and only a minimal role for Boomers.
Furthermore, these findings suggest that the traditional media's ability to communicate campaign information about presidential candidates themselves-leadership qualities, experience and personality traits-has greater impact on older rather than younger age groups.This should not be surprising, as traditional media has been Boomers' central, if not only, source for campaign information.Their life-long dependency on three television networks and a few national-level newspapers continues to satisfy their need for election information, and no alternative source fulfils that need.Millennials, however, have grown up in a vastly different communication environment, one with many alternative forms of media.As the findings show, Millennials' attention to Facebook and candidate websites predicted their candidate evaluations rather than television.Since the media dependency theory was originally developed during the golden age of traditional media, it perhaps does not account for the current communication landscape.Due to the increasing variety of offline and online media sources, it may be time to broaden the scope of the media dependency theory to include different levels, types or strengths of dependency [86].Millennials are not exclusively depending on television for campaign news, nor have they relied on a handful of sources for a lifetime.Millennials depend on a variety of offline and online sources that may constitute an important "information mix" during political campaigns.For example, Millennials may rely on social media sources for candidate traits and characteristics and also rely on television for general campaign information.
Last, it is important to note a few limitations of this study.Our research relies on separate convenience samples of both Baby Boomers and Millennials; therefore, we cannot generalize our results to national samples of youths and older adults.In addition, these samples are one-shot, cross-sectional samples collected during an election period.Cross-sectional survey data are limiting, as they do not allow us to infer a causal direction between media attention and our dependent variables.Indeed, panel data would be a better empirical test of the causal relationship between media attention and media trust and performance, as well as candidate evaluations among different generational groups.It is likely that media preferences and attention to political information change throughout the campaign period.Therefore, it is important that scholars build on this case study, examining online media attention and its impact on political attitudes among specific demographic groups.Clearly, we show that different age cohorts pay attention to different media for politics.As new media platforms change and citizens' news habits evolve, it is critical to continue to examine the relationship between media information and attitudes.

Table 1 .
Demographic and political characteristics of Boomers and Millennials.

Table 2 .
Comparing media attention among Boomers and Millennials.

Table 3 .
Boomer and Millennial media trust and performance levels by media attention.

Table 4 .
Boomer and Millennial candidate evaluations by media attention.