Professional Documents
Culture Documents
Communication Modes and Candidate Image 1
Communication Modes and Candidate Image 1
NCA 2007
Evaluations
This study was designed to examine how young people respond to political candidates by
examining the effects of exposure to the messages presented through television advertisements,
televised debates, and campaign Web sites on ratings of the candidates’ images. Using a 3
(communication mode) x 3 (political party) MANOVA procedure, this study tested two
hypotheses and one research question. The results indicated significant differences in the
candidates’ image ratings based on the type of campaign communication mode to which the
subjects were exposed and significant differences with regard to the importance of political
Evaluations
Research on political candidate images has long supported the argument that voters make
voting decisions based on the perceived image of the candidate (for a review see Hacker, 1995,
2004). As presidential campaigns spend enormous amounts of money on such message outlets
as political advertising, time in negotiation and preparation for televised debates, and creative
energies in formulating messages for multiple audiences on their campaign Web sites, they do so
in order to persuade voters to vote for them. Although multiple sources of campaign information
assist voters in making evaluations about candidate images (Pfau, Cho, and Chong, 2001),
televised advertising and campaign Web sites represent two distinct sources of mass
communication over which candidates can exert the most control over message, design, and
image, as well as reach the voters directly with their message. In other words, the message has,
as Kaid (2002) points out, a linear, “direct effects” transmission (p. 28), uniquely different from
the third party interpretation involved with other transmission sources, such as news media
coverage, talk shows, and blogs. From a mass communication perspective, debates also serve an
important function for candidates. Although the candidates may not have same element of
control, at the moment of message consumption by the viewer they do have the opportunity for
In 2004, young voters were the target of not only these forms of campaign-related
messaging, but also messages from non-profit, governmental, and media sources. The past
decade has witnessed increased concern over the seemingly disengaged young voter population.
Organizations such as MTV, the National Secretaries of State, the Center for Information and
Research on Civic Learning (CIRCLE), the American Democracy Project, the Harvard
Communication Modes and Candidate Image 4
University Institute of Politics, and Campus Compact have pursued initiatives designed to
increase civic engagement among young people, often primarily in the form of voting. These
groups were rewarded with exciting results when the 2004 votes were tallied. In total, more than
11.6 million American youth voted in the presidential election (Lopez, Kirby, & Sagoff, 2005).
Young voter turnout was at a twelve-year high in the 2004 general election with a 47% voter
turnout (Lopez et al.). Considered a triumph, the young voter demographic showed an 11%
increase from the 2000 to the 2004 election—the largest increase of any age demographic during
that time (Lopez et al.). Young voters also represented a larger share of the total votes cast in
2004 (9.3%) than in any of the previous eleven years (Lopez et al.).
Perhaps not surprising, based on increased turnout, the Pew Research Center reported in
the fall of 2004 that young people were more interested in the election and in voting than four
years earlier, although they were less firm in their candidate preference (Pew, 2004). It is the
evaluation of candidates that leads to candidate selection, with which this study takes interest.
With the high interest in improving young people’s turnout at the polls and the subsequent
increase in young voter turnout, the 2004 general election is a unique case study for examining
how young voters responded and connected to the candidates. Specifically, this study seeks to
understand the relationship between young voters image perceptions of the presidential
Literature Review
Communication Modes and Candidate Image 5
Over the past several decades presidential candidates have enjoyed the opportunity to
reach millions of Americans through mass forms of campaign communication, such as television
spot ads and televised debates; and, in the past decade campaign Web sites have emerged as an
important campaign tool. In fact, Kaid and Postelnicu (2004) argue that in 2004 the presidential
candidates’ Web sites became “a very visible part of the campaign communication” (p. 265).
This recent infusion of the Internet and campaign Web sites into the realm of political
communication has again raised the issue of whether message channel, or mode, influences
audience perception and interpretation of campaign messages. In other words, how does our
evaluation of candidates differ when we are exposed to the differing messages available through
Katz and Feldman’s (1962) well-known study of the 1960 Kennedy-Nixon debate
provided early indications that in fact channel does matter, finding that declaration of a candidate
as the debate winner was dependent on the channel to which the audience was exposed. These
results, along with McLuhan’s (1964) suggestion that message is intertwined with medium, lead
many to argue that different effects emerge from exposure to different media (Kaid, 1981; Kaid,
2002; Kaid & Postelnicu, 2005; McKinnon, Tedesco, & Kaid, 1993). However, as new
technologies offer campaigns new channels through which to disseminate messages and
influence voter attitudes and perceptions, the study of message channel—or mode—remains an
extensive and rich understanding of message effects, the frequent choice of scholars to examine
individual campaign communication modes in isolation from other modes does not reflect the
Communication Modes and Candidate Image 6
complex mediated nature of modern campaigns. Pfau, Cho, and Chong (2001) argue that “this
isolated approach is flawed” (p. 88) and call for more work in the study of multi-mode effects. It
is well documented that voters are exposed to many campaign-related messages throughout a
presidential election cycle and scholars have sought to determine which serve as primary sources
of influence. Yet, few studies have focused on a comparison of the different effects that might
particularly the effects these messages have on candidate evaluations and voting behavior.
Indeed, the few studies that have examined to what extent such effects might exist find
support for the presence of mode-based influence, although the inconsistent results do not offer
firm answers. For instance, during the 1992 election McKinnon, Tedesco, and Kaid (1993)
compared reactions to a presidential debate that was simultaneously aired on television and
radio. While the results revealed significant changes in image perception for some of the
candidates based on the mode of communication, the findings were not consistent across all
candidates. In her study of effects resulting from traditional (television) versus nontraditional
(internet) exposure to the same campaign advertising messages during the 2000 presidential race,
Kaid (2002) determined that in fact vote choice changed based on communication mode
exposure although evaluations of the candidates did not. In 2000, Bush benefited from exposure
through television and Gore benefited in the Internet exposure group. Kaid and Postelnicu
(2005) used a similar approach during the 2004 election, exposing a young voter audience to the
same campaign ads on television and online. Their results did find that evaluations of the
candidates significantly changed from the pretest to posttest when voters viewed the ads through
the candidates’ Web sites; Kerry’s evaluations increased significantly and Bush’s evaluations
decreased significantly. However, candidate evaluations did not differ when the candidate ads
Communication Modes and Candidate Image 7
were viewed via television. Importantly, participants from the Internet group reported perceiving
the ads as significantly less credible than did the television group.
During the 2000 presidential campaign Pfau et al. (2001) examined the influence of 13
“modalities” on voters’ perceptions of candidate image for the two major party presidential
candidates. Using a survey approach, the authors asked respondents to report their
communication modality use and attention to messages, some of which were candidate-
controlled and most of which were not. Their results indicated that the use of “non-traditional”
forms of communication (i.e., political talk radio, television entertainment talk shows, television
news magazines) had the most influence on perceptions of candidate image. While debates—a
traditional forms such as newspapers, magazines, and television news exerted limited influence.
Although the previous studies differ in purpose and design from the current study, they
provide support for the notion that the different communication modes can produce different
effects among voters. Clearly, each of the individual modes of communication are important for
—provide an interesting point of comparison with regard to their effects on candidate assessment
and voting behavior. Therefore, this current study seeks to expand the prior work on campaign
communication effects across multiple modes. Specifically, this study explores differences in
young voters’ perceptions of candidate image based on exposure to messages across three
different forms of campaign media. Image effects research based on each of these
messages indicate that exposure to these forms of communication during a campaign aids voters
in evaluating candidates and making vote choices. Viewing television campaign advertising
typically results in positive effects on assessments of candidate image, although negative effects
have been noted (Kaid, 2004; Tedesco & Kaid, 2003). In terms of positive effects, televised
campaign ads have been found to increase candidate name recognition (Kaid, 1982) as well as
improve overall evaluations of candidate image following exposure to the ads (Kaid & Johnston,
2001; Tedesco & Kaid, 2003). Exposure to political advertising has resulted in higher candidate
ratings for characteristics such as intelligence, strength, dependability, honesty, and fairness
(Cundy, 1986). In terms of negative effects on candidate assessment or image, findings typically
reflect that exposure to negative advertising produces negative image evaluations of the
opponent (Friedkin & Kenney, 2004; Jasperson & Fan, 2002; Kaid, 1997; Tedesco & Kaid,
2003). The positive and negative effects are important findings because both indicate that
predictor of voter decision-making (Miller, Wattenberg, & Malanchuk, 1984; Nimmo & Savage,
1976).
The literature on political debates also suggests that viewers develop perceptions of
candidates’ images as a result of exposure to a debate (Benoit, McKinney, & Stephenson, 2002;
Best & Hubbard, 1999; Dailey, Hinck, & Hinck, 2006; McKinney, Dudash, & Hodgkinson,
2003; McKinney, Kaid, & Robertson, 2001; McKinnon, Tedesco, & Kaid, 1993) and that
debates provide an opportunity for voters to assess and better understand the candidates’
personalities and images (McKinney et al., 2003). Specifically, candidate image evaluations
have shifted significantly based on debate viewing (Best & Hubbard, 1999; Dailey, Hinck, &
Communication Modes and Candidate Image 9
Hinck, 2006), with changes occurring in evaluations of the candidate’s competency (Dailey et
al., 2006), character, and viability (Best & Hubbard, 1999). Lanoue and Schrott (1991) conclude
that “viewers are far more likely to use debates to gain insight into each candidate’s personality
and character” (p. 96) than their issue positions. Because of the length of exposure time that is
available to voters for evaluative purposes, as well as the opportunity for direct candidate
comparison and for the candidates to in essence talk “directly” to the viewers; debates remain an
opportunity to exert some control over their preparation, responses, structure, and issues covered
during a debate (Self, 2005). Therefore, this element of control and opportunity for the
candidates to speak directly to the viewing audience make campaign debates an important mode
of communication to study in comparison with television advertising and the more recently
Due to the relatively recent emergence of campaign Web sites as modes of campaign
communication, research is advancing rapidly in its analysis of Web site exposure effects on
voter evaluations of candidates’ images and on voter behavior. In their study of the effects of
Web site design and complexity on candidate image, Sundar, Kalyanaraman, and Brown (2003)
found that more complex Web site designs led to lower candidate evaluations when compared to
less complex Web sites. In this experimental design, participants were exposed to one of three
Web site designs—low, medium, and high interactivity—for a fictitious candidate. Candidate
image ratings were significantly lower in the high interactivity group than either the low or
medium interactivity group. The authors concluded that Web site designers need to be “cautious
in their use of new media tools” (p. 52), and to keep the message as the focus of a campaign site
Tedesco’s (2006) study of young voter efficacy and Web interactivity found that
exposure to the interactive sites increased participants’ internal and external efficacy, and their
feelings of being informed. Other scholars have confirmed that in fact exposure to campaign
information online significantly impacts political knowledge, political efficacy, and participation
(Kenski & Stroud, 2006). Internet use for obtaining political information has also been found to
involvement. Beyond the work by Kaid and colleagues (Kaid, 2002; Kaid & Postelnicu, 2005),
little attention has been paid to understanding how voters develop image perceptions of the
Therefore, this study seeks to build upon prior research on advertising and debate effects
and expand research on campaign Web site effects to examine if and how perceptions of
candidate image develop differently for young voters across these three communication modes.
Delli Carpini (2004) argues that little is known about the impact of different communication
modes on groups, such as young voters. The examination of this target group and of three mass
communication modes offers the potential for a more comprehensive analysis of message effects
than the prior research has thus far assessed, as well as for a greater understanding of what
messages young people are more likely to process. To achieve this purpose, the study first poses
RQ: Does the message mode influence how young people rate the candidates’ images?
A Pew poll (2005) reported that 52% of those who access the Internet for their political
information indicated that “the internet was important in giving them information that helped
them decide how to vote” (p. ii). While convenience remains the main reason people get their
political news online, more than half reported getting their news online because they prefer the
Communication Modes and Candidate Image 11
ability to seek out information beyond that provided by the mainstream media (Pew, 2005).
Additionally, young people are reporting frequent Internet access for news gathering (Pew, 2005)
and Pew reported that 57% of these young web users got political information online for the
2004 election. Tedesco (2004) suggests that this increase in the Web as a source of political
news leads to the likelihood for young voters to access the Internet as the communication
medium of choice for their political information rather than traditional media (i.e., television
news, newspapers). Given the increased use of the Internet among this age group, there is a need
to examine whether this choice influences important variables such as candidate image
evaluations and how this mode of campaign communication compares to other modes for this
specific demographic. The findings of previous research with regard to young people’s media
consumption and political campaign communication modes, leads to the following hypothesis:
H1: Of the three modes tested, the candidate image ratings will be highest when the
Political Affiliation
Political affiliation has also been well documented as a strong indicator of voting
behaviors and candidate image evaluations, with political party affiliates preferring their party’s
candidate (for a review see Kaid, 2004; Paletz, 2002; Rahn, 1993; Powell & Cowart, 2003).
Considering that political socialization develops during young adulthood (Paletz, 2002), it is
quite important to learn whether political party is as strong of an indicator of voting behaviors
and candidate assessments in this demographic as it is in the adult demographic. Given that
research has repeatedly supported the influence of political affiliation within the general public,
the current study does not assume that the relationship will be different with young voters and
H2: Candidates will receive higher image ratings from young people affiliated with their
political party than from Independents or those affiliated with the other political party.
Method
A quasi-experimental procedure was used to respond to the research questions and to test
the hypotheses. In order to examine the reactions to each of the three types of messages—
televised advertisements, debates, and Web sites—groups of young voters were exposed to the
Participants
large midwestern university. The participants signed up for one of the three exposures at random
and received course credit for their participation. Fifty-four percent of the participants were
women (n = 226) and 46% were men (n = 193) with a mean age of 19.97 (range = 18-25). The
American, 3.6% as Asian/Pacific Islander, 3.6% as Hispanic, 3.3% selected “other,” and 3%
refused to report their race/ethnic background. The political party affiliation of the participants
was evenly split at 35% Democrats (n = 147) and 35% Republicans (n = 147); 30% of the
Two hundred thirty seven participants were exposed to the televised campaign
advertisements, 69 were exposed to the televised debates, and 113 were exposed to the campaign
Web sites. While efforts were made to generate comparable sample sizes, the study did result in
unequal sample sizes across the three communication modes therefore, careful attention to the
Procedure
Communication Modes and Candidate Image 13
The data collected for this study were gathered in conjunction with research being
conducted by the UVote 2004 national research team project. The items analyzed in this study
are a combination of questions included on the UVote questionnaire and items added specifically
for this site location. Therefore, this study reports findings from the local site location only. The
television advertising and debate stimuli to which participants were exposed were coordinated
A pre-test/post-test survey design was used for each of the sessions. In the pre-test
images, vote choice for a presidential candidate, strength of that vote, and questions pertaining to
sources to which they turn for political information acquisition. The participants were then
candidates, a televised debate, or the presidential candidates’ Web sites. The participants then
completed the post-test section of the questionnaire, rating their perceptions of the candidates’
images, identifying their vote choice, strength of vote, and providing responses to open-ended
The television advertising group viewed one of two sets of advertisements for George W.
Bush and John Kerry.1 The first set of advertisements were shown during sessions held from
September 28 to October 2, 2004, and the second set of advertisements were shown during
sessions held from October 26 to October 29, 2004. The debate group viewed one of two
debates that aired nationally on September 30, 2004 and on October 8, 2004. Each debate
session was held in a common area of a dormitory on the university’s campus where television
The Web site group viewed the presidential candidate Web sites in a computer lab on the
university’s campus, where access to the Internet was consistent for all participants (i.e.,
download time was the same). Specific instructions for viewing the Web sites were provided for
each session. Participants were asked to open their Internet browsers to the first candidate’s
official Web site and were given eight minutes to view the Web site. The participants were
given no further direction regarding what areas of the site to view. At the end of eight minutes,
they were directed to the second candidate’s official Web site and given eight minutes to view
this site. The Web site viewed first was counterbalanced across groups of participants
(johnkerry.com or georgewbush.com). After viewing both Web sites the participants were asked
to close their Web browsers and continue with the last portion of the questionnaire. The Web
site sessions were conducted between Tuesday, October 19 and Tuesday, October 26, 2004, just
Instruments
Candidate image was measured with a semantic differential scale widely used in research
on political candidate image (e.g., Kaid & Chanslor, 1995; Kaid et al., 1977; Kaid & Tedesco,
1999; Sanders & Pace, 1977; Tedesco & Kaid, 2003). The scale utilized a series of twelve
bipolar adjective pairs, rated on a seven-point scale, describing characteristics of candidate image
research mean candidate image ratings were calculated for each participant. When used in
previous research the scale achieved high reliability with Cronbach’s alpha levels ranging
from .83 (Kaid, Downs, & Raga, 1990) to .92 (Kaid, Leland, & Whitney, 1992). In this study,
Communication Modes and Candidate Image 15
reliabilities for each dimension were assessed using Cronbach’s alpha, which resulted in an alpha
Party identification was measured by asking participants “Which of the following best
represents your political beliefs?” Participants were given four options to choose from
(Democrat, Republican, Independent, Other). For the purposes of data analysis, participants
Pre-test evaluations
speaking with others about the campaign, pre-test data was examined to determine if differences
existed between participants across the three modes on the candidate image variable. Two, one-
way analysis of variance tests were conducted to evaluate the relationship between ratings of
Bush and Kerry’s image and participants in each of the three modes of exposure—televised ads,
television debates, and campaign Web sites—prior to exposure to the stimulus. The ANOVA
test for Bush’s image ratings across modes was not significant, F(2, 146) = 1.20, p = .30 and the
ANOVA for Kerry’s image ratings across modes was also not significant, F(2, 146) = .835, p
= .44. In other words, participants did not differ significantly across the three different modes in
their ratings of Bush’s image nor in their ratings of Kerry’s image prior to exposure to the stimuli
in this study, despite the potential for varied exposure to campaign messages or discussions
about the candidates occurring external to this study. Given this lack of significance in the pre-
test data, any significant differences in the ratings of candidate image following exposure were
Results
Communication Modes and Candidate Image 16
This experiment tested two hypotheses to examine the influence communication mode
and party affiliation on candidate image. It also addressed one research question regarding the
and televised debates—on ratings of candidate image. The hypotheses and research question
variance procedure for judgments of the two candidates on the dependent variable.
Hypothesis 1 predicted that candidate image ratings would be highest when the messages
were viewed through campaign Web sites when compared to the other modes of communication.
To test this hypothesis, a multivariate analysis of variance was conducted with the image ratings
for each candidate as the dependent measures. A preliminary analysis evaluating the covariance
matrices of the dependent measures across groups indicated that the relationship between the
groups was not equal, Box’s M = 209.28, F (24, 72260) = 8.50, p = .000. Due to the unequal
variances, Pilai’s Trace, which takes this violation in account, is reported for this analysis.
Results from this analysis revealed significant main effects of party and communication
mode, and a significant party by communication mode interaction (see Table 1). Univariate tests
showed that these effects were significant for both Bush and Kerry (see Table 1). Therefore, the
Bush Image
Simple main effects analysis indicated that communication mode had a significant effect
on Bush’s image ratings for Republicans, F (2, 144) = 59.08, p = .000, η2 =.45, and
Independents, F (2, 122) = 8.23, p = .000, η2 =.12, but not for Democrats, F (2, 144) = 1.86, p
> .05, η2 =.03. Pairwise comparisons (Green & Salkind, 2003) revealed that Republicans who
Communication Modes and Candidate Image 17
viewed the debates rated Bush’s image significantly higher than those who viewed the televised
ads and his campaign Web site (see Table 2). Independents rated Bush’s image significantly
higher in both the debates and his campaign Web site than in televised ads (see Table 2).
Simple main effect analyses also indicated that political party had a significant effect on
Bush’s image ratings in each of the three modes of campaign communication: Ads, F (2, 234) =
37.95, p = .000, η2 = .25; Debates, F (2, 66) = 31.44, p = .000, η2 = .48; Web sites, F (2, 110) =
16.38, p = .001, η2 = .23. Pairwise comparisons (Green & Salkind, 2003) revealed that
Republicans rated Bush’s image significantly higher in his ads than did Independents and
Democrats. With regard to the debates, participants from each political party varied significantly
in their ratings of Bush’s image with Republicans rating his image highest, followed by
Independents, and Democrats rating his image lowest. Finally, Independents rated Bush’s image
significantly higher in his Web site than did Republicans and Democrats.
Kerry Image
Following the procedures outlined by Green and Salkind (2003), simple main effects of
communication mode on image ratings for each candidate were examined within political party.
These analyses indicated that judgments of Kerry’s image varied significantly across the three
communication modes for participants in each political party: Democrats, F (2, 144) = 33.64, p =
.000, η2 = .32; Republicans, F (2, 144) = 9.64, p = .000, η2 = .12; Independents, F (2, 122) = 4.25,
p < .05, η2 = .07. Pairwise comparisons (Green & Salkind, 2003) revealed that Democrats who
viewed the debates rated Kerry’s image significantly higher than those who viewed televised ads
and his campaign Web site (see Table 2). In contrast, Republicans who saw the debates rated
Kerry’s image significantly lower than those who viewed the televised ads and his campaign
Communication Modes and Candidate Image 18
Web site (see Table 2). Finally, Independents who viewed the debates rated Kerry’s image
significantly higher than those who viewed his campaign Web site, while ratings from
Independents who viewed the ads did not differ significantly from those for the other two
Simple main effect analyses also indicated that political party had a significant effect on
Kerry’s image ratings in each of the three modes of campaign communication: Ads, F (2, 234) =
19.04, p = .000, η2 = .14; Debates, F (2, 66) = 33.31, p = .000, η2 = .50; Web sites, F (2, 110) =
7.69, p = .001, η2 = .12. Pairwise comparisons (Green & Salkind, 2003) revealed that Democrats
rated Kerry’s image significantly higher in his ads than did Independents and Republicans. With
regard to the debates, participants from each political party varied significantly in their ratings of
Kerry’s image with Democrats rating his image highest, followed by Independents, and
Republicans rating his image lowest. Finally, Democrats rated Kerry’s image significantly
higher in his Web site than did Independents, while the Republicans’ ratings of Kerry’s image
Summary
These results provide little support for Hypothesis 1 that participants who viewed the
Web sites would give candidates the highest image ratings. In two instances (Republicans rating
Kerry and Independents rating Bush) participants who viewed Web sites gave higher ratings to
the candidates than those who viewed one of the other communication modes. However in both
of those cases, the Web site ratings were equivalent to the rating from the third mode. Viewing
the debates led to the highest candidate image ratings in three cases: Republicans rating Bush,
Hypothesis two received more support with candidates receiving higher image ratings
Communication Modes and Candidate Image 19
from young people affiliated with their political party than from Independents or those affiliated
with the other political party, with one exception. For Bush, Republicans rated his image higher
in both his ads and the debates than both the Independents and Democrats. In the Websites,
however, the Independents rated Bush’s image higher than did the Republicans and Democrats.
John Kerry’s image was rated highest by the Democrats in each of the three modes of campaign
communication.
In this study, the overall mean image ratings of the candidates centered around the mid-
point of the image scale. The candidate image measure utilized a semantic differential scale with
a series of twelve bipolar adjective pairs, rated on a seven-point scale with higher values
representing higher candidate image ratings. For Bush’s image, the Democrats rated him below
the scale mid-point in each of the three forms of communication, while the Republicans and
Independents rated him above the scale mid-point in each of the three forms of communication.
In contrast, Kerry’s image was rated below the scale mid-point only once (Republicans rating in
the debates). Although several ratings were above the scale mid-point of “4.0,” the average
ratings for each candidate were never greater than 4.88, which indicate a relatively neutral image
Discussion
This study examined how young people respond to political candidates, specifically the
2004 Presidential candidates. Of importance was the effect of three candidate-controlled, mass
on how young voters rated the candidates’ images. From a review of the literature, the study
advanced one research question and two hypotheses. The research question inquired whether
message mode influenced how young people rated the candidates’ images and was coupled with
Communication Modes and Candidate Image 20
a hypothesis that predicted the candidate images would be rated most positively by the group
viewing the campaign Web sites. The results indicated that differences existed in the candidates’
image ratings based on the type of communication mode to which the subjects were exposed.
Of the three modes examined, the study hypothesized that the ratings of candidate image
by young people would be highest when viewed through the Web sites. This hypothesis was not
supported by the results despite research indicating an increase in the use of the Web as a source
of political information among the young voter demographic. For instance, the Pew Research
Center [Pew] reported that the Internet was an important source of information when deciding
for whom to vote for a majority (52%) of Internet users in the 2004 election (Pew, 2005).
Additionally, 57% of young voters who accessed the Internet during the campaign did so to
obtain political information (Pew). However, while young people may prefer to get their
political information online (Pew; Tedesco, 2004), the results in this study indicate that exposure
to candidate messages on campaign Web sites does not necessarily lead to higher evaluations of
candidate image when compared to evaluations generated after exposure to candidate messages
from other modes of campaign communication. While ratings of candidate image were not
significantly higher following exposure to the campaign Web sites, significant differences did
emerge with both candidate’s images being rated highest overall following exposure to the
campaign debates.
This increase in ratings of candidate image following exposure to the debates provides
support for McKinney, Dudash, and Hodgkinson’s (2003) claim that debates have an influence
opportunity for voters to assess and better understand the candidates’ personalities and images.
Specifically, this study’s results indicated that young people who were affiliated with a political
Communication Modes and Candidate Image 21
in the debates that in advertising or campaign Web sites, findings that confirm earlier work on
partisanship and debate effects (McKinnon et al., 1993; Payne, Golden, Marlier, & Ratzan,
1989). In other words, young Democrats rated Kerry’s image significantly higher when they
were exposed to the debates than when exposed to either the televised ads or the campaign Web
site; and young Republicans rated Bush’s image significantly higher in the debates than in the
televised ads or the campaign Web site. Conversely, young party affiliates rated the image of the
opposing party’s candidate lower in the debates when compared with the other modes of
campaign communication, although the some of these differences by mode were not statistically
significant.
While these findings do not support hypothesis one—which predicted candidate image
ratings would be highest when messages were viewed through campaign Web sites—they do
support hypothesis two—candidates will receive higher image ratings from young people
affiliated with their political party than from those affiliated with the opposing party. The results
indicate that, for young people affiliated with a political party, the debates provide a significant
point of comparison between the candidates that may serve to strengthen allegiance to the party
and/or the candidate. Similarly, McKinney and Carlin (2004) noted that debates result in very
little change in voters who have made a candidate choice, but instead debates are more likely to
serve as a means through which vote choices can be solidified for these viewers. As a result,
political candidates may be best served if young party affiliates are exposed to the presidential
Additionally, young people who considered themselves as Independents also rated each
candidate’s image higher following exposure to the debates when compared to other modes of
Communication Modes and Candidate Image 22
communication, thus further providing no support for hypothesis one that ratings would be
highest from exposure to the Web sites. Specifically, young Independents rated Bush’s image
significantly higher in the debates than in his televised ads, and, while not significant, they also
rated him higher in the debates compared to his Web site. For Kerry, the Independent voters
rated his image significantly higher in the debates compared to his Web site and higher in the
debates than his televised ads, although this last comparison was not significant. Thus, the
results of this study indicate that both of the candidates were best able to enhance their image
with young Independents during the debates, just as they were able to do with their own party’s
affiliates. Although the findings are not statistically significant in all instances (see Table 2), the
results are important in an effort to understand the value of campaign debates for presidential
Prior research has indicated that campaign debates may have more influence on
unaffiliated, undecided, and weakly committed voters than on decided voters or party affiliates
(Chaffee & Choe, 1980; Geer, 1998; Kraus & Davis, 1981; McKinney, 1994), particularly with
regard to candidate assessment and vote choice. McKinney and Carlin (2004) have argued that
viewing debates may aid in forming voting preferences, solidifying voting choices, and
providing reasons for a change in candidate selection for undecided voters. Specifically with
regard to the 2004 presidential election, Bush and Kerry courted “undecided voters” and
Independents. In addition, the “swing states,” those on which neither party had a strong hold,
Given that candidate image ratings were highest in the campaign debates and given the
previous literature attesting to the influence of debates on voting decisions for undecided and
unaffiliated voters, it seems that the debates are the mode of campaign communication where
Communication Modes and Candidate Image 23
candidates may best be able to positively influence their image with young Independent voters.
Unlike party affiliates, Independents do not have a personal party cue to assist them in evaluating
and making a connection with a candidate, which may explain why the debates afford them the
opportunity to gather information they can use in forming an image of the candidates to use in
their voting decisions. Although these results echo those that indicate the importance of debates
for party affiliates, the findings with regard to influence of debates on Independent voters
provide a strong argument for campaign debates as a vital mode of communication for
Presidential candidates.
Therefore, the findings of this study serve to reinforce the claim that campaign debates
are vital to the citizenry’s evaluations of the candidates because they serve as platform for
comparing the candidates (McKinney & Carlin, 2004; Racine Group, 2002). The Racine Group
claimed that debates do matter, citizens do learn from them, and citizens use the information in
their voting decisions. This study’s findings support that claim and confirm its application to
young voters in particular as an addition to those studies focused on the public in general.
Although the debates may serve as an important opportunity for young people to
formulate their perceptions of each candidate’s image, this effect may not be widespread given
that few young people actually watch the campaign debates in part or in full when compared to
the general public (Patterson, 2005). While 28% of adults (over age 30) watched all of the first
debate, only 14% of young adults (18-29 year olds) watched the full debate—half the number of
older adult viewers (Patterson). Even more alarming is that 41% of young adults did not tune in
to watch any of the first debate (Patterson). Considering this apparent lack of interest in the
debates by young people, perhaps more needs to be done to encourage young people to watch the
affords young people the opportunity to watch the debates and then discuss and clarify their
learning through open dialogue with other citizens. Buehler’s dissertation (2004) attests to the
importance of a public forum, such as DebateWatch, in which young voters in particular can
discuss their opinions and feel that they are valid. Given that campaign debates provide young
voters a valuable opportunity for learning about candidate images, programs and initiatives
whose purpose is to encourage political engagement of young people would do well to first
encourage young voters to tune in to the debates and then generate continued interest through
teaching young people how to seek out relevant political information through various sources,
such as campaign Web sites, newspapers, network and cable news, and other political news Web
sites. Providing additional emphasis that builds on what they learned in the debates may be
beneficial for encouraging interest by young people during the campaign that can be carried over
This study utilized a quasi-experimental design to examine the responses to the 2004
Presidential candidates by young people. As in any other experimental study the current study
was able to control and manipulate certain variables but may lack the spontaneity of naturally
While conducting this experiment during the course of the actual campaign is one
strength of this study, the campaign itself presents a potential limitation. Given that the
citizens had made a decision early in this campaign season regarding for whom to vote.
Additionally, President Bush was running for re-election, which meant that the public entered the
general election with four years of exposure to him as President, a timeframe in which voters had
Communication Modes and Candidate Image 25
undoubtedly developed their perception of his image as a presidential candidate and as president.
that allowed the opportunity for constructing an image of him in their minds. Therefore,
research comparing multiple mode message effects could be particularly insightful when
conducted during open-seat Presidential elections. Future research should also focus on lesser-
known candidates, such as is often the case in lower-level races, to determine how exposure to
Conclusions
The purpose of this study was to examine the responses of young people to political
candidates. Specifically, the study was designed to examine the influence of three modes of
debates, and campaign Web sites—on ratings of candidate image. This study contributes to the
important variables that might contribute to decisions for whom to vote. The experiment
revealed that communication mode does have an effect on how young people view the
candidates’ image. More importantly, the results indicated that this variable is influenced
differently by the modes with image being rated highest in the campaign debates. It seems as
though the modes of campaign communication in which young people must actively seek
are the ones that are more appealing to this young generation who grew up with computers,
video games, reality television shows, and the Internet as forms of entertainment. As candidates
look to new ways to appeal to this large, but underrepresented voting group, the results seem to
indicate that they can best do this through new or active forms of campaign communication.
Communication Modes and Candidate Image 26
Ultimately, young people have been an enigma in U.S. electoral politics. As a group,
young voters have not yet been mobilized in large numbers in an election, but significant strides
have been made over the past few election cycles. Young voters will continue to be a group
is essential to understand how they respond to and connect with political candidates since
citizens go to the voting booth to cast a vote for a specific person. If candidates better
understand how young people respond to them, then they will not only serve their self-interest by
appealing to this group, but also serve the public interest by encouraging the political
References
Benoit, W. L., McKinney, M. S., & Stephenson, M. T. (2002). Effects of watching primary
debates in the 2000 U.S. presidential campaign. Journal of Communication, 52, 316-331.
Best, S. J., & Hubbard, C., (1999). Maximizing “minimal effects:” The impact of early primary
Chaffee, S. H., & Choe, S. Y. (1980). Time of decision and media use during the Ford-Carter
Cundy, D. T. (1986). Political commercials and candidate image: The effects can be substantial.
Dailey, W. O., Hinck, E. A., & Hinck, S. S., (2005). Audience perceptions of politeness and
advocacy skills in the 2000 and 2004 presidential debates. Argumentation and Advocacy,
41, 196-210.
on citizens’ involvement in political and civic life. In, L. L. Kaid (Ed.), Handbook of
Associates, Inc.
Friedkin, K. L., & Kenney, P. J. (2004). Do negative messages work? The impact of negativity
Geer, J. G. (1988). The effects of presidential debates on the electorate’s preferences for
Green, S. B., & Salkind, N. J. (2003). Using SPSS for Windows and Macintosh: Analyzing and
Understanding Data (3rd ed.). Upper Saddle River, NJ: Prentice Hall.
Hacker, K. (Ed.) (1995). Candidate Images in Presidential Elections. New York: Praeger.
Hacker, K. L. (Ed.) (2004). Presidential candidate images. Lanham, MD: Rowman & Littlefield.
Jasperson, A. E., & Fan, D. P. (2002). An aggregate examination of the backlash effect in
political advertising: The case of the 1996 U.S. senate race in Minnesota. Journal of
Johnson, T. J. & Kaye, B. K. (2003). A boost or bust for democracy?: How the Web influenced
political attitudes and behaviors in the 1996 and 2000 presidential elections.
Press/Politics, 8, 9-34.
Just, M., Crigler, A., & Wallach, L. (1990). Thirty seconds or thirty minutes: What viewers learn
from spot advertisements and candidate debates. Journal of Communication, 40, 120-133.
Political Communication (pp. 249, 271). Beverly Hills, CA: Sage Publications.
Kaid, L. L. (1982). Paid television advertising and candidate name identification. Campaigns
Kaid, L. L. (1997). Effects of the television spots on images of Dole and Clinton. American
Kaid, L. L. (2002). Political advertising and information seeking: Comparing the exposure view
Kaid, L. L. (2003). Comparing Internet and traditional media: Effects of voters. American
Kaid, L. L. & Chanslor, M. (2004). The effects of political advertising on candidate images. In.
Kaid, L. L., Downs, V. C., & Ragan, R. (1990). Political argumentation and violations of
Kaid, L. L. & Johnston, A. (2001). Political advertising content & effects. In, Videostyle in
Kaid, L. L., Leland, C. M., & Whitney, S. (1992). The impact of televised political ads: Evoking
57, 285-295.
Kaid, L. L. & Postelnicu, M. (2004). Political advertising in the 2004 election: Comparison of
traditional television and Internet messages. American Behavioral Scientist, 49, 265-278.
Kaid, L. L., Singleton, D. L., & Davis, D. (1977). Instant analysis of televised political
Katz. E., & Feldman, J. (1962). The debates in light of research: A survey of surveys. In S. Kraus
(Ed.), The Great Debates (pp. 173-223). Bloomington, IN: Indiana University Press.
Communication Modes and Candidate Image 30
Kern, M., & Just, M. (1995). The focus group method, political advertising, campaign news, and
King, D. C., & Matland, R. E. (2003). Sex and the Grand Old Party: An experimental
Kraus, S., & Davis, D. K. (1981). Political debates. In D. D. Nimmo & K. R. Sanders (Eds.),
Lanoue, D. J., & Schrot, P. R. (1991). The joint press conference: The history, impact, and
Lopez, M. H., Kirby, E., & Sagoff, J. (2005, July). The youth vote 2004. The Center for
McKinney, M. S. (1994). Design and implementation of the focus group study. In D. B. Carlin &
M. S. McKinney (Eds), The 1992 presidential debates in focus (pp. 21-35). New York:
Praeger.
McKinney, M. S., & Carlin, D. B. (2004). Political Campaign Debates. In L. L. Kaid (Ed.),
McKinney, M. S., Dudash, E. A., & Hodgkinson, G. (2003). Viewer reactions to the 2000
the 2000 campaign (pp. 43-58). Lanham, MD: Rowman & Littlefield.
Communication Modes and Candidate Image 31
McKinney, M. S., Kaid, L. L., & Robertson, T. A. (2001). The front-runner, contenders, and
McKinnon, L. M., Tedesco, J. C., & Kaid, L. L. (1993). The third 1992 Presidential debate:
Press.
Nimmo, D., & Savage, R. L. (1976). Candidates and their images: Concepts, methods, and
Paletz, D. L. (2002). Political Socialization. In, The media in American politics: Contents and
Patterson, T. E. (2005, February 2). Young voters and the 2004 election. Retrieved March 1,
Payne, J. G., Golden, J. L., Marlier, J., & Ratzan, S. C. (1989). Perceptions of the 1988
Pew Research Center for the People and the Press (2004, September 30). Young people more
Pew Research Center for the People and the Press (2005, March). The Internet and Campaign
Pfau, M., Cho, J., & Chong, K. (2001). Communication forms in U.S. Presidential campaigns:
88-105.
Powell, L., & Cowart, J. (2003). Political campaign communication: Inside and out. Boston:
Racine Group (2002). White paper on televised political campaign debates. Argumentation and
Rahn, W. M. (1993). The role of partisan stereotypes in information processing about political
Sanders, K. R., & Pace, T. J. (1977). The influence of speech communication on the image of a
Self, J. W. (2005). The first debate over the debates: How Kennedy and Nixon negotiated the
Sundar, S. S., Kalyanaraman, S., Brown, J. (2003). Explicating web site interactivity: Impression
Tedesco, J. C. (2006). Web interactivity and young adult political efficacy. In, A. P. Williams
& J. C. Tedesco (Eds.), The Internet Election: Perspectives on the web in campaign 2004
Tedesco, J. C., & Kaid, L. L. (2003). Style and effects of the Bush and Gore spots. In L. L. Kaid,
Communication in the 2000 campaigns (pp. 5-16). Lanham, MD: Rowman & Littlefield.
Communication Modes and Candidate Image 33
Tedesco, J. C. (2004). Changing the channel: Use of the Internet for communicating about
Footnotes
1. The television advertising stimuli to which participants were exposed was coordinated
with the UVote 2004 national research team project and the compilation of
being aired at the time of data collection. This study used two different UVote 2004 ad
compilations—one from late September 2004 and another from late October 2004.
Communication Modes and Candidate Image 35
Table 1
Analysis of Variance for Bush Image and Kerry Image by Participant Party and Communication Mode
Source df Error df F η2 p
Multivariate Test
Party* 4 820 51.47 .20 .000
Communication Mode* 4 820 14.85 .07 .000
Party x Mode* 8 820 17.65 .15 .000
Univariate Tests Party
Bush Image* 2 410 83.90 .29 .000
Kerry Image* 2 410 50.99 .20 .000
Univariate Tests Communication Mode
Bush Image* 2 410 20.47 .09 .000
Kerry Image* 2 410 10.64 .05 .000
Univariate Tests Party X Mode
Bush Image* 4 410 21.69 .18 .000
Kerry Image* 4 410 20.50 .17 .000
*significant at the p<.001 level
Table 2
Means and Standard Deviations for Bush Image and Kerry Image by Participant Communication Mode
and Party
Communication Mode
Ads Debates Web site Total
M SD M SD M SD M SD
Bush Image
Democrat 3.942 0.62 3.671 0.97 3.711 0.91 3.83 0.78
Republican 4.63a1 0.45 6.17b2 1.13 4.33a2 0.53 4.88 0.95
Independent 4.08a2 0.61 4.66b3 1.17 4.52b2 0.49 4.33 0.69
Total 4.26 0.63 4.90 1.60 4.20 0.76
Kerry Image
Democrat 4.51a1 0.40 5.66b1 0.84 4.66a1 0.83 4.75 0.76
Republican 4.07b2 0.58 3.81b2 0.94 4.60a 0.41 4.10 0.68
Independent 4.332 0.41 4.56a3 0.57 4.16a2 0.56 4.29 0.51
Total 4.28 0.52 4.67 1.16 4.42 0.69
Note: Means in each row with different lettered superscripts differ significantly at p<.008.
Means in each column with different numbered superscripts differ significantly at p<.008.
Scale: Candidate image was measured on a 12-item, 7-point semantic differential scale with higher
values representing higher candidate image ratings