This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
         If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line, 
send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").

(Feb 2006)
Thank you.
Elliott Parker
====================================================================

No Laughing Matter: Negative Attribute Agenda Setting on Late Night Television


By
Amy Zerba and Tania Cantrell


School of Journalism
University of Texas at Austin
P.O. Box 8238, Austin, TX 78713
  (512) 699-2347
[log in to unmask] and [log in to unmask]




Submitted to the Entertainment Studies Interest Group of the AEJMC 
National Convention,
San Antonio, Texas, August 2005


Abstract

No Laughing Matter: Negative Attribute Agenda Setting on Late Night Television

[Key Words: Late Night Television, Agenda Setting, Daily Show, Election]

This attribute agenda setting study explores the negative attributes 
of Bush and Kerry jokes on Leno, Letterman, Conan and The Daily Show 
and the negative attributes stated by watchers and non-watchers of 
the shows during the 2004 U.S. presidential election. Findings from 
this content analysis, Web survey and experiment study show attribute 
agenda setting effects for Bush; increased campaign interest with 
show(s) exposure; and the significant influence of party affiliation 
with respondents' negative descriptions and the jokes' negative attributes.



3



No Laughing Matter: Negative Attribute Agenda Setting on Late Night Television
Introduction
What do former President Bill Clinton playing the saxophone, Sen. 
John Kerry riding a Harley-Davidson, and President George Bush 
reading a Top 10 White House joke list have in common? These 
presidential candidates performed these routines on late-night 
television in an attempt to reach younger voters. Sixty-one percent 
of people aged 18-29[1] regularly or sometimes learn about the 
campaign or get news about the candidates from non-traditional news 
sources, such as comedy shows like Saturday Night Live or The Daily 
Show (Pew Research Center, 2004). Studying young voters' perceptions 
of candidate attributes and opinions of the candidates grows more 
important as non-traditional news, or what has been termed "new 
news," increases in popularity. Prior to the 2004 presidential 
election, young voter turnout had declined 13 percentage points among 
young citizens during the 1972 to 2000 presidential election years, 
excluding a 1992 presidential election voting spike (Center for 
Information & Research on Civic Learning & Engagement, 2002 & 2003). 
In the 2004 election, young voter turnout amplified, with at least 
20.9 million Americans under age 30 -- an increase of 4.6 million 
from 2000 – voting (CIRCLE, 2004).
  Specific to the 2004 presidential election, the two candidates 
targeted the young voter through their appearances on new news 
programs, such as The Daily Show, Letterman, Leno, Live with Regis 
and Kelly, Dr. Phil and the BET Network, and The O'Reilly Factor. 
This study explores the relationship between the young voters' 
perceptions of the 2004 candidates and late-night television jokes. A 
content analysis of negative jokes from four shows (the Late Show 
with David Letterman, The Tonight Show with Jay Leno, Late Night with 
Conan O'Brien Show, and The Daily Show), a Web survey of college-aged 
students and an experiment of non-late-night TV watchers were 
conducted prior to the 2004 presidential election to tap into the 
influence of non-traditional news on young adults' negative 
perceptions of the candidates. Attribute agenda setting theory founds 
this study. Why is studying agenda setting effects among college-aged 
students so important?  People get cues about what is important via 
personal experience, interpersonal communication and mass 
communication (Lasorsa & Wanta, 1990), with the mass medium chosen 
varying from person to person, depending on age, gender, education, etc.
Literature Review
Attribute Agenda-Setting Theory, Candidate Image and Need for Orientation
Attribute agenda setting theory extends object salience, or 
first-level agenda setting, which considers the issues the public 
thinks about (McCombs & Shaw, 1972), to attribute salience, or 
second-level agenda setting, which concerns how the public thinks 
about issues. Kim and McCombs (2004) describe the agenda of 
attributes, noting that issues, or objects, in the news have many 
attributes that "fill out the picture of that object" (pp. 2-3). Each 
issue can have an agenda of attributes, which can be ranked according 
to relative salience of those attributes.  Second level 
agenda-setting effects are particularly important to understanding 
presidential candidates' attributes.  Campaign coverage of 20 U.S. 
newspapers during the last four weeks of the 1968 presidential 
campaign was found to be similar; coverage mostly focused on the 
candidates' personal qualities, primarily their personality traits, 
from trustworthiness to the ability to inspire confidence, 
compassion, ability to empathize with the needs of people, and the 
ability to show restraint in his style of operations (Graber, 1972). 
Also, candidate attributes, which can subsume issue and persona 
perceptions that voters find important, vary from election to 
election (Hacker, Zakahi, Giles & McQuitty, 2000).
The public has been found to link factual information and personal 
characteristics to candidates based on media coverage, even when that 
coverage has been purposively falsified in an experimental lab 
setting (Kiousis, Bantimaroudis & Ban, 1999).  However, a study by 
Golan & Wanta (2001) showed the public is less likely to link 
positive/negative evaluations of these same issues and 
characteristics to candidates in proportion to the positive/negative 
coverage the candidates received in the news media. When viewers 
chose television news in a longitudinal study spanning seven years, 
the findings showed that agenda setting effects are unidirectional; 
television news, particularly lead stories and when the president is 
featured, influences public concern, not vice versa (Behr & Iyengar, 
1985).  Additionally, having at least some college education as well 
as prior awareness of news topics through print media exposure 
contributes to increases in television-mediated agenda-setting 
effects (Hill, 1985).  Further, exposure quality, rather than 
quantity, leads to greater agenda-setting effects (Hill, 1985, p. 348).
People receive and/or seek out information in accordance with their 
need to know, or their need for orientation. The degree to which 
attribute agenda-setting effects occur depends largely on the 
public's need for orientation (Weaver, 1991), which is determined by 
one's degree of interest in the message content; uncertainty about 
the subject of the message; and effort required to attend to the 
message (including the perceived likelihood that a reliable source of 
information is available) (Weaver, 1991). Political campaign 
information, particularly synthetic presidential candidate images 
(Nimmo & Savage, 1976), is linked to this need for orientation 
because of indirect source availability and unobtrusiveness. To 
orient themselves, young voters turn to various media, including 
traditional and non-traditional forms.  The researchers hypothesize:
H1. Viewers with a greater need for orientation about the 
presidential candidates are more likely to be exposed to late-night television.

Non-Traditional News:  Presidential Debates and Political Humor
Learning about candidates and their images and personalities as well 
as increasing factual political knowledge is becoming popular through 
soft news (Baum, 2003). Two types of soft news, which has been termed 
"infotainment" for its supposed decrease in emphasis on political 
participation (Brants, 1998), that are particularly important to 
young voters are the presidential debates and political humor. A 
study of the media during the 1992 election showed that the 
interview/talk format, which included the debates, added to voter 
learning about the candidates (Chaffee, Zhan & Leshner, 1994). 
According to a USA Today survey, 36 percent of people considered 
televised debates to be the most valuable information source in their 
decision-making (Carlin, 2000). Reasons for this could include the 
reasoning, clearer statements and greater specificity in issues, 
along with more accurate and precise discourse within the 
give-and-take exchanges that audiences witness between the candidates 
(An & Pfau, 2004). Debate watching has been linked to participation 
in the campaign and vote likelihood (McLeod et al., 1996). Also, 
attention to such talk shows by those with a higher education has 
been shown to be positively associated with actual campaign knowledge 
(Hollander, 1995).
  Non-traditional news sources in the 1992 presidential campaign were 
found to be more influential than traditional news media on 
prospective voters' perceptions of the candidates' competence and 
relational traits, the latter being intimate characteristics such as 
personable/unpersonable, warm/cold, sincere/insincere, to name a few 
(Pfau & Eveland, 1996). The researchers hypothesized this may be 
because non-traditional news is personal by nature.
Evidence negating non-traditional media's influence also exists. 
Nontraditional media in the form of TV talk shows and morning network 
shows also have been shown not to contribute significantly to greater 
knowledge of candidate issue positions, greater intention to vote or 
higher levels of interest in the 1992 election campaign (Weaver & 
Drew, 1995). However, exposure to non-traditional news may have an 
influence on individuals' attitudes, including even their voting 
behavior – particularly among those not intrinsically interested in 
politics or foreign affairs – without necessarily having a comparable 
effect on their long-term factual knowledge about specific political 
issues or events (Baum, 2003). Because evidence about non-traditional 
media's influence is mixed, this present study specifically examines 
campaign interest and exposure to late-night television, predicting:
H2. The greater the amount of campaign interest, the greater the 
exposure to late-night television.

Attention to night time television shows, which are a form of 
political humor, has proven an important source in candidate-issue 
knowledge (Chaffee, Zhao & Leshner, 1994), even though some argue 
that there is no new way to present information; use of new media is 
just "more information about the same things" (Kerbel, 1995).  But 
political humor has many social functions. From the point of view of 
politicians, humor can be used to define political concepts, to 
disarm critics, to establish a position or make a point and to 
relieve tension.  From the point of view of political critics, humor 
can be used to expose chauvinism, ineptitude, oppression, and even 
pretentiousness (Nilsen, 1990). From viewers' perspective, political 
humor plays a large role in not only how the public obtains political 
information, but also what kind of information the public 
obtains.  "Comic sages" (Schutz, 1977) such as Jay Leno, David 
Letterman, Jon Stewart, and Conan O'Brien rely on a basic 
understanding of current events to joke with their audiences about 
"the candidates' verbal gaffes, their off-beat proposals, and their 
character flaws …" (Amundson & Lichter, 1998).
Young voters seek not only some of their news from late-night 
television (Schechter, 2003), but a significant amount of their 
political information, especially from The Daily Show during the 2004 
election. For example, reports on surveys from the 1996 campaign 
showed that 40 percent of Americans under 30 learned something about 
presidential politics from late-night television (Bennet, 1996), 
while reports on surveys during the 2000 campaign said that almost 
half (47 percent) of Americans under the age of 30 said they get at 
least some of their news about presidential campaigns from late-night 
shows (Halonen, 2000).  These numbers remained at least as high in 
the 2004 presidential election and demonstrated a significant 
difference in political knowledge between those who watch late-night 
television and those who do not. Polling conducted over a two-month 
period in 2004 showed that The Daily Show watchers answered more 
items correctly on a six-item political knowledge test (3.59 items) 
compared to Leno watchers (2.95), Letterman viewers (2.91) and 
non-watchers of late-night comedy (2.62) (Young, 2004).
Why do The Daily Show viewers differ so greatly from other late-night 
television show viewers? This study, which treats The Daily Show and 
the other "Big 3" (Leno, Letterman, O'Brien) late-night television 
shows separately, considers this, examining the attributes of each 
show and projecting:
H3. The negative attribute agendas of Bush and Kerry on The Daily 
Show, Letterman, Leno and Conan will be similar.

	According to the Pew Research Center's national study, 27 percent of 
young people under 30 who say they regularly or sometimes learn 
campaign news from comedy and late-night shows indicate they also 
learn new things about the candidates and campaigns that they didn't 
know before (2004). In other words, the audiences rely on the "sages" 
to gain perspective into the candidates' personalities, to see that 
"Hey, they're people, too" (Wizda, 2000). And young voters, primarily 
those who have a higher education level, tend to want to learn about 
presidential candidates' personal attributes (Glass, 2001; Hollander, 
1995). This current study specifically examines exposure to 
late-night television and attribute agenda setting, stipulating
H4. The strength of the relationship between the shows' negative 
presidential candidate attributes and viewers' negative attributes of 
presidential candidates is contingent upon the amount of exposure to 
late night television.

	In controlling for exposure in an experimental setting, this study 
further hypothesizes
H5. Attribute agenda-setting effects will increase when never or 
seldom viewers of late- night television are exposed to negative 
presidential candidate jokes.

Political Humor and Partisanship
 From the late-night shows, monologues and jokes influence culturally 
savvy voters' perceptions of political candidates (Weinraub, 2000). 
Rob Burnett, executive producer of the Letterman Show, acknowledged 
that his show and others had an impact, but said, the shows only take 
"what people understand about a candidate and make fun of it. We 
don't really invent anything." Said Leno:
… Let me explain what happens with a joke. You have to repeat 
information they already know. For example, if it made the headlines 
today, you go, 'Thank you for coming to the show tonight. As you 
know, you all signed your loyalty oaths.' If they go, 'What's he 
talking about?' then I'm still talking, and they're still explaining 
the joke to each other. (Finke, 2004)

He also stipulated, "You don't change anybody's mind with comedy… You 
just reinforce what they already believe" (Weinraub, 2000).
In a study that content analyzed first-run episodes of the Tonight 
Show, Late Show, Late Night and Politically Incorrect over a span of 
five years (1996-2000), Leno was found to be the most prolific 
political jokester (Niven, Lichter & Amundson, 2003). In the past 
five years, late night television jokes have tended to be non-issue 
oriented, targeted more at Democrats, and directed more at the 
executive branch, such as the president, major presidential 
candidates and those surrounding the candidates such as family, the 
cabinet, advisors and accusers (Niven et al.).
Studies that indicate partisanship influences the image that voters 
have of candidates are mixed (Sigel, 1964; Campaign for Young Voters, 
2004). The image that Democrats held of their ideal president in the 
1960 election correlated positively with their image of Kennedy 
(Sigel, 1964). The Republicans' ideal image correlated positively 
with their image of Nixon. The Democratic view of Nixon did not 
relate significantly to their idealized image of the president. The 
same was true for Republicans' image of Kennedy. No significant 
correlations were obtained when the independents' ideal image was 
compared to that of either candidate. Partisanship is important in 
determining the ideal candidate but not so much in formulating the 
ideal non-preferred candidate image.  Voters, though, do consider 
competence, character and sociability, followed by composure and 
extroversion, as the most important dimensions among political 
candidates (McCroskey & Jenson, 1975).
Questions regarding the connection between voter political 
affiliation and candidate image attributes arise. Specifically, this 
study explores
R1. Does party affiliation affect the relationship between the shows' 
negative attributes of the presidential candidates and regular 
viewers' negative attributes of the candidates?

	To address the research question and each of the five preceding 
hypotheses, a four-pronged method was used to investigate late night 
television's negative attributes of candidates and the public's 
negative attribute agenda. Do attribute agenda setting effects cross 
over to late night television?
Method
The research design of this study merges a traditional agenda-setting 
study (survey and content analysis) with a laboratory experiment.  It 
consisted of  (a) a content analysis of a constructed week (Riffe, 
Aust, & Lacy, 1993) in October of presidential jokes on four 
late-night television shows: the Big 3, or the Late Show with David 
Letterman, The Tonight Show with Jay Leno, and Late Night with Conan 
O'Brien Show, and The Daily Show; (b) a Web-based survey of randomly 
selected college students conducted two weeks prior to the 2004 U.S. 
presidential election; (c) a content analysis of presidential 
candidate jokes on a 25-minute tape of edited segments from these 
shows spliced together from two weeks of mid-October programs used in 
an experiment; and (d) a paper survey given to college students who 
reported never watching these four late-night television shows or 
watching no more than three of them one day a week.
Because most jokes on late night television about presidential 
candidates are negative, only the negative attributes were used in 
this study. See Table 4.1. Too few jokes on late night television 
described the candidates in positive or neutral ways. For this 
reason, this study specifically examined only negative attributes 
about the candidates on late night television and by viewers and 
non-viewers of these four shows (Letterman, Leno, Conan and The Daily Show).
The study was conducted during a time when the United States was at 
war with Iraq; Bush and Kerry had finished their third debate four 
days prior to the study's survey part; some prospective voters did 
not know where the Democratic nominee Sen. John Kerry stood on key 
issues; and Bush remained ahead of Kerry in the CNN/USA Today/Gallup 
poll by a 52 to 44 margin (Gallup Organization, 2004).
The Web-based Survey and Sample. A systematic random sample with a 
random start point      was used to select 30 percent of currently 
registered students at a large southwestern university who had agreed 
to list their email address publicly with the university. The sample 
of 15,077 students was emailed a request to participate in a 
30-question Web-based survey. A link in the email pointed students to 
the online survey. Respondents were given one week (Oct. 17-24, 2004) 
to take the survey. The invitation email to take part in the study 
mentioned an incentive (a random drawing for one $50 gift certificate 
to the university's bookstore) to participate and assured 
confidentiality. A reminder email was sent on the fourth day. Survey 
results were automatically sent to a database, where open-ended 
answers were categorized. The results were then rank ordered and 
analyzed in SPSS. The response rate was 10 percent (N=1,556). Because 
this study controlled for two variables – age and U.S. citizenship – 
1,250 students who took the survey qualified. This final total -- 
1,250 participants -- included only those ages 18 to 29, similar to 
the Pew Research Center studies, and those who were U.S. citizens.
Presidential Candidate Attributes. To measure the public's attribute 
agenda for each presidential candidate, the Web-based survey asked 
students the attribute agenda setting question, "Suppose that one of 
your friends has been away a long time and knows nothing about the 
candidates for president.  What would you tell your friend about 
[George W. Bush or John Kerry]?" (Kihan & McCombs, 2004). The first 
100 open-ended answers for each candidate were then categorized and 
collapsed into 10 categories by two graduate students. The 
categories, which were based on a previous attribute agenda setting 
presidential campaign study (McCombs, Lopez-Escobar & Pablo Llamas, 
2000) but with additional categories were: Ideology; Biography, 
Record and Leadership; Integrity and Values; Personality and Image; 
War, Terrorism and Allies; Public Speaking/Debates; Flip Flopping and 
Decision-making; Intelligence; and Other. These same categories were 
used for the content analysis of the presidential candidate jokes on 
the four late-night television shows as well. Further, the tone of 
the comment or joke, depending on its assertion valence, was 
categorized into positive, neutral or negative (Kihan & McCombs, 
2004). The intercoder reliability coefficient was .90 using Holsti's 
method (Neuendorf, 2002). A second test for reliability midway 
through the coding stage produced a .85 intercoder reliability 
coefficient using Holsti's method.
The nine categories were defined as follows:
1.) Ideology/Issues was characterized as comments made about where a 
candidate stood on specific issues, his party platform, 
liberal/moderate/conservative and his vision/goals.
2.) Biography, Record and Leadership incorporated references made 
about the candidates' past record either in office or not in office. 
For example, one student wrote "[Kerry] is a democratic senator from 
Boston who is married to the Heinz Ketchup lady." All of these 
references (except "democratic," which is Ideology) fall under 
Biography and Record. Again, for this study the researchers only used 
the first attribute listed. So, this example would fall under the 
Ideology category, because "democratic" was the first attribute 
listed. Another student wrote, "George Bush has done a good job in 
leading this nation through the difficult times that we have faced 
since Sept. 11 …" This was coded as Bush positive Record.
3.) Integrity and Values touched on the candidates' morals, 
especially truth. For example, the response, "He is a liar who does 
not know what he is doing" was coded as Bush negative Integrity.
4.) Personality and Image incorporated any inferences made about a 
candidate's overall image. For example, "typical politician," "evil," 
"close-minded," "stubborn." Mostly adjectives about their image or 
personality fell into this category.
5.) War encompassed anything that dealt with the war in Iraq, 
terrorism or allies. If war was mentioned, it trumped all other 
categories. For example, the statement "He lied to us about the war 
in Iraq" fell into War, not Integrity and Values.
6.) Public Speaking/Debates included references to speaking in 
general. Because this study took place four days after the third 
presidential debate, many references were made to the candidates' 
speaking ability; hence, this category surfaced.
7.) Flip Flopping/Decision-making incorporated references made about 
changing positions. Examples include "wishy washy," "flip flopper," 
"will follow through with what he says" and "is willing to make tough 
decisions."
8.) Intelligence encompassed any descriptive word or phrase to 
describe the candidates' intelligence, such as "stupid," "moron," 
"idiot," "intelligent" and "smart."
9.) Other referred to references that did not fit into any of the 
above categories, such as "He is the lesser of two evils," "vote for 
him," or "I don't like him."
Because humor is a difficult construct to content analyze, a system 
of operationalizations was developed, as described below.
Content Analysis.  Comments said about the two candidates were 
content analyzed using a count system for the description and tone of 
the comment. To illustrate, if a respondent replied with several 
comments about a candidate, only the first description was coded. The 
following is a sample of open-ended answers, along with the coded 
category, for the question "If you had a friend who knew nothing 
about John Kerry, how would you describe Kerry to your friend?"
War hero from Vietnam, Massachusetts's senator … [positive Record]

He has some faults, but overall will do a much better job than Bush 
[negative Image]

He supports stem cell research [neutral Ideology]

Vote for him [positive Other]

He is very indecisive and cannot stick to one story [negative Flip Flopping]

Liar but a great debater [negative Integrity and Values]
Late Night Shows' Presidential Attributes. Jokes about Bush and Kerry 
on the Late Show with David Letterman, The Tonight Show with Jay 
Leno, Late Night with Conan O'Brien Show and The Daily Show were 
content analyzed during a constructed week in October and Nov. 1, the 
night prior to the election. The dates were as follows: Tuesday, Oct. 
5; Wednesday, Oct. 13; Thursday, Oct. 21; Friday, Oct. 29; and 
Monday, Nov. 1, 2004. Because of technical difficulties (a power 
outage), only the monologue of the Oct. 5, 2004 Conan O'Brien Show 
was content analyzed. The monologue is typically when most jokes 
about the presidential candidates take place (Weinraub, 2000), with 
the exception of The Daily Show, which is one continuous monologue 
with Jon Stewart sprinkled with skits and interviews. Aside from the 
Oct. 5, 2004 Conan monologue-only episode, each show in its entirety 
was content analyzed for mentions of either candidate. Interviews 
with guests, however, were not coded even if jokes about Bush or 
Kerry were mentioned. In addition, only verbal jokes – not facial 
gestures or sounds, such as applause -- were counted and coded. The 
tone of the joke – positive, negative and neutral – also was coded.
Jokes often contained a lead-in, the joke, and the punch line. Each 
part was coded. Lead-ins without any humor were coded as neutral. If 
the skit made fun of the candidate, it was coded accordingly for each 
individual remark about the candidate, followed by the punch line, if 
there was one. The "punch line" refers to the host making fun of what 
the audience just saw. For example, David Letterman introduces a joke 
by saying "And now, ladies and gentleman, a message from John Kerry," 
and a flag appears with the words: "A Message from John Kerry" [coded 
as neutral Other].  A 15-second edited montage of clips appears from 
Kerry speaking at various venues. Kerry is shown saying, "Let me tell 
you straight up, I'm wishy-washy [coded as negative Flip Flopping], I 
change my mind [coded the same], I'm a flip flopper [coded the same], 
I can't be president [coded as negative Image]." The scene then cuts 
to the flag with the same words "A Message from John Kerry" to end 
the joke, but not coded because it's not necessarily a punch line.
Experimental Design. To examine late-night television effects more 
closely, a group of 56 students volunteered to take part in a second, 
smaller portion of the original Web-based study during Oct. 26-28, 
2004. A question on the original Web-based survey requested 
volunteers for this second part of the study with the following 
statement, "We would like to know more about your TV viewing habits 
and opinions about the election. If you supply your email address 
here, we may contact you in the near future for the second part of 
this study that also has an incentive." The incentive for the 
experiment portion was free pizza for one hour of their time.
Of those students who supplied usable email address (N=1,026), only 
those students who were ages 18 to 29, U.S. citizens and never or 
seldom watched late night television or watched the show(s) one day a 
week were considered (N=345). To increase the sample pool, students 
could watch up to three of the shows per week but only one day a week 
for each, no more[2]. Those who qualified under these conditions 
(N=345) were divided into two groups by an alternating method. One 
person was assigned to the control group; the next person was 
assigned to the experiment; next person to the control group, etc. A 
total of 56 participants were divided into two groups. The 
experimental group watched a montage of edited clips from all four 
shows with non-presidential jokes in the mix. The control group 
watched a third-season episode of Friends, with no references to 
politics or the election.
The two groups – the experimental group and the control group – were 
sent the same email inviting them to participate in the three-night 
study. Each night consisted of three time slots. Similar to the 
alternating placement of participants in the two groups, the one-hour 
time slots alternated between a control and experiment group 
[3].  So, on all three nights, a control group and an experiment 
group participated in the study. After the television viewing, both 
groups of students then were given the same Web-based survey, with a 
few added questions and in paper-and-pencil form. After surveys were 
collected, the participants were debriefed and promised confidentiality.
Measures
	Late night television exposure. Students were asked: "In a typical 
week, how many days do you watch [the Daily Show with Jon Stewart, 
The Late Night Show with David Letterman, The Tonight Show with Jay 
Leno and Late Night with Conan O'Brien Show]?" The six choices for 
Letterman and Leno were: No days; one day; two days; three days; four 
days and five days. The choices for The Daily Show differed from the 
first two, for this show only runs Monday through Thursday, with a 
repeat every Friday. The five choices provided for The Daily Show 
were: No days; one day; two days; three days; and four or five days. 
For consistency with The Daily Show, responses for "four days" and 
"five days" were collapsed into one answer of "four or five days" for 
Letterman, Leno and Conan.
Responses to exposure to the four late-night television shows were 
then factor analyzed. Using a rotated varimax factor analysis, 
Letterman, Leno and Conan loaded similarly, and The Daily Show loaded 
separately. Therefore, this study examined Letterman, Leno and Conan 
as a newly created "Big 3" late night television variable.  The 
"Daily Show" was treated as a separate late- night television 
variable. When testing exposure, students who answered that they 
watched one or more of the Big 3 shows (Letterman, Leno and Conan) 
one or more days a week were counted once as a late night show 
watcher. Those who answered no days to all three shows were 
considered non-watchers of late night television. The Daily Show 
exposure had the same two categories of exposure – non-watchers and watchers.
	Late night television exposure for experiment study. The initial 
exposure to late night television was controlled for in the 
experiment by recruiting only those students who reported never or 
seldom watching these four late-night television shows or watching up 
to three of them at the most one day a week. The exposure variable, 
however, for the experiment was a stimulus: a 30-minute tape blending 
presidential candidate and non-political jokes from the four 
late-night television shows airing Oct. 1-15, 2004. The negative 
candidates jokes were split evenly. Half of the students (N=28) in 
the experiment watched the tape. The control group (N=27) was not 
exposed to this stimulus. The presidential candidate jokes on the 
edited tape were content analyzed, and the same nine categories were 
used. The frequencies of the negative attribute jokes within these 
categories were tabulated and compared to the salience of attributes 
for each candidate by the participants in the study. Spearman's rho 
was used to determine the rank order similarities of the attributes 
between the edited tape (stimulus) and the participants' presidential 
candidate attributes.
	Public's attributes of presidential candidates on Web survey. 
Students were asked, "Suppose that one of your friends has been away 
a long time and knows nothing about the candidates for president. 
What would you tell your friend about George W. Bush?" Then asked, 
"Suppose that one of your friends knows nothing about John Kerry. 
What would you tell your friend about Kerry?" The open-ended 
responses for this dependent variable were collapsed into: 
Ideology/Issue/Vision; Biography; Record and Leadership; Integrity 
and Values; Personality and Image; War, Terrorism and Allies; Debates 
and Public Speaking; Flip Flopping and Decisiveness; Intelligence; 
and Other. The frequency of the negative-only responses for each of 
the nine categories was tabulated separately for Kerry and for Bush. 
These frequencies were then compared to the frequencies of the 
late-night television attributes of both candidates. A Spearman's rho 
was calculated to determine rank order similarities between the 
late-night television shows' salience of attributes of the 
presidential candidates and students' salience of attributes of Bush and Kerry.
	 Late night television negative attributes of the presidential 
candidates. The four late-night shows were content analyzed and jokes 
were coded using the nine categories stated earlier. Only the 
negative jokes were compared, for the number of positive and neutral 
jokes in the constructed week were too few to compare, with many 
categories missing. The frequencies of each category of negative 
attributes for each candidate were tabulated in SPSS, and a 
Spearman's rho was used to calculate the rank order similarities with 
the public's negative attribute agenda for each candidate.
	Experiment participants' attributes of presidential candidates. 
Again, students were asked the same "If you had a friend …" attribute 
agenda setting question for both Bush and Kerry on the paper survey. 
Similar to the Web survey, the open-ended responses for this 
dependent variable were collapsed into the same nine categories. The 
negative-only open-ended responses for those exposed to the stimulus 
(edited montage of presidential candidate jokes from the four shows) 
were compared to the negative only presidential candidate jokes on 
the edited tape. The same test was conducted for the control group. 
Frequencies were tabulated for the nine categories for each group and 
for each candidate, and a Spearman's rho was calculated to determine 
rank order similarities. The correlations between the two groups were compared.
Exposure to candidate attributes on experimental video. The 
experiment pulled together jokes from each of the four shows to 
examine possible attribute agenda setting effects. The negative 
attributes of the jokes were closely balanced, with Bush having 32 
and Kerry having 31. The jokes were coded into the nine categories. 
The frequencies of jokes for each category were tabulated, and a 
Spearman's rho was calculated to determine rank order similarities 
between the negative attributes of the candidates separately on the 
video clips and the negative attributes of each candidate by those 
exposed to the video.
  	Party affiliation. Students were asked with which political party 
they identify themselves. The choices were: Republican, Democrat, 
Independent and Other. For purposes of this study, only the 
Republican, Democrat and Independent responses were used. Students' 
presidential candidate attributes for Bush and Kerry were examined by 
controlling their party affiliation and late night television shows' 
attributes of both Bush and Kerry. These frequencies were tabulated 
and then compared to the frequencies of the late-night television's 
attributes of both candidates. A Spearman's rho was calculated to 
determine rank order similarities between the late-night television 
shows' salience of attributes of the presidential candidates and 
students' salience of attributes of Bush and Kerry.
Need for orientation. Students were asked the question, "Concerning 
the two main candidates for president – George W. Bush and John Kerry 
– do you feel that you: Know enough about them; Need some additional 
information; or Need a great deal of information?" For an even 
distribution, responses were collapsed into two categories  -- "No 
need for orientation" and "Need for orientation," the latter included 
both "Need some additional information" and "Need a great deal of 
information" responses. A chi square test was used to measure the 
relationship between students' exposure to late night television and 
their need for orientation. A separate chi square test was run for 
watchers and non-watchers of The Daily Show and their need for orientation.
	Campaign interest. This dependent variable was measured by the 
number of debates a student watched. The question asked was, "How 
many presidential debates have you watched this month?" The four 
response choices (none, one debate, two debates and three debates) 
were collapsed into three categories – low political activity (none); 
medium political activity (one to two debates); and high political 
activity (three debates). A chi square test was used to measure the 
relationship between students' exposure to late night television and 
their political activity. A separate chi square test was run for 
watchers and non-watchers of The Daily Show and their political 
activity using the same three levels.
Study Weaknesses
Surveying college students has its strengths and weaknesses. Notable 
weaknesses include that these students are all college-educated, 
which represents a smaller subset of the national population. The 
results cannot be generalized to the general population, but only 
this group of students at this university. However, three fourths of 
youth voters who voted in the 2004 election were college students, 
making this sample type an influential voting population (Center for 
Information & Research on Civic Learning & Engagement, 2003).  Also, 
the study had a low 10 percent response rate[4].
Controlling for coder political bias needs to be addressed. The two 
coders recognized their political biases – one was a Democratic coder 
and the other a Republican – and took this into consideration when 
coding. For difficult open-ended responses or jokes that leaned 
positively and negatively for more than one category depending on the 
political perspective, the two researchers collaborated to decide an 
appropriate category and tone. All of The Daily Show was coded together.
Sample
	Of the 1250 respondents, two-thirds were female. Ages ranged from 18 
to 29 with the average age at 21. Three-fourths of the participants 
were undergraduate students, and 20 percent were graduate students. 
Seventy-one percent of the respondents were white with 
Hispanic/Latino as the largest minority at 10 percent. Asian 
Americans represented 9 percent, followed by blacks at 3 percent. 
Ninety-five percent of the respondents reported being registered to 
vote in the United States, and 360 of these students voted in the 
2000 U.S. presidential election. This sample is somewhat interested 
in the presidential election, for 88 percent reported that they 
"probably would" or "definitely would" vote in the 2004 U.S. 
presidential election. Also, 91 percent of these students watched one 
or more of the debates (41% watched all three). More Democrats (38%) 
took the survey than Republicans (28%). Independents (26%) made up 
the third largest group followed by other parties listed (5%). The 
other 3 percent didn't know, didn't respond or reported being 
non-partisan. Almost the same number of students who considered 
themselves moderates (39%) took the survey as those who considered 
themselves to be liberal (38%). Self-reported conservatives made up 
19 percent.
Results
About one-third of these participants watch late night television one 
or more days a week with The Daily Show and Conan O'Brien tied as the 
most popular choice (36%), and Leno (31%) and Letterman (28%) 
trailed. These students tend to use more than one source for news 
about the election; three-quarters report obtaining news about the 
election from TV news sources, 66 percent from the Internet, 60 
percent from newspapers, 46 percent from friends, and 51 percent 
receive news from family.	
As stated earlier, because of the overwhelming negative attributes on 
late night television, only the negative attributes of candidates on 
the shows and by viewers and non-viewers were compared. Personality 
and Image (26%) was the top ranked negative attribute of Bush (N=107) 
on all four shows (Letterman, Leno, Conan and The Daily Show), 
followed by a three-way tie for second-place with War, Public 
Speaking/Debates, and Intelligence (15%), and Ideology following 
(10%). (See negative attributes on Table 4.1.) Personality and Image 
(35%) dominated the 91 negative attributes for Kerry on the four 
shows, followed by Speaking/Debates (15%), Flip Flopping/Decisiveness 
(13%), Other (10%) and Ideology and Vision (9%). 	
Personality and Image (17%) ranked as students' top negative 
attribute of Bush (N=572), followed closely by Bio and Record 
(16.4%), Intelligence (16.3%), Ideology (12%) and Other (11.7%). See 
Table 4.2. Personality and Image (36%) also ranked at the top of 
Kerry's negative attributes (N=437) by students, followed by Flip 
Flopping/Decisiveness (25%), Other (15%), Ideology/Vision (13%) and 
Integrity and Values (5%). See Table 4.1.
H1. Viewers with a greater need for orientation about the 
presidential candidates are more likely to be exposed to late night television.

	There is a significant relationship between need for orientation and 
exposure for The Daily Show only, but the correlation is weak (X2 = 
9.60, df=1, p<.001). See Table 4.3. There was no significant 
relationship found between need for orientation and exposure to the 
Big 3 Shows (X2 = .005, df=1, p>.05). See Table 4.4
H2. The greater the amount of campaign interest, the greater the 
exposure to late-night television.

A significant correlation was found between those regular watchers of 
the Big 3 Shows and a medium to high campaign interest (X2 = 30.262, 
df=2, p<.001). See Table 4.5. The same held true for The Daily Show, 
as shown in Table 4.6. The higher the viewers' campaign interest, the 
more exposure to The Daily Show (X2 = 51.82, df=2, p<.001). Low 
campaign interest was equal to low exposure to both The Daily Show 
and the Big 3 Shows.
H3. The negative attribute agendas of Bush and Kerry on The Daily 
Show, Letterman, Leno and Conan will be similar.

	The late night shows did not have similar negative attribute agendas 
for each of the candidates except in one case; Conan O'Brien and 
Letterman had significantly similar negative Kerry attribute agendas 
(p<.01) with a Spearman's rho of +.793.  See Table 4.7 for the 
negative presidential candidate attributes for the shows.
H4. The strength of the relationship between the shows' negative 
presidential candidate attributes and viewers' negative attributes of 
presidential candidates is contingent upon the amount of exposure to 
late-night television.

The salience of negative Bush attributes on the Big 3 Shows 
(Letterman, Leno and Conan) positively parallels that of negative 
Bush attributes by those who watch one or more of the shows at least 
one day a week (p<.05, with a Spearman's rho of +.733). See Table 
4.8. Negative attributes of Bush on the Big 3 Shows and non-viewers' 
negative Bush attributes were not significant. Neither was the 
salience of attributes for Kerry by both viewers and non-viewers.
No evidence of attribute agenda setting effects appeared for The 
Daily Show's negative attributes of Bush or Kerry and those negative 
attributes of the candidates by both viewers and non-viewers, as 
shown below by Table 4.9.
H5. Attribute agenda-setting effects will increase when never or 
seldom viewers of late night television are exposed to negative 
presidential candidate jokes.

The salience of negative Kerry attributes strongly correlates with 
the negative attributes of Kerry by those exposed to late night 
television (p<.01, with a Spearman's rho of +.761). The negative Bush 
attributes by the exposed participants and the late night television 
clips' negative Bush attributes were not significantly similar. See 
Table 4.10. However, in testing internal reliability, the experiment 
group's responses about Kerry after viewing the video have a moderate 
significant relationship to the non-exposed group's responses about 
Kerry (p<.05, with a Spearman's rho of +.609). Also, the experiment 
group's negative attributes of Kerry matched their previous responses 
given on the Web survey (p<.05, with a Spearman's rho of +.625).
R1. Does party affiliation affect the relationship between the shows' 
negative attributes of the presidential candidates and regular 
viewers' negative attributes of the candidates?

The salience of negative Bush attributes on the Big 3 Shows is 
strongly significantly similar to the salience of negative Bush 
attributes by the Republican viewers who watch at least one of the 
Big 3 Shows one or more days a week (p<.01, with a Spearman's rho of 
+.824). See Table 4.11. The same held true for The Daily Show with 
Republican viewers' negative attributes of Bush matching the shows' 
negative attribute agenda of Bush (p<.05, with a Spearman's rho of 
+.660). See Table 4.12.
Evidence of attribute agenda setting also was found between the 
negative Kerry attributes on the Big 3 Shows and the negative 
attribute agenda of Kerry attributes by Democrat viewers who watch 
one or more shows at least once a week (p<.05, with a Spearman's rho 
of +.613).  See Table 4.13.
Table 4.14 shows the lack of a relationship, or no evidence of 
attribute agenda setting effects, between negative Kerry attributes 
on The Daily Show and negative attributes by Democrat or Republican 
watchers of The Daily Show. The negative attribute agenda of the two 
candidates by Independent viewers and the negative attributes on the 
Big 3 Shows or on The Daily Show were not significantly similar.
Discussion

	The most surprising finding of this study is the moderate to strong 
relationship between young voter partisanship, negative candidate 
attributes and exposure to late night television. The negative 
comments that Republican late-night television watchers say about 
their own party's candidate is similar to the negative attributes of 
Bush on late night television. The negative comments that Democrat 
late night television watchers say about their own party's candidate 
– Kerry – is similar to the negative attributes of Kerry on the Big 3 
Shows. Why would a person agree with the negative comments made about 
his or her own party's presidential candidate? This contradicts a 
previous finding that ideal images of candidates are tied to images 
of favored candidates but not for the opposing candidate (Sigel, 
1964), but that study addressed traditional news. Several 
possibilities could explain this.
1. How political humor occurs may answer why late night television or 
the Daily Show viewers see their own party candidate as the shows do 
– negatively. Joke listeners must have some prior knowledge of a 
topic, or in this case person, to understand a political joke. When 
this new knowledge (or joke content) is consistent with prior 
knowledge or builds on it, the humor of it comes out. Two questions 
arise: Is laughing at one's own party candidate easier than finding 
laughable faults in the opposing candidate? And could negative, 
jesting attributes possibly be more memorable than positive ones? 
Political jokes have this "safe softness" attached to them; they do 
not seriously attack the candidates, but they poke fun at them in a 
way that both Republicans and Democrats will laugh. For example, 
David Letterman showed video clips on a few of his shows of Bush 
blinking rapidly while listening during the debates. The joke was set 
up as the "Presidential Blink Count," and both Republicans and 
Democrats could understand and laugh at it.
        2. Repeated exposure of a negative attribute can stick. 
Negative attributes of candidates are more likely to be seen on 
non-traditional news sources than traditional media. Republicans most 
likely have heard repeatedly negative attributes of Bush. They either 
acknowledge the negative descriptions and counter them, feel angry 
toward the Bush administration and agree with the jokes, compare the 
negative attributes with the negative attributes of the other 
candidate (i.e. "he's the lesser of two evils" was commonly used in 
responses) or simply recite what they've heard repeatedly either by 
the media or through discussion with others. In their responses, some 
Republican students acknowledged these negative attributes of Bush, 
but used them as qualifiers, such as "Bush is a terrible speaker, but 
he isn't a dumb man," "He is not the most intelligent person but is a 
good leader and has a great wife," or "That he sticks by what he 
believes in even if he might be wrong." The same can be said about 
Kerry and Democrats. Democrat respondents may have heard or read that 
Kerry had been accused of flip flopping on decisions but until they 
see it as parodied on The Daily Show, Letterman, Leno and Conan, the 
negative attribute is simply what they have read or heard about.
	3. Bush's history in the media is substantially longer than that of 
Kerry's, which could explain the significant relationship between 
Bush and exposure to the Big 3 shows. Because of greater exposure to 
Bush over time via traditional and nontraditional media sources and 
the tendency of media to be cynical to and critical of political 
figures (McCombs, 2004), pre-existing knowledge of Bush and his 
media-portrayed weaknesses may account for these results. No 
significant findings were found between negative Kerry attributes on 
the Big 3 Shows or on The Daily Show and watchers' negative Kerry 
attributes. There was no evidence of attribute agenda setting effects 
between non-viewers of all four shows and the four shows' negative 
candidate attributes.
4. Why no significant effects for non-viewers' negative attributes 
and late night television? It could be that the young voter relies on 
a variety of media sources for campaign news. Three-quarters of the 
students surveyed reported receiving news about the election from TV 
news sources, 66 percent from the Internet, 60 percent from 
newspapers and about half do from family and/or friends. Individuals 
have different media use habits, and the extent of media effects 
depends on people's motivations for using a specific medium or even 
interest in following the campaign (Miller & MacKuen, 1979).
  	Campaign Interest. Campaign interest is linked to exposure to late 
night television. The findings show significant relationships between 
those with a medium or high interest in campaign news and exposure to 
the Big 3 Shows and The Daily Show, which had a stronger correlation. 
The more debates watched, considered a strong predictor campaign 
interest by previous studies, the more exposure to late-night 
television. But a 1976 presidential election survey of adult citizens 
found that among those with a low degree of political activity, 40 
percent watched none of the televised presidential debates compared 
to 6 percent of the most attentive respondents (Miller & MacKuen, 
1979). They concluded that the most attentive are likely to get their 
political news from additional sources, not solely the debates. For 
this study, campaign interest was treated as an antecedent to 
late-night exposure. But could campaign interest be a consequence of 
exposure to late-night television?
Need for orientation. A reverse correlation was found between need 
for orientation about the candidates and exposure to The Daily Show. 
The hypothesis predicted the greater the need for orientation, the 
greater exposure to the shows. Of those who watch The Daily Show, 
more viewers who marked that they "know enough" about the candidates 
watched the show more than those with a greater need for orientation 
about the candidates. Can the need for orientation in the attribute 
agenda setting process be treated the same with non-traditional news 
as it is with traditional news? The entertainment gratification 
sought on late-night television may cloud this need for information 
about a topic, issue or person. Perhaps this is why The Daily Show 
has become so popular with young adults, for Jon Stewart discusses 
what is happening in the news but then pokes fun at it and especially 
how the media have covered it. It is this infotainment that young 
adults may be drawn to in learning campaign news, rather than the 
format of traditional news, which may successfully or not 
successfully strive for objectiveness. Reasons for tuning into 
late-night television need to be explored.
The experiment. The experiment setting found that the negative 
attributes of Kerry given by the exposed group significantly matched 
the negative attributes of the clips shown. However, internal 
reliability checks of rank order between groups – control and 
experiment – and within the group – experiment group's pretest and 
posttest rank orders – negate the finding that the stimulus did 
transfer salience of attributes for Kerry. The researchers recommend 
further experimental testing of late night television's effects on 
attribute agenda setting of politicians and non-politicians. Further, 
manipulating attributes, such as using minor attributes, and the 
tones of jokes (positive and negative), along with controlling for 
other influential factors, such as party affiliation or media 
exposure, are possible ways to explore nontraditional news' influence.
Conclusion
Comedy, "humanizes the [campaign] process"(Wizda, 2000) and, because 
of its -- albeit questionable -- nontraditional role as an 
influential news source, "is no laughing matter" (Wizda). This study 
has shown that the new news is setting a negative agenda; Republicans 
agree with the negative attributes of their party's candidate (Bush), 
and Democrats agree with the negative attributes of Kerry. This could 
be because of the very nature of the shows – humor, and knowing 
candidates well enough to recognize their faults. How much 
information young voters obtain versus bring to nontraditional 
late-night television is still unknown. Perhaps future studies could 
test political knowledge prior to exposure of the shows followed by a 
post-test.
This study's main strength draws on how political humor contributes 
to the media and public attribute agenda setting conundrum. That the 
traditional media agenda influences the public agenda and even a 
policy agenda is an established by the more than 400 – including 
national and international -- agenda setting studies.  New into this 
inquiry is the question, "Do new, or non-traditional, news set 
agendas differently than traditional media? " In addition, what other 
elements mediate the relationship between the late-night television 
and young voters' perceptions of candidates? This study has 
investigated party affiliation and need for orientation as two 
factors, yet more factors may be influential.
Whispering through this study is the complexity of today's 
media-saturated society and audiences' abilities to and motivations 
for selectively exposing and retaining or rejecting the information 
before them.  In so doing, industry implications are raised.  Is 
non-traditional news really an effective medium for candidates to 
reach their audiences?  Or is traditional news more effective? 
Previous studies show mixed findings. Also, with the resulting 
increase in young voter participation, what are the long-term effects 
of differentiated messaging on democracy?  Much work lies ahead.


Tables
Table 4.1:  Tones of Presidential Candidate Jokes on Letterman, Leno, 
Conan and The Daily Show
BUSH
KERRY
Attribute
Positive
Neutral
Negative
Attribute
Positive
Neutral
Negative
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
5  (5)
2 (6)
10 (5)
Ideology / Vision
11 (4.5)
7 (5)
9 (5)
Bio and Record
10 (3)
4 (5)
7 (6.5)
Bio and Record
0 (8)
12 (4)
5 (7)
Integrity and Values
0 (8)
0 (8)
2 (8.5)
Integrity /Values
5 (6)
0 (8)
2 (9)
Personality / Image
14 (2)
6 (4)
26 (1)
Personality / Image
11 (4.5)
5 (6)
35 (1)
War
5 (5)
30 (2)
15 (3)
War
0 (8)
15 (3)
7 (6)
Speaking / Debates
5 (5)
22 (3)
15 (3)
Speaking / Debates
16 (2.5)
27 (2)
15 (2)
Flip Flop / Decisiveness
0 (8)
0 (8)
2 (8.5)
Flip Flop / Decisiveness
16 (2.5)
0 (8)
13 (3)
Intelligence
0 (8)
0 (8)
15 (3)
Intelligence
0 (8)
0 (8)
3 (8)
Other
62 (1)
37 (1)
7 (6.5)
Other
42 (1)
34 (1)
10 (4)
N=
21
54
107
19
41
91


Table 4.2:  Negative Attributes of Presidential Candidates by Web 
Survey Respondents

BUSH
KERRY
%
%
Ideology / Vision
11.9 (4)
13 (4)
Bio and Record
16.4 (2)
4 (6)
Integrity and Values
7 (7)
5 (5)
Personality / Image
17 (1)
36 (1)
War
11 (6)
.7 (8)
Speaking / Debates
6 (8)
2 (7)
Flip Flop / Decisiveness
2 (9)
25 (2)
Intelligence
16.3 (3)
.5 (9)
Other
11.7 (5)
15 (3)
N =
572
437
According to Web survey responses.








Table 4.3:  Need for Orientation about Candidates and Exposure to Daily Show

Non-Viewers
Viewers
Know enough info
54
63
Need more info
46
37
Valid Cases
(800)
(450)
X2 = 9.60, df=1 p<.001

Table 4.4:  Need for Orientation about Candidates and Exposure to Big 3 Shows

Non-Viewers
Viewers
Know enough info
57
57
Need more info
43
43
Valid Cases
(609)
(641)
X2 = .005 df=1 p>.05

Big 3 Shows: Conan, Leno and Letterman

Table 4.5:  Big 3 Shows and Campaign Interest

Low
Campaign Interest (%)
Medium
Campaign Interest  (%)
High
Campaign Interest (%)
Non-Viewers
73
45
48
Viewers
27
55
52
Valid Cases
(117)
(616)
(517)
X2 = 30.26, df=2, p<.001

Table 4.6:  Daily Show Exposure and Campaign Interest

Low
Campaign Interest (%)
Medium
Campaign Interest (%)
High
Campaign Interest (%)
Non-Viewers
87
68
54
Viewers
13
32
46
(Valid Cases)
(117)
(616)
(517)
X2 = 51.821, df=2, p <.001.









Table 4.7:  Late Night Television Negative Attributes of Bush and Kerry

BUSH
KERRY
Daily Show
Conan
Leno
Letterman
Daily Show
Conan*
Leno
Letterman*
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
19 (2)
0 (7)
12 (4)
0 (7)
8 (5)
9 (4)
0 (9)
21 (2)
Bio and Record
8 (5.5)
25 (2)
9 (5.5)
0 (7)
4 (7)
9 (4)
3 (6.5)
8 (4.5)
Integrity and Values
3 (9)
13 (3.5)
0 (8.5)
0 (7)
4 (7)
0 (7.5)
3 (6.5)
0 (8)
Personality / Image
11 (4)
50 (1)
21 (2)
45 (1)
23 (2)
27 (2)
40 (1)
46 (1)
War
16 (3)
0 (7)
9 (5.5)
24 (2)
12 (4)
0 (7.5)
10 (3.5)
0 (8)
Speaking / Debates
24 (1)
0 (7)
6 (7)
17 (3)
19 (3)
45 (1)
3 (6.5)
13 (3)
Flip Flop / Decisiveness
5 (5.5)
0 (7)
0 (8.5)
0 (7)
27 (1)
0 (7.5)
10 (3.5)
8 (4.5)
Intelligence
8 (7)
13(3.5)
26 (1)
14 (4)
0 (9)
9 (4)
3 (6.5)
4 (6)
Other
5 (8)
0 (7)
18 (3)
0 (7)
4 (7)
0 (7.5)
27 (2)
0 (8)
N=
37
8
34
29
26
11
30
24

* Conan and Letterman shows had similar negative Kerry attributes 
(Spearman's rho = .793, p<.01)



Table 4.8: Exposure to the Big 3 Shows, Participants' Negative 
Attributes of Candidates and Negative Attributes of Presidential 
Candidate on Late Night Television

BUSH
KERRY
TV / Jokes*
Viewers*
Non-Viewers
TV / Jokes
Viewers
Non-Viewers
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
6 (7)
12 (5)
12 (4.5)
9 (4)
15 (3)
11 (4)
Bio and Record
7 (6)
16 (3)
17 (1)
6 (6)
4 (6)
4 (6)
Integrity and Values
1 (8)
6 (8)
8 (7)
2 (9)
5 (5)
5 (5)
Personality / Image
34 (1)
18 (1)
16 (2)
40 (1)
32 (1)
40 (1)
War
14 (3)
11 (6)
12 (4.5)
5 (7.5)
.5 (8.5)
.4 (7.5)
Speaking / Debates
10 (4)
7 (7)
6 (8)
14 (2)
2 (7)
.4 (7.5)
Flip Flop / Decisiveness
0 (9)
1 (9)
2 (9)
8 (5)
27 (2)
23 (2)
Intelligence
20 (2)
17 (2)
15 (3)
5 (7.5)
.5 (8.5)
.4 (9)
Other
8 (5)
12 (4)
11 (6)
12 (3)
14 (4)
16 (3)
N =
71
298
274
65
214
223
Conan, Leno and Letterman Combined
Late Night Television is Letterman, Leno and Conan
* Spearman's rho =  .733, p<.05

  Table 4.9: Exposure to Daily Show, Participants' Negative 
Attributes of Candidates and Negative Attributes of Presidential 
Candidate on Daily Show

BUSH
KERRY
TV / Jokes
Viewers
Non-Viewers
TV / Jokes
Viewers
Non-Viewers
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
19 (2)
15 (3.5)
9 (6.5)
8 (5)
19 (3)
10 (4)
Bio and Record
8 (5.5)
20 (1)
14 (3)
4 (7)
4 (5)
4 (6)
Integrity and Values
3 (9)
7 (7)
8 (8)
4 (7)
3 (6)
5 (5)
Personality / Image
11 (4)
18 (2)
16 (2)
23 (2)
37 (1)
35 (1)
War
16 (3)
11 (5)
12 (5)
12 (4)
1 (8)
1 (8.5)
Speaking / Debates
24 (1)
2 (9)
9 (6.5)
19 (3)
2.3 (7)
2 (7)
Flip Flop / Decisiveness
5 (7.5)
3 (8)
1 (9)
27 (1)
24 (2)
26 (2)
Intelligence
8 (5.5)
15 (3.5)
17 (1)
0 (9)
0 (9)
1 (8.5)
Other
5 (7.5)
10 (6)
13 (4)
4 (7)
11 (4)
17 (3)
N =
37
245
327
26
151
286


Table 4.10:  Experiment Video And Experiment Group's Negative 
Candidate Attributes

BUSH
KERRY
Video
Viewers
Video*
Viewers*
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
9 (5)
25 (1.5)
6 (4.5)
13 (4)
Bio and Record
25 (1.5)
19 (3)
3 (6)
0 (8)
Integrity and Values
0 (9)
6 (6)
0 (8)
13 (4)
Personality / Image
13 (3.5)
6 (6)
23 (2)
13 (4)
War
13 (3.5)
25 (1.5)
0 (8)
0 (8)
Speaking / Debates
6 (6.5)
6 (6)
10 (3)
13 (4)
Flip Flop / Decisiveness
3 (8)
0 (8.5)
52 (1)
38 (1)
Intelligence
25 (1.5)
0 (8.5)
0 (8)
0 (8)
Other
6 (6.5)
13 (4)
6 (4.5)
13 (4)
N =
32
16
31
8
* Spearman's rho =  .761, p<.01

Table 4.11: The Big 3 Shows Viewers' Negative Bush Attributes, the 3 
Big Shows' Negative Attributes of Bush and Party Affiliation

Late-Night Jokes
Democratic
Republican
Independent
%  (RO)*
%  (RO)
%  (RO)
%  (RO)
Ideology / Vision
6 (7)
11 (5)
12 (4.5)
19 (1.5)
Bio and Record
7 (6)
18 (1)
9 (6)
19 (1.5)
Integrity and Values
1 (8)
7 (7)
6 (8)
6 (7.5)
Personality / Image
34 (1)
16 (3)
29 (1)
6 (7.5)
War
14 (3)
10 (6)
12 (4.5)
10 (5.5)
Speaking / Debates
10 (4)
5 (8)
15 (2.5)
10 (5.5)
Flip Flop / Decisiveness
0 (9)
1 (9)
0 (9)
0 (9)
Intelligence
20 (2)
17 (2)
15 (2.5)
16 (3)
Other
8 (5)
15 (4)
3 (7)
13 (4)
N =
71
164
34
31

Conan, Leno and Letterman Combined
* Spearman's rho =  .824, p<.01





Table 4.12: Daily Show Viewers' Negative Bush Attributes, Daily 
Show's Negative Attributes of Bush and Party Affiliation

Daily Show Jokes*
Democratic
Republican*
Independent
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
19 (2)
13 (4)
23 (1.5)
16 (3)
Bio and Record
8 (5.5)
23 (1)
9 (5.5)
17 (2)
Integrity and Values
3 (9)
8 (7)
5 (7)
6 (7)
Personality / Image
11 (4)
15 (2.5)
23 (1.5)
22 (1)
War
16 (3)
10 (6)
14 (4)
14 (4)
Speaking / Debates
24 (1)
1 (9)
9 (5.5)
3 (8)
Flip Flop / Decisiveness
5 (7.5)
4 (8)
0 (8.5)
1 (9)
Intelligence
8 (5.5)
15 (2.5)
18 (3)
13 (5)
Other
5 (7.5)
11 (5)
0 (8.5)
7 (6)
N =
37
132
22
69

* Spearman's rho =.660, p<.05

Table 4.13: The 3 Big Show Viewers' Negative Kerry Attributes, the 3 
Big Shows' Negative Attributes of Kerry and Party Affiliation

Late-Night Jokes*
Democratic*
Republican
Independent
%  (RO)
%  (RO)
%  (RO)
%  (RO)
Ideology / Vision
9 (4)
8 (4)
18 (3)
11 (3)
Bio and Record
6 (6)
5 (5)
4 (6)
0 (6.5)
Integrity and Values
2 (9)
5 (5)
4 (6)
0 (6.5)
Personality / Image
40 (1)
41 (1)
25 (2)
56 (1)
War
5 (7.5)
0 (8.5)
1 (8)
0 (6.5)
Speaking / Debates
14 (2)
2 (7)
1 (8)
0 (6.5)
Flip Flop / Decisiveness
8 (5)
24 (2)
31 (1)
33 (2)
Intelligence
5 (7.5)
0 (8.5)
1 (8)
0 (6.5)
Other
12 (3)
17 (3)
13 (4)
0 (6.5)
N =
65
59
106
9
  Conan, Leno and Letterman Combined
* Spearman's rho =  .613, p<.05


Table 4.14: Table 4.10: Daily Show Viewers' Negative Kerry 
Attributes, Daily Show's Negative Attributes of Kerry and Party Affiliation

Daily Show Jokes
Democratic
Republican
Independent
% (RO)
% (RO)
% (RO)
% (RO)
Ideology / Vision
8 (5)
15 (4)
23 (3)
19 (2)
Bio and Record
4 (7)
4 (5)
2 (6.5)
5 (5.5)
Integrity and Values
4 (7)
2 (6)
6 (5)
3 (7)
Personality / Image
23 (2)
35 (1)
27 (2)
43 (1)
War
12 (4)
0 (8)
2 (6.5)
0 (8.5)
Speaking / Debates
19 (3)
0 (8)
0 (8.5)
5 (5.5)
Flip Flop / Decisiveness
27 (1)
25 (2)
31 (1)
14 (3)
Intelligence
0 (9)
0 (8)
0 (8.5)
0 (8.5)
Other
4 (7)
19 (3)
8 (4)
11 (4)
N =
26
48
48
37
p>.05






























References

Amundson, S., & Lichter, S. R.  (1988, July/August). Heeeeeere's 
politics.  Public Opinion, 11, p. 46.

An, C. and Pfau, M. (2004) The Efficacy of Inoculation in Televised 
Political Debates. Journal of Communication, 432.

Baum, M.A.  (2003).  Soft news and political knowledge:  Evidence of 
absence or absence of evidence?  Political Communication, 20, 173-190.

Baxter, L. A., & Babbie, E.  (2004).  The basics of communication 
research.  Wadsworth/Thomson Learning:  Belmont, CA.

Bennet, J.  (1996, July 9). Did you hear the one about the'96 
campaign? New York Times, p. A17.

Behr, R. L. & Iyengar. S. (1985) Television News, Real-World Cues and 
Changes in the Public Agenda. Public Opinion Quarterly, 49 (1), 38-57.

Brants, K. (1998) Who's Afraid of Infotainment? European Journal of 
Communication, 13, 315-336.

Campaign for Young Voters. (2004) How they decide their vote. 
Retrieved Dec. 11, 2004.
   	http://www.campaignyoungvoters.org/toolkit/find/howdecide/index.html.

Carlin, D.P. (2000) Watching the Debate:  A Guide for Viewers. In S. 
Coleman (Ed.), Televised Election Debates (pp. 157-177). New York, 
St. Martin's Press.

Center for Information & Research on Civic Learning & Engagement 
(2002), "Youth Voter Turnout in
the States during the 1998 Midterm and 2000 Presidential Election," 
Retrieved Dec. 11, 2004
from http://www.civicyouth.org.

Center for Information & Research on Civic Learning & Engagement 
(2003), "Youth Voter Turnout Up
Sharply in 2004," Retrieved Dec. 20, 2004 from http://www.civicyouth.org.

Center for Information & Research on Civic Learning & Engagement 
(2003), "College Students
Mobilized, Engaged and Backed Kerry," Retrieved Dec. 20, 2004 from
http://www.civicyouth.org

Center for Media and Public Affairs (2004, March 22) "Late Night 
Jokes Jab Democrats More Than Bush,"
  Retreived Dec. 19, 2004 from
http://www.cmpa.com/LateNightJokesJabDemocratsMoreThanBush.htm..

Chaffee, S. H., Zhao, X., & Leshner, G.  (1994, June) Political 
knowledge and the campaign Media of 1992.  Communication Research, 21 
(3), p. 313

Finke, Nikki. (2004, September 17) Does Mr. Middle-of-the-Road Lean 
Left? [Electronic version]. LA
Weekly, Retrieved Dec. 12, 2004, from 
http://www.laweekly.com/ink/04/43/deadline-finke.php

The Gallup Organization (2004, October 18) "Bush Gains After Debates."
  Retrieved Dec. 12, 2004 from 
http://www.gallup.com/poll/content/login.aspx?ci=13657

Glass, D.P. (2001) Evaluating Presidential Candidates: Who Focuses on 
Their Personal Attributes? Public Opinion Quarterly, 49, 517.

Golan, G. and Wanta, W. (2001) Second-level Agenda Setting in the New 
Hampshire Primary: A Comparison of Coverage in Three Newspapers and 
Public Perception of Candidates. Journalism & Mass Communication 
Quarterly, 78, 247-259.

Graber, D. A. (1972) Personal Qualities in Presidential Images: The 
Contribution of the Press. Midwest Journal of Political Science, 16, 46-76.

Hacker, K. L., Zakahi, W. R., Giles, M. J., & McQuitty, S.  (2000) 
Components of candidate images:  Statistical analysis of the 
issue-persona dichotomy in the presidential campaign of 
1996.  Communication Monographs, 67 (3), 227-238.
Halonen, D.  (2000, February 14).  For election news, young stay up 
late. Electronic Media, 19 (7).

Hill, D. B.  (1985).  Viewer Characteristics and Agenda Setting by 
Television News.  Public Opinion Quarterly, 49 (3), 340-350.

Hollander, B.A. (1995) The New News and the 1992 Presidential 
Campaign:  Perceived vs. Actual Political Knowledge. Journalism & 
Mass Communication Quarterly, 42, 786-798.

Kerbel, M. R.  (1995).  Remote and controlled:  Media politics in a 
cynical age.  Westview Press:  Boulder, CO, p. 133

Kihan, K. and McCombs, M., (2004) Impact of Attribute Agenda-Setting 
on Attitudes presented at the annual conference of the International 
Communication Association in New Orleans, LA, 8-9.

Kiousis, S., Bantimaroudis, P. and Ban, H. (1999) Candidate Image 
Attributes:  Experiments on the Substantive Dimension of Second Level 
Agenda Setting. Communication Research, 26, 414-428.

Lasorsa, D. L., & Wanta, W.  (1990)  The effects of personal, 
interpersonal and media experience on issue salience. Journalism 
Quarterly, 67 (4), 804-813.

McCombs, M. (2004) Setting the agenda:  The mass media and public 
opinion. Cambridge, UK: Polity Press.

McCombs, M., Lopez-Escobar, E., and Pablo Llamas, J. (2000) Setting 
the Agenda of Attributes in the
  	1996 Spanish General Election. Journal of Communication, 50, 77-92.

McCombs, M. and Shaw, D. (1972) The agenda setting function of the 
mass media. Public Opinion Quarterly, 36, 176-187.

McLeod, J.M., Guo, Z., Daily. K., Steele, C.A., Huang, H., Horowitz, 
E. and Chen, H. (1996) "The Impact of Traditional and Nontraditional 
Media Forms in the 1992 Presidential Election. Journalism & Mass 
Communication Quarterly, 73, 401-416.

McCroskey, J.C., & Jenson, T. A.  (1975). Image of mass media news 
sources.  Journal of Broadcasting, 19 (2), 169-180.

Miller, A. H. and MacKuen, M. (1979) Learning About the Candidates: 
The 1976 Presidential Debates.
  	Public Opinion Quarterly, 326-346.

Neuendorf, K. A.  (2002). The content analysis guidebook. Thousand 
Oaks, CA: Sage
  	Publications, p. 149.

Nilsen, D. L. F, (1990) The Social Functions of Political Humor. 
Journal of Popular Culture, 24, p. 35.

Nimmo, D., & Savage, R. L.  (1976).  Candidates and their images: 
Concepts, methods, and findings. Pacifica Palisades, CA: Goodyear 
Publishing Company, Inc., 5.

Niven, D., Lichter, S.R. and Amundson, D. (2003) The Political 
Content of Late Night Comedy. Press/Politics, 8 (3), 118-133.

Pfau, M. and Eveland Jr., W.P., (1996) Influence of Traditional and 
Non-Traditional News Media in the 1992 Election Campaign. Western 
Journal of Communication, 60.

Pew Research Center (2004) "Cable and Internet Loom Large in 
Fragmented Political News Universe,"
  	Retrieved Dec. 12, 2004 from 
http://people-press.org/reports/display.php3?ReportID=200

Riffe, D., Aust, C.F., & Lacy, S.  (1993). The effectiveness of 
random, consecutive day and constructed
  	week samples in newspaper content analysis. Journalism Quarterly, 
70, 133-139.

Schechter, D.  (2003).  Are we reaching da youth? Nieman Reports, 57 
(4), p. 12.

Schechter, D.  (2003).  Solving some mysteries about the habits of 
youth. Nieman Reports, 57 (4), pp. 14-17.

Schonlau, M., Fricker, R. D. & Elliott, M. N. (2002) Conducting 
Research Surveys via E-Mail and the Web. Santa Monica, Calif: RAND.

Schutz, C. E.  (1977).  Political humor:  From Aristophanes to Sam 
Ervin.  Cranbury, NJ: Associated University Presses, Inc., p. 197.

Sigel, R. S. (1964) Effect of Partisanship on the Perception of 
Political Candidates. Public Opinion
  	Quarterly, 28, p. 483-496.

Weinraub, B.  (2000, January 19). Election's barometer: Barbs of 
late-night TV. New York
  	Times, p. A16.

Weaver, D. H.  (1991). Political issues and voter need for 
orientation. In Protess, D. L., & McCombs, M. (Eds.) Agenda 
Setting:  Readings on Media, Public Opinion, and Policy Making. 
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc., 131-139.

Weaver, D. and Drew, D. (1995) Voter Learning in the 1992 
Presidential Election:  Did the 'Nontraditional' Media and Debates 
Matter? Journalism & Mass Communication Quarterly, 1995, 72, 7-17.

Wizda, S.  (2000, November) Late night with Al and Dubya. American 
Journalism Review, p. 10.

Young, G. D.  (2004, September 21).  Daily Show viewers knowledgeable 
about presidential campaign, National Annenberg Election Survey 
Shows. Annenberg Public Policy Center news release, p. 1.

[1]  This age category is by Pew Research studies, rather than other 
organizations that classify young voters as ages 18-24.
[2]  It is important to note the different exposure definition of 
"never or seldom" watchers from the larger Web-based survey study of 
non-watchers and watchers. The experiment was related to the overall 
study, but should be seen as a separate smaller study to examine 
those who have little exposure to late night television.
[3]  The alternating experiment and control rooms were set up 
similarly with a TV, couches, chairs and tables in an on-campus 
student building.
[4]  Response rates for Internet surveys range from 7 to 44 percent. 
(Schonlau, Fricker & Elliot, 2002). Web surveys continue to yield a 
variety of response rates.