Citation
What Media Do to Bi-Polar, 50-50 States

Material Information

Title:
What Media Do to Bi-Polar, 50-50 States
Creator:
Yun, Hyun Jung
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (96 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Mass Communication
Journalism and Communications
Committee Chair:
Kaid, Lynda L.
Committee Members:
Ferguson, Mary Ann
Kiousis, Spiro K.
Hedge, David M.
Graduation Date:
8/11/2007

Subjects

Subjects / Keywords:
Cynicism ( jstor )
Political attitudes ( jstor )
Political awareness ( jstor )
Political campaigns ( jstor )
Political candidates ( jstor )
Political change ( jstor )
Political ideologies ( jstor )
Political parties ( jstor )
Politics ( jstor )
Voting ( jstor )
Journalism and Communications -- Dissertations, Academic -- UF
attitude, bias, candidate, cynicism, efficacy, election, geopolitics, information, media, political
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Mass Communication thesis, Ph.D.

Notes

Abstract:
As geopolitics assumes, people in different places have different ways of thinking, personalities, and behaviors because they experience and are exposed to different social and political environment. In addition, as selective exposure and hostile media phenomena assume, individuals have a tendency to prefer information corresponding to and avoid information inconsistent with their own beliefs. For these reasons, individuals perceive even the same political information differently based on where they are placed. If individuals are placed in a political battle ground, they are naturally exposed to multiple sources and types of political information, while other individuals in strong Democrat or Republican districts would have smaller amounts of campaign information. Liking and disliking candidates is strongly associated with individuals? political predispositions, but media exposure tends to reinforce or decrease their political beliefs towards targeting political candidates and issues they are interested in. Therefore, by knowing the media dynamics on political preferences among people within certain political boundaries, we are able to predict with a better explanatory power. This study particularly looks at the dynamics between the amounts of political information available and individuals' political attitudes in battle and non-battle ground states. Using experiment data (N=2965) gathered by the Uvote research team during the 2004 presidential election at multiple locations, this study compares two groups of individuals who have been exposed to multiple campaign information in battle-grounds and who have been exposed to limited, no, or indirect campaigning information in non-battle grounds to measure individuals? political attitudes based on political information availability. In addition, this study also measures how these individuals with various levels of information accessibility evaluate political candidates, recognize issues, and perceive media messages differently by equally given sets of political ads and debates. The study found that individuals in battle-ground states with multiple political information are more likely to have higher levels of information efficacy and lower levels of cynicism, tend to evaluate candidates positively, more easily change their previous opinions, and tend to have a broader range of issue awareness than people in non-battle ground states with limited or no campaigning information, with some variations within demographic factors. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2007.
Local:
Adviser: Kaid, Lynda L.
Statement of Responsibility:
by Hyun Jung Yun.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Yun, Hyun Jung. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
662508846 ( OCLC )
Classification:
LD1780 2007 ( lcc )

Downloads

This item has the following downloads:

yun_h.pdf

yun_h_Page_86.txt

yun_h_Page_51.txt

yun_h_Page_65.txt

yun_h_Page_26.txt

yun_h_Page_17.txt

yun_h_Page_15.txt

yun_h_Page_24.txt

yun_h_Page_89.txt

yun_h_Page_18.txt

yun_h_Page_32.txt

yun_h_Page_64.txt

yun_h_Page_49.txt

yun_h_Page_42.txt

yun_h_Page_88.txt

yun_h_Page_19.txt

yun_h_Page_06.txt

yun_h_Page_48.txt

yun_h_Page_62.txt

yun_h_Page_54.txt

yun_h_Page_60.txt

yun_h_Page_81.txt

yun_h_Page_10.txt

yun_h_Page_33.txt

yun_h_Page_29.txt

yun_h_Page_28.txt

yun_h_Page_93.txt

yun_h_Page_57.txt

yun_h_Page_91.txt

yun_h_Page_27.txt

yun_h_Page_63.txt

yun_h_Page_21.txt

yun_h_Page_58.txt

yun_h_Page_25.txt

yun_h_Page_53.txt

yun_h_Page_43.txt

yun_h_Page_87.txt

yun_h_Page_69.txt

yun_h_Page_45.txt

yun_h_Page_82.txt

yun_h_Page_02.txt

yun_h_Page_79.txt

yun_h_Page_03.txt

yun_h_Page_71.txt

yun_h_Page_67.txt

yun_h_Page_84.txt

yun_h_Page_14.txt

yun_h_Page_35.txt

yun_h_Page_09.txt

yun_h_Page_47.txt

yun_h_Page_55.txt

yun_h_Page_01.txt

yun_h_Page_59.txt

yun_h_Page_37.txt

yun_h_Page_46.txt

yun_h_Page_44.txt

yun_h_Page_78.txt

yun_h_Page_08.txt

yun_h_Page_20.txt

yun_h_Page_38.txt

yun_h_Page_77.txt

yun_h_Page_31.txt

yun_h_Page_72.txt

yun_h_Page_41.txt

yun_h_Page_13.txt

yun_h_Page_61.txt

yun_h_Page_85.txt

yun_h_Page_68.txt

yun_h_Page_11.txt

yun_h_Page_94.txt

yun_h_Page_66.txt

yun_h_Page_16.txt

yun_h_Page_50.txt

yun_h_Page_04.txt

yun_h_pdf.txt

yun_h_Page_30.txt

yun_h_Page_34.txt

yun_h_Page_22.txt

yun_h_Page_23.txt

yun_h_Page_36.txt

yun_h_Page_75.txt

yun_h_Page_90.txt

yun_h_Page_07.txt

yun_h_Page_12.txt

yun_h_Page_52.txt

yun_h_Page_39.txt

yun_h_Page_80.txt

yun_h_Page_96.txt

yun_h_Page_73.txt

yun_h_Page_95.txt

yun_h_Page_92.txt

yun_h_Page_40.txt

yun_h_Page_83.txt

yun_h_Page_56.txt

yun_h_Page_70.txt

yun_h_Page_05.txt

yun_h_Page_74.txt

yun_h_Page_76.txt


Full Text





individuals' perceptions of media and even distinguish between different media outlets, and

these are not easily changed (Gussin & Baum, 2004). Individuals' assessments of source

credibility are determined by their internal factors such as political awareness or partisanship.

Individuals who have a particular political tendency are likely to evaluate or interpret media

coverage accordingly. For instance, strong partisans believe that media tend to be biased and

perceive different implication from the same sort of messages; thus, they tend to be more

selective for information and strengthen pro and con attitudes than weak partisans (Schmitt,

Gunther, & Liebhart, 2004, pp.623-41).

The more committed an individual is to an attitude, the greater he or she is to resist

attempts to change it (Hovland, Campbell, & Brock, 1957). The more knowledgeable a person

becomes about an attitude abj ect, the harder it will likely be to change his or her attitude toward

the obj ect because there is so much support for the existing viewpoint (Petty & Krosnick, 1995,

p.4).

Political audiences who have strong attitudes are actively involved in the processing of

political information. If they perceive counter-attitudinal messages from media, they tend to

think the media message is biased. "It is not just a difference of the opinion but a difference of

perception" (http://www.answers. com/topic/ hostile-media-effect) Thus, either multiple

political messages or a limited single piece of information can be faulty facts as long as the

media messages contain information that discomfort audiences' preexisting political beliefs and

preferences. Audiences' political beliefs do influence their perception of media messages.

Vallone, Ross and Lepper (1985) illustrated that Arab and Israeli students both felt broadcast

news on the Middle East conflict was biased against them.










Independents in information rich states. Political Independents tend to be cynical about political

views of both sides of Democratic and Republican parties. Independents distinguish themselves

by rejecting the major parties. Recalling American history, unrest among youth, civil right

groups, and disengaged people from strong party ties, especially related to issues of war and

human rights, have been political Independents. Therefore, these groups tend to be naturally

more cynical, but somewhat flexible in their political choice since they use policy issues that are

given for each particular election in order to make political decisions (Erikson & Tedin, 2007,

pp.84, 119, & 270).

Again, although direct information exposure did not make any difference in change of

political cynicism between people from information rich and poor states, information availability

interacts with individuals' party identification in changing individuals' cynical attitudes. After

political information exposure, Democrats and Independents in the political information rich

states reduced their levels of cynicism, but Democrats and Independents in the information poor

states remained with their previous level of political cynicism. This tendency suggests that

although the high level of political information always tends to reduce political cynicism,

individuals with low political information often have no particular tendency or motivation to

change their previous attitudes (Erikson & Tedin, 2007, p.86).

Candidate Evaluations

This study found that different levels of political information availability influence

individuals' candidate evaluations for Democrat candidate, Kerry, but not for the Republican

candidate, Bush. More specifically, individuals in the political information rich states were more

likely to evaluate Kerry positively than individuals in political information poor states.

In terms of candidate evaluation, although political information tends to increase

candidates' evaluations (Berman, 1997; Shaw, 1999), there have been somewhat consistent










Kaid, L. L.(Ed.), (2004). The H~andbook of Political Communication Research. Mahway, NJ:
Lawrence Erlbaum Associates, Publishers

Kaid, L.L., & Holtz-Bacha, C. (1995). Political advertising in Western democracies. Thousand
Oaks, CA: Sage.

Kaid, L.L., Landreville, K., Postelnicu, M., and Martin, J. (2005, May). Enhancing information
efficacy for young voters: The effects of political advertising and debates. Paper presented
at the International Communication Association Conference, New York City, NY

Kaid, L. L., Mc~inney, M. S. & Tedesco, J. C. (2000). Civic dialogue in the 1996 presidential
campaign: Candidate, media, and public voices. Cresskill, NJ: Hampton

Kaid, L. L., Mc~inney, M. S., & Tedesco, J. C. (2007). Political information efficacy theory and
younger voters. American Behavioral Scientist, 50(9), 1093-1111

Kaid, L.L., Postenicu, M., Landrevill, K., Yun, H. J., and Hendren, A. G. (2007). The Effects of
Political Advertising on Young Voters. American Behavioral Scientists, 50(9), 1137-1151

Kamieniecki, S. (1988). The Dimensionality of Partisan Strength and Political Independence.
Political Behavior, 10(4), 364-376.

Katz, E., & Feldman, J.J. (1962). The debates in light of research: A survey of surveys. In S.
Kraus (ed.), The great debates: Kennedy vs. Nixon, 1960 (pp.173-223). Bloomington:
Indiana University Press.

Klapper, J. T. (1960). The Effects of Ma~ss Communication. New York: Free Press.

Koch, J. W. (1998). Political Rhetoric and Political Persuasion: The Changing Structure of
Citizens' Preferences on Health Insurance During Policy Debate. Public Opinion
Quarterly, 62, 209-229

Krosnick, J. A., Boninger, D.S., Chuang, Y. C., Berent, M.K., & Carnot, C. G. (1993). Attitude
Strength: One Construct or many related constructs? Journal ofPersonality and Social
Psychology, 65, 1132-1151

Kubey, R., & Csikszentmihalyi, M. (1991). Television and the Quality ofLife. Mahwah, NJ:
Lawrence.

Lane, R. E. (1969). Political Life: Why People Get Involved in Politics. NY: Free Press.

Langton, K. P., & Karns, D. A. (1969). The Relative Influence of the Family, Peer Group, and
School in the Development of Political Efficacy. The Western Political Quarterly, 22(4),
813-826.

Layman, G. C., & Carsey, T. M. (2002). Party Polarization and Party Structuring of Policy
Attitudes: A Comparison of Three NES Panel Studies. Political Behavior, 24 (3), 199-236.










personal trait appeal can also be explained by a new and fresh image of a challenger above

heavily negative tags on the incumbent (Romero, 1996).

Issue Awareness

As expected, information availability influenced the range of political issues recognition.

Individuals who were in information rich states listed larger numbers of issues that they believed

important than people from information poor state. Direct information exposures to a series of

political information reduced the gap in range of policy issues awareness and the rank of issue

importance among individuals in different political regions (Shaw, 1999; Funk, 1999; Atkin &

Gary Heald,1976).

There have been consistent findings that the political audience learned candidates'

substantive policy issues and image through paid campaign ads (Kaid, McKinney, & Tedesco,

2007; Freedman, Franz, & Goldstein, 2004). Less informed individuals are more likely to be

susceptible to new information than well-informed individuals, thus easily perceiving new issues

(Nelson, Oxley, & Clawson, 1997, p.233). With higher levels of information, individuals who

are in battleground states are more likely to understand current issues comprehensively (Benoit

et al., 2004). In other words, opinion consistency on political issues is higher among more

politically informed individuals than the less knowledgeable (Erikson & Tedin, 2007, p.81). The

result proved that media and the interaction with geopolitical constraints on level of information

shape individuals' perception of current issues (Iyengar & Kinder, 1987).

Media Bias

As hostile media phenomena assume, individuals with strong political preference in non-

battleground states are more likely to think that media are biased than individuals in flexible and

moderate political tendencies in battleground states. Even when media messages are neutral,

individuals believe that perceived media contents are biased if the contents discomfort them









one of the indexes, cynicism, had a relatively low level of Cronbach' s score (.66); this could lead

to less significant results.

Theoretically, political psychologists stress the mechanisms that shape political opinions

by the way information is processed (Erikson & Tedin, 2007, p.66). With multiple media

channel choices and specialized media contents, political audiences could be both homogeneous

and heterogeneous. The one obvious phenomenon is segmentation and fragmentation of

audiences. In the results, media tend to be much more diverse to manage different consumers

with appropriate products and provide more individualized contents. Such segmentation and

fragmentation create difficulties for media with extensiveness and unpredictability to reach both

the larger general public and smaller target groups (Mcquail, 1997, p. 133; Barnes & Thomson,

1994, p.89). Hence, political information becomes more partisan and sensationalistic.

Interpretative journalism is unavoidable phenomenon. As a mediator, the media is in the middle

of political communication between politicians and public, and control the content types and size

of political information for the public (Erikson & Tedin, 2007, pp.229-3 5).

Therefore, the theoretical generalization of this study faces some limitation by inherent

elements of media information and individual diversity. In the multimedia era, if there is no

direct political information available, citizens use other channels of informal political

information such interpersonal political discussion. In other words, individuals who are in lack

of political information somehow fill the vacuum by increasing informal communication about

politics (Converse, 1962, p.595-97)

In addition, there might be a great dimension of individual differences that any theoretical

research cannot completely track. As we age, individuals exposed to different components of

media, symbolic roles, and life patterns adopt different processes of information perception.









However, unlike the fourth hypothesis predicted, direct information exposure did not

make any difference in change of political cynicism between people from information rich and

poor states. Nonetheless, the information availability is again interacted with individuals' party

identification in changing individuals' cynical attitudes. After political information exposure,

Democrats (M=-0.05, SD=0.44) and Independents (M=-0.05, SD=0.39) in the political

information rich states reduced their levels of cynicism, but Democrats (M=-0.01, SD=0.40) and

Independents (M=0.01, SD=0.39) in the information poor states remained at their previous level

of political cynicism, while controlling other demographic variables (F[2, 2580]=3.076, P<.046)

(Figure 4-4 & Tables 4-4 & 4-5).

Candidate Evaluations

Overall, older adults were more supportive of Bush (t=2.914, P<.004), and younger

adults were more supportive of Kerry (t=-4.398, P<.001). Ethnic maj orities (M=54.45,

SD=34.89) evaluated Bush more positively than ethnic minorities (M=39.40, SD=31.67) (F[1,

2657]=11.373, P<.001), however, ethnic minorities (M=58.69, SD=26.84) evaluated Kerry more

positively than ethnic majorities (M=45.96, SD=30.64) (F[1, 2596]=11.971, P<.001). As

expected, Republicans were strong supporters of Bush (M=81.50, SD=18.57), but Democrats

evaluated Bush relatively negatively (M=24. 11, SD=23.68) (F[2, 2657]=798.540, P<.001). In

contrasts, Democrats were strong supporters of Kerry (M=71.39, SD=20.03), but Republicans

evaluated Kerry significantly more negatively (M=25.54, SD=23.45) (F[2, 2596]=524.947,

P1.001). In addition, previous political information consumption induced the tendency of

negative evaluation toward Bush (t=-3.253, P<.001).

Hypothesis 5 arguing that different levels of political information availability influence

individuals' candidate evaluations was partially confirmed, in that it appeared true for the

Democrat candidate, Challenger Kerry, but not for the Republican candidate, incumbent Bush.









APPENDIX E
IVEDIA BIAS

* News organizations, such as newspapers and television news, try to manipulate public
opinion

* News organizations often fail to get all of the facts straight.

* News organizations often don't deal fairly with all sides of a political or social issue.

* News organizations do a poor j ob of separating facts from opinions.

* News organizations are concerned with the community's well-being.

* News organizations watch out for my interests.

* News organizations are concerned mainly about the public welfare.









BIOGRAPHICAL SKETCH

Hyun Jung Yun grew up in South Korea and completed her undergraduate work maj oring

in Political Science at Aj ou University. Starting with her undergraduate senior year as an

exchange student, she continued her master' s degree studying in political science and doctorate

in the fields of 'political science' and 'j ournalism and communications' at the University of

Florida in the United States of America

Her research interests are in political perception, the political communication process,

policy attitudes, persuasion, and geopolitics across different levels of the individual, small group,

and aggregate group. More specifically, her research in the Hield of political communication

explores the relationship between political information process and individuals' political

attitudes in different geopolitical circumstance. In the same line of interdisciplinary research, her

research in political science investigates how individuals' beliefs about various policies are

influenced by varying levels of multi-dimensional social capital and communication networks.

Her research in journal publications demonstrates how individuals' political perceptions

and attitudes are influenced by political predispositions within a group and by political resources

within a given political and media system at the aggregate level. In addition, she had coauthored

several book chapters examining news coverage of policy issues and political candidates across

different political regions to observe the relationship between different political characteristics

and political information effects. She is currently working on analyzing the dual spirals of

silence in policy opinion formation between issue minority and issue maj ority, effects of

relationship between media and politics on voter perceptions, as well as political cynicism and

information efficacy in young voters.

She also has participated in grant-supported research proj ects including the Florida

Department of Health' s 2004 Proj ect in Media Terrorism Preparation under Dr. Mary Ann









According to an empirical study, Republican audiences are more likely to see news

coverage as leaning to Democratic candidates, but Democratic identifiers are more likely to see

the coverage as leaning to the Republican candidates. As the hostile media effects theory

assumes, partisans may see even neutral political information as hostile to their own viewpoint

(Gunther & Schmitt, 2004; Schmitt, Gunther, &Liebhart, 2004). Another empirical study found

that each individual reacts to the same media messages differently based on their preexisting

political preferences and issue relevance. When two groups on opposite points of the Middle

East issue were exposed to an identical news story about the conflict, both groups evaluated the

content as biased against to their views (Dalton, Beck, & Huckfeldt, 1998).

According to belief congruence theory by Rokeach and Rothman (1965), audiences'

previously built values, attitude, and belief system serve as protection layers for external

influence. Krosnick et al.'s (1993) estimates of correlations between latent attitude dimensions,

talking and thinking (R=.84), previous knowledge and talking (R=.79), and previous knowledge

and thinking (R=.76) are highly correlated. Thus, hostile media effect at the level of the

individual needs to be considered as a precondition for political communication.

At the Macro-Level

Hovland, Irving, and Kelley (1953) argued that individuals' conforming tendencies are

highly influenced by expectations or perspectives of committed groups with whom they are

associated. Consequently, political candidates use various strategies to appeal their positions in

order to get political support from different groups of people. Different communication channels

have different effects on contrasting types of people. Various political audiences find different

types of media to acquire desirable information (Schooler, Chaffee, Flora, & Roser, 1998).

Hence, effective campaigns use more than one type of media, combine mass media with other

types of activities, such as interpersonal interactions, and represent their view through experts of





Table 4-11. Information Availability on Personal Attribute Evaluation
Bush Kerry
Low Attention States M=51.90 (SD=13.74) M=52.30 (SD=12.49)
High Attention States M=53.04 (SD=14.89) M=53.49 (SD=11.56)
Not significant Not significant


Change in Bush's Trait
M=0.49 (SD=6.59)
M=0.54 (SD=5.92)
Not significant


Change in Kerry Trait
M=1.18 (SD=7.82)
M=1.94 (SD=6.91)
F [1, 2636]=7.215, P<.007









various fields for more effective political appeals to various political audiences (Hofstetter,

Shultze, & Mulvihill, 1992).

Eveland and Shah (2003) found that perception of media bias was more likely related to

consistent information exchange among ideologically similar individuals rather than amount of

discussion. Group preferences are another important factor for exploring media effects. For

instance, their study found that such a phenomenon is more prominent among Republicans than

Democrats (Eveland & Shah, 2003). Therefore, by understanding dissimilarity between groups

of people who share different political characteristics in different political boundaries, we are

able predict media effects with a better explanatory power. Thus, this study particularly looks at

the dynamics between the amount of political information availability and possible political

perspectives of individuals across different political boundaries.

Demographics and Media Geopolitics

Political audiences exist in different geographical spaces, and this factor interacts with

various individuals' predispositions such as gender, social context, and subculture (McQuail,

1997, pp.50-52). Where an individual belongs is a crucial indicator in predicting political

behavior. According to USA TODAY's survey results in the 2004 presidential election, party

identification and ethnicity as well as previous voting behavior were important factors for

potential political behaviors. In fact, 89% of Democrats voted for Kerry, and 93% of

Republicans voted for Bush. White voters (57%) selected Bush while Hispanics (55%) and

Asians (57%) voted for Kerry. Considering 90% of people who voted for a certain party in the

previous presidential election voted for the same party for the 2004 presidential election, political

demographics should be considered, or at least be controlled, to understand media geopolitics

(USA Today, 2004).





























To all who nurtured my intellectual curiosity, academic interests, and sense of scholarship
throughout my lifetime, making this milestone possible










composition of sub-populations of gender (X2=21.445, p<.001), ethnicity (X2=1 10.493, p<.001),

and partisans (X2=8.749, p<.013) between information rich and poor states. In other words, there

were higher proportion of female, ethnic maj ority, and political Independents in battle-ground,

information rich states. In order to avoid statistical noise from these sub-population variations, I

controlled for them in each set of main analyses.

In the first set of analyses, individuals' political attitudes, such as political cynicism,

information efficacy, candidate evaluations, and issue awareness, were compared in battleground

and non-battleground states. In the second set of analyses, individuals' changes in their political

cynicism, information efficacy, candidate evaluations, issue awareness, perception on media

bias, and voting intention were compared by determining equally given sets of political

advertising and debates to see how people interpret the same message in various ways based on

their preexisting information levels, controlling for all demographic variables and political

interests and attentions.









CHAPTER 1
INTTRODUCTION

Geopolitics

The Republican candidates' red swept across the South, the Great Plains, and most of the

Rocky Mountain West, while the Democratic candidates' blue covered almost all of New

England, the Mid-Atlantic, and the West Coast in the 2004 presidential election (Marlantes,

2004). There are various scholarly approaches to exploring the causes of the divided nation. A

number of scholars explain the red and blue division through the preexisting bipartisan political

system. Other people argue that it is because of the voters' strategic political tendency toward

balancing a two party system. Some other scholars argue that it is the sum of local political

characteristics and decisions (Burden & Kimball, 2000, p.13).

Although different scholars argue with different explanations, they admit that territorial

and geographical units are the basic concept of societal and political transmission. Geographical

boundaries determine characteristics of members within the group, societal form, and functional

capability based on given economic, sociological, or political assets and the patterns (Arensberg,

1961, p.248). Especially, during campaigning season, hundreds of millions of dollars are

poured into political communication; however, these amounts of money are spent unevenly

across different political regions (Ansolabehere & lyengar, 1995). Current scholars have started

to realize the implication of geographic location as one of crucial variable to measure political

phenomena (Miron & Bryant, 2007, p.399).

Political ideology and political region of a physical boundary are interchangeable in U.S.

politics. "As long as political boundaries remain fixed and population movements are modest,

we expect that the political characteristics will stay nearly constant .. and it is also unlikely

voters' preferences [will] shift markedly between elections" (Burden & Kimball, 2000, pp.39,










Won-Ho Park, Dr. Chang-Hoan Cho, Dr. YooJin Choi, and Dr. Hyojin Kim, have always offered

to lend their hands to me. I also would like to give my special thanks to Dr. Joo Myung Song,

Dr. Do Kyung Ahn, and Dr. Yoo Hyang Kim who have shared their hearts even giving me 'luck

money' for my job interviews. In addition, Dr. Richard Scher and Dr. Badredine Arfi's

compliments and support on my teaching experience have built me up with positive confidence,

thus make me be able to overcome all unnecessary self-consciousness as an international student

My great colleagues and best friends, David Conklin, Dr. Monica Postelnicu, Dr. Jun Soo

Lim, Dr. Seung Eun Lee, Dr. Jong Hoon Lee, Michele Kim, Dr. Hyung Koo Kang, Dr. Eyun

Jung Ki, Soo Yeon Km, Hyung Suk Lee, Dr. Nadia Ramutar, Sarah Urriste, and Shari Kwon,

deserve to have my whole heart. I was able to stay human due to their sense of humor, faith,

love, and trust. In addition, I would like to thank to Mrs. Jody Hedge, Mrs. Sue Lawless-

Yanchisin, and Mrs. Debbie Wallen who have been my personal life advisers beyond academia

by giving me important information and sharing our stories. They always shared their great

smiles and open-minds with me. They are the core people for the programs of Journalism &

Mass Communications and Political science. They made all this process easier, smoother, and

enjoyable. My special thanks also goes to Gordon Tapper, who has a wonderful heart and

cultural insights. Sharing my cultural background, he has observed my pattern of English

speaking to guide me in how a foreign speaker at the advanced level can reach to the level of

native English speaker, even dedicating his free Friday for me.

Lastly, I would like to express my special thanks to my parents. My daddy, who is very

protective and would do anything he can do for me, thus it is not easy for him in the beginning to

let me go away even for an academic program, has been an important reason for me to be a good

person. My mom has dedicated every single breath to me and has been the best supporter for my









the level of information (Erikson & Tedin, 2007, pp.87-8), and strong partisans tend to believe

that they have higher levels of political information than politically independent people

(Coleman, 1996).

Ethnicity was another important interacting factor. Ethnic minorities who are often

isolated from the mainstream media coverage and political attention tend to be more vulnerable

to newly incoming information (Erikson & Tedin, 2007, pp.204-5). Therefore, ethnic minorities,

such as Asians, Hispanics, Blacks, and other mixed groups in information poor states changed

their attitudes more significantly after exposure to new political information compared to ethnic

maj ority or other ethnic minority in information rich states. It tells us that it would be more

efficient for politicians and campaigners to provide campaigning information for minority people

in low of political information environments.

Cynicism

As predicted, the factor of information availability determined individuals' levels of

political cynicism. Individuals in information poor states were more cynical than those in

political information rich states regardless of other demographic factors. Less informed

individuals tend to be cynical than individuals who have access to multiple political information

(Rosenberg, 1955). Frequent information exposures decease citizens' cynical attitudes and

increase trust toward general political process and outcomes (Berman, 1997, p. 111; Cappella &

Jamieson, 1997, p.83). Therefore, one of the best strategies to reduce public cynicism is to

provide "persistent, divers and consistent information campaigns" (Berman, 1997, p.106).

However, information availability still interacted with other demographic factors, such as

party identification, in determining the cynicism level. Interestingly, political Independents

tended to be more cynical than Democrats and Republicans in both information poor and rich

states and Independents in information poor states were more likely to be cynical than









CHAPTER.5
DISCUSSION/ INTERPRETATION

My research examined the co-effects of media information availability and political

predisposition within different geopolitical boundaries in predicting individuals' political

attitudes and media perceptions. Individuals' previous beliefs or attitude and external political

information circumstances co-affect their political attitudes, preferences, and changes in their

previously existing political characteristics. As expected, individuals' cynical attitudes and

perception of media bias were more prevalent in information poor states, and the level of

political information efficacy for strong partisans was much higher in information rich states.

Geopolitical constraint over information availability or accessibility even influences candidate

evaluations, issue awareness, and voting intentions. High information availability increased

positive evaluations and voting intention for the challenger and more comprehensive issue

awareness. Various demographic factors were also influenced by information availability in

influencing political attitudes and preferences. Certain demographic groups, such as political

Democrats and Independents in battleground states with high levels of information availability

and ethnic minorities in non-battleground states with low campaigning information, were more

vulnerable to newly received political information. The implication of each finding will be

discussed in the following sections.

Political Information Efficacy

When the relationships between information availability and political information

efficacy were measured, different levels of information availability in batter- and non-battle

ground states solely did not influence individuals' political information efficacy and the changes

after direct information exposure. However, the combined effect with party identification, one of

the strong indicators in political characteristics of individuals, differentiated individuals'





Table 4-1. Information Availability & Party Affiliation on Political Information Efficacy
Campaign Attention Indexa Party IDb Mean (SD)
Low Attention States Democrat 3.53 (1.03)
Independent 3.41 (0.96)
Republican 3.48 (1.00)
High Attention States Democrat 3.49 (0.99)
Independent 3.14 (1.08)
Republican 3.66 (0.98)
a. Not significant, a~b. F[2, 2708]=7.635, P<0.001, c. Min=1, Max=5

Table 4-2. Information Availability & Ethnicity on Changes in Political Information Efficacy
Campaign Attention Indexa Ethnicityb Mean (SD)
Low Attention States Minority 0.25 (0.56)
Maj ority 0.16 (0.46)
High Attention States Minority 0.18 (0.46)
Maj ority 0.19 (0.46)
a. Not significant, a~b. F[1, 2686]=3.756, P<.053

Table 4-3. Information Availability & Party Affiliation on Political Cynicism
Campaign Attention Indexa Party IDb Mean (SD)


Low Attention States Democrat 3.24 (0.59)
Independent 3.32 (0.57)
Republican 3.16 (0.61)
total 3.22 (0.60)
High Attention States Democrat 3.11 (0.63)
Independent 3.19 (0.58)
Republican 2.92 (0.61)
total 3.05 (0.62)
a. F 1, 2620] =12.237, P10.001, a~b. F 2, 2620] =4.038, P10.018,
c. Min=1, Max=5


Table 4-4. Information availability & Party affiliation on Changes in Political Cynicism
Campaign Attention Indexa Party IDb Mean(SD)
Low Attention States Democrat -0.01 (0.40)
Independent 0.01 (0.39)
Republican -0.02 (0.39)
High Attention States Democrat -0.05 (0.44)
Independent -0.05 (0.39)
Republican -0.02 (0.42)
a. Not significant, a~b. F[2, 2580]=3.076, P<.046










political outcomes throughout American presidential elections that individuals having higher

level of political information were more likely to support the challengers, but people with lower

levels of information tended to support incumbents who were often better known. Incumbents

are often well known to even politically ignorant people through multiple media coverage, but at

the same time, those incumbents are also more vulnerable to negative attacks on their previous

political history. Therefore, more negative issue ads targeting incumbents' failing policies and

other mistakes are prevalent in media coverage. The campaigning of the 2004 presidential

election repeated the tendency (Kaid, et al, 2007). As a result, individuals who obtained more

political information could be more likely to be exposed to negative messages about Bush, thus

evaluating him more negatively and less likely to vote for him than individuals who were in low

political information states. This tendency is more apparent for ethnic minorities who were

originally more political Democratic and isolated from mainstream information (Jasperson &

Yun, 2007).

Interestingly, individuals did not change their feeling thermometers toward both

candidates, Bush & Kerry, by a series of political information; however, the information

exposure changed their perceptions on candidate personal traits, especially for challenger.

Individuals in battleground states with high levels of information were more likely to increase

positive perception about Kerry's personal traits than people in information poor states. This

result is well supported by previous research arguing that the political audience tend to learn

about the candidates' personal traits and issue, but they are less likely to change their attitudes

toward candidates by campaigning information (Kaid, McKinney, & Tedesco, 2007; Shaw,

1999; Kaid & Chanslor, 1995; Funk, 1999; Kenney & Rice 1988;). Moreover, Kerry's advanced

















Cynicism


3.ao-( campaign attention
l ow attention states
op15 attention states


3.20-




m 3.10-




~j3.OO-
L.I



2.90-

Dernocrat Independent Republican

Party ID


Figure 4-3. Political Cynicism





Information Exposure and Changes in Cynicism



0.02- campaign attention
= ow attention states
Po 15 attention states


00--



-0.02 -



-0.04 -



-0.06 -


)-I


-.08

Democrat Independent Republican

Party ID


Figure 4-4. Changes in Political Cynicism










more meaningful and representative of the particular political tendencies and outcomes of the

election year. According to the 2004 presidential election result, a large majority (85%) of

young people declared that they were paying close attention to the campaign, 80% registered to

vote, and 42% actually showed up and voted. These figures were significantly higher among the

college student population, a group that reached a 66% voter turnout rate (Circle, 2004).

However, there could still be possibilities for excessive presenting of cynical attitudes

and insufficient presenting of political information efficacy using young Americans who have

been known as politically ignorant, highly cynical, and highly apathetic throughout American

history (Deli Carpini, 2000; Mann, 1997; Third Millenium, 1999). Thus, this study may have

some limitation to track age variations in different political tendencies and the effects of

information exposure in different political regions.

Measurement

The survey asked a series of questions about participants' demographic factors, such as

age, gender, party ID, race, and political and media tendencies, such as levels of cynicism,

information efficacy, candidate feeling thermometers, perceptions on candidates' personal

attributes, issue awareness, media consumption patterns, media bias, and voting intentions. To

measure individuals' cynical attitudes in a more consistent way, eight different questions

measuring individuals' levels oftrust on government, political leaders, and general political

systems were adopted from reliable previous research, such as the American National Election

Panel Studies (ANES) and many other j oumal articles (Kaid, McKinney, & Tedesco, 2000;

Rosenstone et al, 1997). These eight categories were moderately correlated based on Cronbach's

Alpha reliability scored of .66. Based on the general acceptance level of .70 for Cronbach Alpha


SCronbach's alpha reliability score indicates a coefficient of consistency among different variables. High
Cronbach' s alpha reliability refers to how well various different items can be constructed into a single









Therefore, it is more appropriate to control each individual's previous political information

exposure regardless of channels or sources than controlling particular media consumption

patterns or sum of media time span.

Procedures

Participants were asked to fi11 out a pre-test asking about general political tendencies such

as party ID, ideology, cynicism, information efficacy, and candidate evaluations. Following the

pretest, respondents were then exposed to sets of political advertisements or one of the

presidential TV debates. The ad spots shown consisted of a series of Bush and Kerry spots,

consisting of a mixture of positive and negative spots for each candidate (Appendix B). The first

advertising experiments took place during September 28-30, just prior to the first presidential

debate, and the second advertising experiments were held from October 28-29. Presidential

debates were shown in real time as the events occurred. The participants completed a pre-test

questionnaire and then viewed first, second, and third presidential debates in real time as it

occurred in Miami, Missouri, and Arizona on September 30, October 8, and October 13

respectively .

After viewing, respondents completed post-test questionnaires asking about the same

categories of political characteristics, attitudes, and candidate evaluations. The experimental

setting that allowed both single or multiple information exposure(s) were more appropriate to

reflect the reality of individuals' political information receptions since general audiences were

exposed to some mixture of these political information.

Analyses

The study mainly measured media exposure diversity and the attitudinal diversity by

different geopolitical regions. This study compares two groups of individuals who are able to

directly access multiple campaign information and who can obtain only limited, indirect, or no









Therefore, the political information should be explored in order to determine an

appropriate explanation related to political cynicism. Therefore, my study looks particularly at

the interaction between information accessibility and constraints and the preexisting political

characteristics crossing different political locations.

Two counter-attitudinal measurements of political information efficacy and cynicism are

highly related to the level of individuals' information exposure, information accessibility, and

external information constraints (Cappella & Jamieson, 1997; Kaid, McKinney, & Tedesco,

2000; Shearer, 2005). Therefore, rather than looking at types of media or specific contents to

predict a single or limited political attitude, this research investigates the amount of political

information that is available for individuals in different geopolitical regions and its multiple

functions across contradictory and procedurally different political attitudes.

Candidate Evaluations

Previous research argues that political information exposures through political ads or

debates influence audiences' candidate evaluations. Although there is a mixture of evidence, the

"valence of information influence candidate evaluations" (Funk, 1999, p.701). According to

Kaid, political information exposure, such as political advertising, shapes audiences' perceptions

of targeted candidates' image, inform them about candidates' issues, and influences general

political attitudes about political systems and voting choice (Kaid, 1995). Benoit, McKinney,

and Holbert (2001) argue that exposure to direct political information, such as debates, changes

viewers' evaluations of the candidates and strengthens confidence in preferred candidates, as

well as the factors that determine vote choice. These studies have also found a positive

correlation between political information and candidate evaluations. Exposures to political

advertising, TV debates, and candidate visits increase positive evaluations of political candidates









it is necessary to understand how media make certain political information available for

particular audiences and how those different levels of information availability influence

individuals' political attitudes in order to measure the theoretical connection between selectivity

and given accessibility.

Media distribute information based on the cost-efficiency calculations, and the levels and

types of information are determined by the demands of information providers and preference of

the target audiences (McQuail, 1997, pp. 52-55). One of the most popular strategies of media

planning, presentation, and editorial orientation is to respond to 'a specific social-demographic

category or to a taste of culture' (McQuail, 1997, p. 85). Politicians, media, and audiences want

to maximize their own utilities in the process of political communication. Whereas politicians

wish to gain further support by distributing their agendas, the media want to satisfy information

sources and attract more audiences, and audiences want to obtain the type of information they

need and prefer (Nightungale, 2004, pp. 234-3 5). Therefore, the levels and types of information

in a particular place are determined by these three contributors to the point that meets all

efficiency and preferences. In other words, a different political environment or geographical

location has different information available by the interaction of political interests, media

efficiency, and audience predispositions.

According to an experimental study on restricted versus unrestricted availability of

information, individuals with limited information are more selective than people without such

restriction. "If the information search is restricted, people experience scarcity and increase their

cognitive activity in order to achieve their goal of finding the best pieces of information, either

by testing information more critically or by being more strongly oriented towards information

quality" (Fischer et al., 2005, p.482). Therefore, information quality is very important,










(Eveland & Shah, 2003; Vallone et al., 1985). Strong partisans tend to be more selective for

information and strengthen pro and con attitudes than weak partisans; therefore, the hostile

media phenomena is more prevalent in non-battleground states with limited political information

availability and strong political tendencies (Schmitt, Gunther, & Liebhart, 2004).

Behavior Intention

Interestingly, voters from political information poor states were equally likely to vote for

both Bush and Kerry, while people from information rich states were more likely to vote for

Kerry over Bush. This tendency relies on information flow. With low information flow,

individuals are more likely to be stable in their political choices. Therefore, individuals in non-

battleground states with low political information tend to stick with their previous preference.

However, individuals with high information flow are more likely to be flexible in their political

attitudes (Converse, 1962). Excessive negativity directed at an incumbent could again increase

positive reaction to the opposite challenger (Romero, 1996). And, it leads to anti-candidate

voting from "a desire to vote against one of the two presidential candidates" (Sigelman & Gant,

1989, p.84). Thus, we can infer that political information reinforcement is liable to be translated

into actual political behaviors (Harris, 1989, p.171; Clarke & Acock, 1989).

The level of political information does lead to active participation in election and makes

the American democracy better functional. According to 2004 NES data, 92% of highly

informed voters voted while only 65% of less informed voters voted (Erikson & Tedin, 2007,

p.270-72, & 336). Therefore, higher political information tends to reduce political audiences'

cynical attitudes and behaviors. In addition, geopolitics reflect both media discrimination and

audiences' ideological critics. Thus, political information is the necessary condition for working

democracy .










(F[1, 2686]=6.013, P<.014). Political Independents were more likely to be cynical (M=3.26,

SD=.58) than Democrats (M=3.18, SD=.61) and Republicans (M=3.06, SD=.62). (F [1,

2686] = 16.704, P<.001). People who had higher levels of information about the campaigns

from previous weeks tended to be less cynical than those that had lower levels of previous

political information exposure (t=-13.398, P<.001).

The factor of information availability by different regions had a strong effect on

individuals' political cynicism. Individuals in information poor states (M=3.22, SD=0.60)

tended to be more cynical than those in political information rich states (M=3.05, SD=0.62),

while controlling all individual media consumption, party ID, age, gender, and race (F [1, 2620]

=12.237, P<0.001). Thus, this result confirmed the third hypothesis. At the same time,

information availability is also interacted with other demographic factors, such as party ID, in

determining the cynicism level. Although political Independents tended to be more cynical than

Democrats and Republicans in both information poor and rich states, the Independents in

information poor states (M=3.32, SD=0.57) were more likely to be cynical than Independents in

information rich states (M=3.19, SD=0.58) (F[2, 2620] =4.03 8, P<0.018) (Figure 4-3 & Tables

4-3 & 4-5).

After exposure to a mixture of positive and negative political spots and candidate debates

in an experiment setting, participants, who already consumed certain levels of campaigning

information from the previous week, increased their levels of cynicism more significantly than

people with lower previous campaign information (t=2.772, P<.006). In addition, ethnic

minorities reduced more significantly their previous levels of cynicism (M=-.07, SD=.44)

compared to ethnic maj ority after viewing sets of political information (.M=0, SD=.39) (F[1,

2580] =7.939, P<0.005)










test, it did not have a very satisfactory level of reliability. However, this study still created a

measurement of cynicism on a scale of 5 by averaging those eight variables giving high credits to

previous studies' repeated reliable tests on these categories (Appendix D).

In measuring political information efficacy, four items asking the levels of respondents'

confidence on their political knowledge and information were also adopted from previous

studies, and these items have been asked in the series of ANES (Kaid, McKinney, & Tedesco,

2000; Rosenstone et al, 1997). They are also measured on 5-point agree-disagree scales and

comprised the political information efficacy scale, which achieved a high Cronbach's alpha level

of .85 (Appendix C).

For candidate evaluation measurement, the survey asked respondents to give any number

of degrees between 0 and 100 for each presidential candidate. This measure was adopted from a

"feeling thermometer" traditionally used by ANES studies (Rosenstone et al.,1997). A higher

number is referenced to a more positive evaluation towards the candidate. The measurement of

candidate favorability was also proved as a very reliable indicator for a comparison of relative

evaluations for different candidates (Kaid, 2004b).

Besides overall feeling thermometers toward candidates, the survey also measured

individuals' perception about various personal traits of each candidate. These personal attribute

scales were adopted from Lynda Lee Kaid and her colleagues' series of national election surveys

and have been proved as excellent measurements with multiple reliability tests (Kaid, 1995 &

2004). These scales were well incorporated into this experimental research in order to measure

how each individual in multiple locations with multiple levels of information perceives

candidates' personal attributes in various ways and how much direct information exposure


unidimensional scale. A reliability coefficient of .70 or higher is considered "acceptable" in most social science
research (http://www .ats.ucla. edu/STAT/SPS S/faq/alpha.html).









Political audiences perceive either pure or manifested political information through the media's

view (Schultz, 2000, p.53; Schumpeter, 1993, p.85). Therefore, "ciien ed h edat

provide a broader range of viewpoints, an awareness of the complexity of politics, and a richer

contextual understanding of politics" in order to appropriately grasp current politics (Schultz,

2000, p.68).

However, mass media today do not "deliver a mass audience." There are also particular

media structures that represent "the relatively constant array of channels, choices, and contents"

in different places, given individuals' affinity for political choice, preference, interests, habits,

and expectation (McQuail, 1997, p.67; McDonald, 1990). Politicians tend to "offer piecemeal

bits of policy proposal a little bit for everyone" (Leighley, 2004, p.32). Media themselves select

specific types of contents and narrow the scope of information for local audiences based on the

political interests and partially audience agreements (McDowell & Lee, 2006, pp.179, 185).

Political information seen through the media, thus, is often a set of "differentiated messages that

are more complex and [that] force receivers to search for cues about which messages to believe."

Therefore, as an alternative, individuals often use their preexisting political cues to filter

the external information (Dalton, Beck, & Huckfeldt, 1998, pll2). In addition, individuals are

exposed to different levels of political information and opinion based on individuals' patterns of

social interaction and political predisposition (Huckfeldt, et al., 1995). Therefore, factors like

who you are and where you belong determine the actual level of information exposure and the

ways of perceiving or interpreting given information. For instance, people are much more

vulnerable to political exposure when they do not have preexisting political characteristics, but

individuals who are occupied by strong political beliefs are less likely to be influenced by given

political information. The media may not directly influence individuals' perceptions for those










96). As "Geopolitics" assumes, people in different places have different ways of perceiving,

interpreting, and behaving (Schultz, 2000, pp. 85-87). Value appeals that resonate with

individuals' predispositions are more likely to have influence on political audience as seen in

their attitude and behavior changes (Gordon & Miller, 2004). Differences in societal status in

wealth, occupation, education, and political experience also shape individuals' traits and beliefs

among a particular group in particular places (Leighley, 2004, p.148).

Media Politics

Media and politics interact for special interests especially during campaigning season.

Politicians want to allocate particular information in particular locations. Political parties

intentionally present themselves through media to gain supports from various audiences.

Campaigners strategically choose to access different political market to persuade various types of

national audiences. Media often reflect these demands in the amount of information and

programming and thus provide "an environment where political adverting and propaganda can be

advantageously deployed" (Nightungale, 2004, p. 233).

The media provide citizens both obj ective factual and subj ective commentary information,

and citizens make political decisions based on given information that is more accessible and

available for them. "Information about politics is mediated or learned through the press and we

look to them to inform us." The media have set current agendas, interpret the cause of events,

and predict possible results (Schultz, 2000, pp.15-16). Therefore, "media information is never

pure and simple", especially in politics (Miron & Bryant, 2007, p.392).

There are a significant number of scholarly findings about the dynamics of political

information exchanges between voters and the media. First, political audiences consume

political information through various but accessible media sources. Second, citizens are simple

minded if they do not have a strong preference or ideology when they enter the political arena.










Denton, R. E., & Woodward, G. C. (1990). Political Communication in America. Westport, CT:
Praeger.

Druckman, J. N., & Nelson, K. R. (2003). Framing and deliberation: How Citizens'
Conversations Limit Elite Influence. American Journal of Political Science, 47 (Oct.),
729-745

Eagly, A., & Chaiken, S. (1993). Psychology ofAttitudes. NY: Harcourt, Brace Jovanovich.
Easton, D. (1965). A Systems Analysis of Political Life. NY: John Wiley.

Fairvote (2004). Who picks the president? Takoma Park, MD: The Center for Voting and
Democracy's Presidential Elections Reform Program. Retrieved on March 15, 2005 from
http:.//www.fairvote. org/pre si denti al/

Freedman, P., Franz, M., & Goldstein, K. (2004). Campaign Advertising and Political
Citizenship. American Journal of Political Science, 48 (Oct.), 723-741

Erikson, R. S., & Tedin, K. L. (2007). American Public Opinion: Its Origins, Content, and
Impact. NY: Pearson Longman.

Eveland, W.P., & Shah, D. V. (2003). The impact of individual and interpersonal factors on
perceived news media bias. Political Psychology, 24(1, 101-117

Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press

Fischer, T., Jonas, E., Fray, D., and Schulz-Hardt, S. (2005). Selective exposure to information:
The impact of information limits. European Journal of Social Psychology, 35, 469-492

Fiske, S. T., & Taylor, S. E. (1991). Social Cognition. (2nd Ed.). New York: McGraw-Hill.

Funk, C. L., (1999). Bringing the candidate into models of candidate evolutions. The Journal of
Politics, 61(3), 700-720

Gamson, W. A., & Modigliani, A. (1989). Media Discourse and Public Opinion on Nuclear
Power: A Constructionist Approach. The American Journal of Sociology, 95(1, 1-37.

Garnett, J. L. (1992). Communicating for results in Government: A Strategic approach for public
managers. San Francisco, CA: Jossey-Bass.

Gordon, A., & Miller, J. L. (2004). Values and persuasion during the first Bush-Gore
Presidential debate. Political Communication, 21, 71-92

Gunther, A.C., & Schmitt, K. M. (2004). Mapping boundaries of the hostile media effect.
Journal of Communication, 54, 55-70.

Gussin, P., & Baum, M. (2004). In the Eye of the Beholder: An ExperimentalInvestigation into
the Foundation of the Hostile M~edia Phenomenon. Presented at the American Political
Science Association, Chicago.









political participation (Lane, 1969, p. 152; Almond and Verba, 1963, p257; Dahl, 1961, p. 286;

Campbell et al., 1960, p 105). Specifically, political campaigning information is positively

correlated to political information efficacy, which possibly leads to active voting behaviors

(Clarke & Acock, 1989, pp.551-55). According to another experimental study, both political

advertising and televised debate exposure increase individuals' confidence in their level of

political knowledge (Kaid, Landreville, Postelnicu, and Martin, 2005)

People reported the lack of information about candidates as one of the top reasons for not

participating in politics (Declare Yourself, 2003). Regarding this rationale, this study assumes

that the factors of information variability and accessibility are important indicators that

determine individuals' level of confidence in their political knowledge and capability, and thus

finally leading them to be good citizens.

Political Cynicism

Political Cynicism is disbelief in politics and mistrust of political actors (Levin & Eden,

1962). Contrasted to political information efficacy, political cynicism has negative implication

in democracy. Political cynicism could lead to "lower voter turnout and to a lessening of

information seeking about political issues and candidates" (Kaid, McKinney, & Tedesco, 2000,

p.680). Voters feel powerlessness and insignificance (Miron & Bryant, 2007, p. 395). The level

of cynicism has been quite high in American politics.

Previous scholars have argued about the undesirable phenomenon with different

theoretical perspectives. Cognitively, American publics believe that political participation such

as voting does not make any significant difference (Agger, Goldstein, & Pearl, 1961).

Systemically, even more scholars discuss the unsatisfactory outcomes of the political process as

the cause of American citizens' political cynicism (Levin & Eden, 1962).









individuals' interpretation of issue thus reducing the gap between individuals' diverse attitudes

toward a public issue (Gamson & Modigliani, 1989; Shaw & Maxwell, 1977). The concept of

issue awareness here is consistent with the theoretical arguments of MaCombs and Shaw' s

'agenda setting' (1972) that looked at the interaction between media coverage on issues and

audiences' issue salience. However, rather than investigating the direct correlations between the

amount of media coverage on particular issues and audience's issues, this study intends to look at

how the range of mixture of political information during the campaigning period determines

individuals' perceptions of important issues varied in different political locations.

Such media information interacts with individuals' levels of information in determining

public issue awareness. According to empirical research, the public tend to easily become aware

of currently prominent issues over a relatively short period of time. Instant issue awareness is

more likely to occur among people who have low and medium levels of political information.

Moderately aware citizens are more flexible in their attitudes toward particular political issues

(Koch, 1998, pp.224-27; Zaller, 1992). Citizens' awareness of important political issues has

changed substantially by given political attentions to certain political issues. Therefore, the

relationship between the political elites, the media, and public opinion are more critically

interacted in determining overall issue awareness by given political information than for any

other political consequence. The media's interpretative package of a given range of scope and

symbols on fluctuating and time-sensitive issues, such as the nuclear weapons issue, tend to

reduce different schemas and ambivalence among individuals and induce more homogenous

tendency of issue awareness and simultaneous reactions of people who experience different life

events (Gamson & Modigliani, 1989, pp.3, 33-36).









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

WHAT MEDIA DO TO BI-POLAR, 50-50 STATES

By

Hyun Jung Yun

August 2007

Chair: Lynda Lee Kaid
Major: Mass Communication

As "Geopolitics" assumes, people in different places have different ways of thinking,

personalities, and behaviors because they experience and are exposed to different social and

political environment. In addition, as selective exposure and hostile media phenomena assume,

individuals have a tendency to prefer information corresponding to and avoid information

inconsistent with their own beliefs. For these reasons, individuals perceive even the same

political information differently based on where they are placed.

If individuals are placed in a political battle ground, they are naturally exposed to

multiple sources and types of political information, while other individuals in strong Democrat or

Republican districts would have smaller amounts of campaign information. Liking and disliking

candidates is strongly associated with individuals' political predispositions, but media exposure

tends to reinforce or decrease their political beliefs towards targeting political candidates and

issues they are interested in. Therefore, by knowing the media dynamics on political preferences

among people within certain political boundaries, we are able to predict with a better explanatory

power.

This study particularly looks at the dynamics between the amounts of political

information available and individuals' political attitudes in battle and non-battle ground states.










Hall, S. (1980). Coding and encoding in the television discourse. In S, Hall et al. (eds.),
Culture, media language. London: Hutchinson.

Harris, R. J., (1989). A cogritive Psychology of~a~ss Conanunication. NJ: Lawrence Erlbaum
Associates.

Heesacker, M., Petty, R.E., & Cacioppo, J. T. (1983). Field dependence and attitude change:
Course dependence and attitude change: affecting message-relevant thinking. Journal of
personality and social Psychology, 51, 653-666

Hofstetter, C. R., Shultze, W. A., & Mulvihill, M. M. (1992). Communications media, public
health, and public affaires: Exposure in a multimedia communality. Health
Conanunication, 4, 259-271

Hovland, C. I. (1959). Reconciling conflicting results derived from experimental and survey
studies of attitude change. American Psychologist, 14, 8-17

Hovland, C. I., Campbell, E.H., & Brock, T. C. (1957). The effects of 'commitment' on opinion
change following communication In C. I. Hovland, et al., The order ofpresentation on
persuasion (pp.22-32). New Haven, CT: Yale University Press.

Hovland, C. I., Irving, J., and Kelley, H.H. (1953). Conanunication andPersua~sion. New
Haven, CT: Yale University

Huckfeldt, R., Beck, P. A., Dalton, R. J., & Levine, J. (1995). Political Environments, Cohesive
Social-Groups, and the Communication of Public-Opinion. American Journal of Political
Science, 39(4), 1025-1054.

Hyman, H. H., & Sheatsley, P.B. (1947). Some reason why information campaigns fail. Public
Opinion Quarterly, 11, 412-423.

lyengar, S., (1991). Is Anyone Responsible?: How Television Frames Political Issues. MA:
University of Chicago Press

lyengar, S., Hahn, K., & Prior, M. (2001). Has Technology Made Attention to Political
Ca~npaigns M~ore Selective ? An Experimental Study of the 2000 Presidential Ca~npaign.
Prepared for Presentation at the Annual Meeting of the American Political Science
Association, San Francisco.

lyengar, S., & Kinder, D. R. (1987). News That Matters. Chicago: University of Chicago Press.

Jasperson, A., & Yun, H. J. (2007). Political Advertising Effects and Anerica's Racially Diverse
Newest Voting Generation. American Behavioral Scientist, 50(9), 1112-1123

Kaid, L. L. (1995). Measuring candidate images with semantic differentials. In K. Hacker
(Ed.), Candidate images in Presidential election canspaigns (pp. 131-134).
Westport, CT: Praeger.












TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES ................ ...............9............ ....


LIST OF FIGURES .............. ...............10....


AB S TRAC T ............._. .......... ..............._ 1 1..


CHAPTER


1 INTRODUCTION ................. ...............13.......... ......


Geopolitics ................. ...............13.................
Media Politics ................. ...............14.......... .....


2 LITERATURE REVIEW ................. ...............18................


Selective Exposure and the Dynamics in Political Communication .............. ....................18
Information Accessibility in Polarized Political Communication ................. ............... ....20
Information Trap between Media, Politics, and the Public .............. ...............22....
Hostile Media Phenomenon: Geopolitics and Media ................. ..............................23
At the Micro-Level .............. ...............23....
At the Macro-Level ................. ...............25...

Demographics and Media Geopolitics............... .. .........................2
Dynamics of Political Attitudes in Political Information Availability ................. ................27
Political Information Efficacy .............. ...............29....
Political Cynicism .............. ...............30....
Candidate Evaluations ................. ...............32........ ......
Issue Awareness. ............. ...............33.....

Voting Intention............... ...............3


3 HYPOTHESES & METHOD............... ...............37.


Hypotheses.................. ........ ........3
Political Information Efficacy .............. ...............37....

Cynicism............... ...............37
Candidate Evaluations ................. ...............37.................
Issue Awareness .............. ...............38....
Media Bias ................. ...............3.. 8..............
Behavior Intention ................. ...............3.. 8..............
M ethod ................ ...............3.. 8..............

Purpose .............. ...............3 8....
Samples............... ...............39
Measurement .............. ...............40....









stand for their own perspectives and relevance (Leighley, 2004, p. 143; McQuail, 1997, p. 118).

For that reason, media often reflectively "encourage individuality by offering ideas that are

congenial to a person's self-interests" rather than transmitting all messages they deliver into

random audiences (McDonald, 2004, p. 184).

Individuals have tendencies to expose themselves to messages and retain information that

is consistent with their own beliefs, and they are likely to interpret messages in a way they

already believe (Klapper, 1960). Individuals are more likely to prefer information corresponding

to, and avoid information inconsistent with, their own beliefs and attitudes (Festinger, 1957).

Audiences watch what they like and like what they watch (Barwise & Ehrenberg, 1988).

Therefore, individuals who are exposed to multiple viewpoints presumably select a piece of

information that fits their point of view and react to the same information differently based on

their own interests and perceptions.

Although some scholars argue that there are limited effects of selective exposure in

political attitude or behavior changes (Chaffee; Saphir; Sandvig, &; Hahn, 2001; Klapper, 1960;

Lazarsfeld, Berelson, & Gaudet, 1948), individuals who have different levels of political

information and different selective tendencies have different political perspectives (Klapper,

1960; Milburn, 1979). Moreover, individuals decode the meaning of media messages in order to

fit their own perspectives and values (Hall, 1980). In other words, although political audiences

may not be completely free from unwanted information and block themselves from changes in

their attitudes and behaviors from such information, they have particular tendencies to focus on

political appeals from their preferred candidates or parties. If individuals are exposed to counter-

attitudinal information, they consider the message as distorted information, and they only

remember information that supports their initial political beliefs (Iyengar, Hahn, & Prior, 2001;









APPENDIX F
MEDIA CONSUMPTION

* Level of media exposure of presidential campaign in past week
* Level of talk with other people about presidential campaign in past week



















IMedia EBias


3.70


campaign attention
- = Iow attention states
opp 15 attention states


3.65-




3.60-

35--


*


Dernocrat


Inclepencient

Party ID


I
Republican


Figure 4-7. Media Bias


3.66-


3.64-


3.62-


3.60-


3.58 -


3.56 -


3.54 -


3.52-


campaign attention
a low attention states

Mop 15 attention states


mammmemmm ---


I
Majority


Minority


Minority vs. Majority


Figure 4-8. Media Bias


Media Bias










and peripheral interaction with multiple social properties of individuals (McClelland & Judd,

1993).

One of the critical intervening factors in the relationship between political information

availability in different political regions and political attitudes was party identification. It has

been well documented that the level of political information was strongly related to party

identification in determining individuals' political tendencies (Erikson & Tedin, 2007, pp.216-

17). One of the interesting results is that Democrats and Independents in information rich states

somehow had lower levels of political information efficacy than the same partisans in

information poor states could be interpreted by uneasiness with available information. As issue

ownership assumes, Republicans can easily handle hard issues, and Democrats are perceived as

more capable of dealing with soft issues (Campbell et al. 1960; Petrocik, 1997; Damore, 2004).

Recalling that the 2004 presidential election was a war-related election, political information on

hard issues such as security, war, and terrorism were less familiar to or discomforted people who

lean to the left or distinguish themselves from either side of the political tendencies (Source

Watch, 2005; Layman & Carsey, 2002). Therefore, Democrats and Independents who were

exposed to a great degree of uncomfortable discussions felt less confident of their knowledge

levels on such issues.

In addition, in terms of party ID strength and information efficacy, party identifiers

recognize the levels of political knowledge that are consistent with their political disciplines

(Layman & Carsey, 2002). "Party loyalists .. tend to be interested in and pay attention to

campaigns, follow government and public affairs, possess knowledge about politics, watch

debated between candidates on television, register to vote, and turn out to vote" (Kamieniecki,

1988, p.373). Therefore, there is a higher correlation between strength of partisan affiliation and









Table 4-8. Information Availability, Ethnicity & Gender on Kerry Evaluation


Total 48.35 (29.37)
High Attention States Minority Male 58.48 (27.44)
Female 66.52 (22.77)
Total 63.73 (24.73)
Maj ority Male 48.57 (30.49)
Female 49.32 (32.40)
Total 48.42 (31.65)
Total 51.36 (31.03)
a. F[1, 2596]=6.028, P<.014, a~b. Not Significant, a~b~c F[1,
2596]=7.081, P<.009, c. Min=0, Max=100

Table 4-9. Information Availability, Ethnicity & Gender on Changes in Bush Evaluation
Campaign Attention Indexa Ethnicityb Gendere Mean (SD)
Low Attention States Minority Male 0.48 (15.50)
Female -0.08 (15.12)
Total 0.18 (15.29)
Maj ority Male -0.42 (14.89)
Female -0.12 (12.72)
Total -0.26 (13.22)
Total -0.09 (14.03)
High Attention States Minority Male -2.36 (17.07)
Female 2.09 (13.73)
Total 0.55 (15.08)
Maj ority Male 0.30 (14.06)
Female -1.27 (14.18)
Total -0.77 (14.14)
Total -0.52 (14.33)
a. Not Significant, a~b. Not Significant, a~b~c F[1, 2646]=5.079, P<.024,
c. Min=0, Max=100


Campaign Attention Indexa
Low Attention States


Ethnicityb Gendere
Minority Male
Female
Total
Majority Male
Female
Total


Mean (SD)
56.69 (
56.58 (
56.63 (
39.72 (
46.50 (
43.48 (


25.33)
29.25)
27.41)
28.78)
29.57)
29.40)










Table 4-12. Information Availability on the Range of Issue Awareness
Campaign Attention Index Information Exposure Mean (SD)
Low Attention States Pre 3.12 (1.17)
Post 3.26 (1.22)
High Attention States Pre 3.23 (1.17)
Post 3.20 (1.91)
Pre issue comparison: F [1, 2082]=4.977, P<.026, Post issue comparison: Not significant


=.979, Spearman's rho R (c~d)=.972, Spearman's rho R


Table 4-13. Information Availability on Issue Salience


Low Attention
Rank Prea
1 War
2 Economy
3 Terrorism
4 Education
5 Health
6 Tax
7 Foreign policy
8 Elderly
9 Environment
10 Welfare
11 Crime
12 Children
Spearman's rho R (a~b)


High Attention
Pree
Economy
War
Education
Health
Terrori sm
Foreign policy
Education
Elderly
Environment
Crime
Welfare
Children


Postb
War
Economy
Health
Terrori sm
Education
Tax
Foreign policy
Elderly
Environment
Welfare
Crime
Children


Postd
War
Economy
Health
Education
Terrori sm
Tax
Foreign policy
Elderly
Environment
Welfare
Crime
Children


(a~c)=.958, Spearman's rho R (b~d)=.993


Table 4-14. Information Availability on Media Bias
Campaign Attention Index Mean(SD)
Low Attention States 3.62 (0.64)
High Attention States 3.53 (0.68)
F 1, 2726 -10.613, P<.001










approaches were investigated. In other words, levels of political information accessibility and

availability in each state that constrains the amount of information available for political

audiences within the political regions were considered at the macro level. At the micro level,

individuals' political attitudes and evaluations under different information circumstance and

degrees of attitude changes after exposure to the same set of political information in different

political regions were measured.

A theoretical connection between micro levels of individual attitudes and macro levels of

mass communication was established by looking at how individuals who are in information rich

states differ from individuals who are in information poor states in their political attitudes and

information perceptions. These connections of micro and macro level approaches are recently

recognized by many mass communication scholars and are useful ways to deal with a more

concrete pattern of political communication (Mutz et al., 1996).

Samples

This research used experimental data conducted and collected by the Uvote 2004 research

group, a non-partisan, inter-university research team, during the peak campaigning season of the

2004 presidential election from September to November across 22 different locations in the U.S.

(Note A). Respondents (n=2965) were enrolled at the participating universities in various

courses and voluntarily participated in the study. The average age of participants is 20.78 years.

The sample consists of 43% males and 57% females and about equal numbers of Republicans

(37%) and Democrats (40%), but lesser numbers for Independents/Others (23.2%) in their party

ID.

Since the maj ority of participants were young adults, there could be particular political

tendencies reflective of younger people's political characteristics. Considering the noticeable

increase of young adults' participation in the 2004 presidential election, the sample would be

































O 2007 Hyun Jung Yun









analyses allow researcher to investigate how the dynamics of each layer interact and influence

political consequences in a more transparent way (Steenbergen & Jones, 2002).

Therefore, this study intends to answer the question of "why do individuals who are in

different geopolitical locations perceive and interpret political information variously?" Thus, it

tests the interaction between individuals' political predispositions and political information

availability in different political geography of states as the cause of various political

campaigning effects. For a more concrete finding of the theoretical connection of geopolitics,

we should understand different levels of units not only at individuals and the nation but also at

state and local districts. Such an approach is theoretically logical since election outcomes are

determined by aggregating votes within those geographical electoral units. Voters within each

geographic boundary are critical political units for red vs. blue and battleground vs. non-

battleground state compositions in the U.S. (Wright, 1998). Therefore, this study adopts

geographical regions as a given factor for campaigning information availability that influences

individuals' political attitudes and differentiates the effects of incoming political information.










significant. In other words, the post test correlation is statistically higher than the pre-test

correlation (Z'=-2.2172, P<.001). 4 Therefore, we can infer that political information do change

the perceptions of important policy issues and the sets of political information make individuals'

issue recognition more identical (Table 4-13).

Media Bias

As the hypothesis 10 predicted, information availability determined individuals'

perception about media. After seeing a series of balanced political ads or a candidate TV debate,

individuals who were in information poor states (M=3.62, SD=0.64) were more likely to think

that media were biased compared to individuals who were originally in information rich states

(M=3.53, SD=0.68) (F[1, 2726]=10.613, P<.001) (Figure 4-7 & 4-8 & Table 4-14 & 4-15).

People who have limited previous information with strong political characteristics in non-

battleground states tended to believe that counter-attitudinal information was biased, unfair, and

exaggerated, although the information were correct and balanced. Although there is "no

systematic and pervasive ideological bias in the media", the preexisting media pattern and the

audiences' routine in interpreting media messages deviate the perception of given political

information (Erikson & Tedin, 2007, pp.236-7)

Voting Intentions

When the survey asked participants for whom they planned to vote in the coming 2004

presidential election, participants in different levels of information availability had differing


4 Fisher's z' (called r-to-z transformation) is used for computing confidence intervals on correlation and difference
between correlations. This technique transfers r to z-score in order different sets of correlations to be compared
statistically. In the first step, z scores are calculated based on the formula, Z' = In[|(r+1)/r-1)|]/2 using r scores
(Z'1=1/2 In (1+.958/ 1-.958)=1.921, Z'2=1/2 In (1+.993/ 1-.993)=2.826). In the second step, the standard error of
difference between the two correlations is estimated based on formula, SE = SQRT[(1/(nl 3) + (1/(n2 -
3)].(-.SQRT (.0833+.0833)=.408). In the third step, the difference between two z-scores is divided by the standard
error (Z'=1.921-2.826/.408)=-2.218). Since the absolute z-score is higher 1.96. Therefore, it is statistically
significant at a=.05. In other words, the post test correlation between information rich and poor states is statically
higher than the pre test correlation between information rich and poor states.









whose political identities were previously established, thus less likely to change those previous

attitudes; but, at least in part, media exposure still "stimulates voters' interests and involvement

in issues" (McCombs, 1994, p.9) and interact with audiences' political characteristics in political

communication.

There should be a combined political outcome of geopolitical factors, media, and political

audiences (Mcquail, 1997, p.68). However, the connection between media information, politics,

and the impact on citizens has not been explored with any conceptual clarification (Miron &

Bryant, 2007, p.398). A number of previous scholars have failed to provide a concrete way of

understanding geopolitical patterns either by only taking a snap shot of a certain political event at

a particular time at the state level, looking at particular groups of individuals in a specific

location or by considering one type of media effect on an audience.

For instance, although Benoit and his colleagues (2004) found different audience issue

salience by different levels of information exposure in battleground and non-battleground states,

they have not looked at how media information and audience' political predispositions interact,

thus even the same amount of political information may induce differently individuals' attitude

changes. In addition, the maj ority of research in the macro perspectives of political systems and

governments often ignore media as the primary connection between an individual and the

political system in the political communication processes (Semetko, 2004, p.355).

In order to overcome such limitations, we need to provide a connection between the

macro-level of the bipolar political system and the micro-level of individuals' political

preferences and ideological pattern and then consider mediating factors, such as media, bridging

macro and micro levels of political dynamics (Burden & Kimball, 2000, pp.32, 34). Multilevel









confidence on their political knowledge. For instance, Republicans in information rich states had

higher levels of political information efficacy than Republicans in information poor states, but

Democrats and Independents in information poor states had higher levels of efficacy than other

Democrats and Independents in information rich regions. Overall, strong partisanship increased

the levels of individuals' confidence on their levels of political information.

In addition, after seeing political ads and debates, the geographical factor solely did not

have any statistical significance in changes in political information efficacy. In other words,

although direct political information exposure increased overall audiences' information efficacy,

the differences between the level of change in people who were in information rich and poor

states was very marginal. However, it had an interaction effect with ethnicity. Ethnic minorities

in political information poor states tended to more significantly increase their levels of

information efficacy than any minority in the information rich states and ethnic maj orities in both

information poor and rich states, while controlling other demographic variables

Throughout previous research, individuals' self-perception that they are capable of

understating politics was not easily changeable and often turned out as a given factor in political

communication processes (Rudolph, Gangl, & Stevens, 2000). There could be both theoretical

and methodological reasons for the constant levels of political information efficacy.

Theoretically, individuals' political dispositions and characteristics are established by primary

groups of people, such as family, school, and peer group. These primary groups develop solid

levels of individuals' political knowledge and efficacy; thus, political efficacy is not easily

changeable by temporally available levels of political information (Langton & Karns, 1969,

pp.813-14). Methodologically, experimental studies theorizing stimulus effects have difficulties

in finding statistically significant results because of complicating components of main effects










LIST OF TABLES


Table page

4-1 Information Availability & Party Affiliation on Political Information Efficacy ...............56

4-2 Information Availability & Ethnicity on Changes in Political Information Efficacy........56

4-3 Information Availability & Party Affliation on Political Cynicism .............. .................56

4-4 Information availability & Party affliation on Changes in Political Cynicism.................56

4-5 ANCOVA Tests of Political Information Effcacy and Cynicism............... ................5

4-6 Information Availability, Ethnicity & Gender on Bush Evaluation .............. .................58

4-7 Information Availability & Party Affliations on Kerry Evaluations ............... ...............58

4-8 Information Availability, Ethnicity & Gender on Kerry Evaluation ................ ...............59

4-9 Information Availability, Ethnicity & Gender on Changes in Bush Evaluation ...............59

4-10 ANCOVA Tests of Candidate Evaluations............... ..............6

4-11 Information Availability on Personal Attribute Evaluation............... ...............6

4-12 Information Availability on the Range of Issue Awareness ................ ........... ..........62

4-13 Information Availability on Issue Salience ................. ...............62...............

4-14 Information Availability on Media Bias ................ ...............62...............

4-15 ANCOVA Test of Media Bias ................. ...............63...............

4-16 Information Availability on Pre-Vote Intention .............. ...............63....









Although information availability was not strongly influential on evaluation towards Bush,

individuals who were in information low states (M=51.57, SD=34.32) seemed to evaluate Bush a

little more positively than people in information rich states (M=47.97, SD=34.99) (F[1,

2657]=2.855, P<.091). However, the factor of information availability, once again, interacted

with other demographic factors, such as ethnicity and gender, in influencing candidate

evaluations. Ethnic minorities in information rich states (M=31.97, SD=29.21) tended to

evaluate Bush more negatively than ethnic minorities in information poor states (M=42.34,

SD=32. 15). However, ethnic majorities in either information rich (M=51.77, SD=35.19) or poor

states (M=57. 15, SD=34.41) evaluated Bush in similar degrees (F[1, 2657]=3.742, P<.053)

(Figure 4-5 & Tables 4-6 & 4-8). In addition, males who belonged to ethnic majorities in

information poor states (M=60.61, SD=33.84) evaluated Bush most positively and females who

belonged to ethnic minorities in information rich states tended to have the least positive Bush

evaluations (M=28.03, SD=27.97) compared to other people with different combinations of these

demographic factors (Tables 4-6 & 4-8).

Geographic variances in political information availability significantly influenced

evaluations of Democratic Party candidate, Challenger Kerry. Individuals in the political

information rich states (M=51.36, SD=31.03) were more likely to evaluate Kerry positively than

individuals in political information poor states (M=48.35, SD=29.37), while controlling for other

demographic factors (F[1, 2596]=6.028, P<.014).

In addition, even under unequal information availability, Democrats were always more

supportive of Democratic Party candidate, Kerry. However, the geopolitical factor of

information availability influenced the effect of party ID for candidate support. Democrats in

information rich states (M=73.78, SD=18.72) more positively evaluated Kerry than other









CHAPTER 3
HYPOTHESES & IVETHOD

Hypotheses

Political Information Efficacy

H1: Individuals in the political information rich states tend to have higher levels of political

information efficacy than individuals in the political information poor states, controlling other

demographic variables.

H2: After a series of political information exposures, individuals in information poor states are

more likely to increase their levels of political information efficacy than individuals already in

the information rich states.

Cynicism

H3: Individuals in the political information rich states tend to have lower levels of political

cynicism than individuals in the political information poor states, controlling other demographic

variables.

H4: After a series of political information exposures, individuals in the political information poor

states are more likely to reduce their levels of cynicism than individuals already in information

rich states.

Candidate Evaluations

HS: Different levels of political information availability influence individuals' evaluations of

different candidates.

H6: After a series of political information exposures, individuals in limited political campaign

information states are more likely to increase their positive evaluation towards political

candidates than individuals already in information rich states.





Table 4-5. ANCOVA Tests




Intercept
Political Information
Age
Information Availability
Party ID
Ethnicity
Gender
Information Availability
* Party ID
Information Availability
* Race
Party ID* Race
Information Availability
* Party ID* Race
Information Availability*
Gender
Party ID* Gender
Information Availability*
Party ID *Gender
Race Gender
Information Availability*
race *Gender
Party ID* Race *Gender
Information
Availability*Party
ID*Race *Gender
***p<.01, **p<.05, p<.10


of Political Information Efficacy and Cynicism
Political Information Political Information
Efficacy Efficacy Changes
F(Sig.) F(Sig.)
369.099(.001)*** 104.752(.001)***
1050. 172(.001)*** 40.350(.001)***
16.152(.001)*** 10.389(.001)***
0.886(.347) 0.001(.971)
14.831(.001)*** 0.001(.999)
1.036(.309) 0.903(.342)
81.610(.001)*** 4.264(.039)**


Cynicism
F(Sig.)
3878.160(. 001)***
118.437(.001)***
1.150(.273)
12.237(.001)***
16.704(.001)***
6.013(.014)**
0.644(.422)

4.038 (.018)**

0. 816(. 366)
4.096(.017)**

1.737(. 176)

0.886(.347)
0.205(.815)

0.912(.402)
3.269(.071)*

2.658(.103)
0.342(.710)


1.836(. 160)


Cynicism Changes
F(Sig.)
1.064(.302)
4.221(.040)**
1.217(.270)
1.348(.246)
0.437(.646)
7.939(.005)***
0.550(.459)

3.076(.046)**

1.241(.265)
2.299(.101)

1.701(.183)

0.279(.598)
0.463(.629)

0.413(.662)
0.504(.478)

1.652(.199)
1.237(.290)


1.432(.239)


7.635(.001)***

0.008(.930)
6.037(.002)***

5.122(.006)***

0.191(.662)
1.550(.213)

1.152(.316)
0.085(.771)

0.749(.387)
0.065(.937)


0.691(.501)


1.649(.192)

3.756(.053)*
1.886(.152)

0.815(.443)

0.019(.890)
0.901(.406)

0.386(.680)
0.256(.613)

0.074(.786)
2.121(.120)


2.069(.126)









Therefore, many academic generalizations may not work perfectly for each single individual

(Mcquail, 1997, p.121). For that reason, further research can be strengthened with a more

qualitative approach by investigating personal motivation and life experience.

Although there has been evidence that regional impact on politics has diminished because

of easier mobility, universal media penetration, citizenship delusion and the arrival of new media

era, the long-term trend of distinctive political geographic phenomenon has persisted (Erikson &

Tedin, 2007, pp.216-7). Despite the fact that political audiences immunize themselves from

political information influence from media using personal beliefs (Druckman & Nelson, 2003),

previous research has confirmed that political information exposure influences political

audiences (Erikson & Tedin, 2007, p.251).

Political audiences, politicians, and media are within triangular relations in political

communication processes. Although there are evidence of minimal effects, political information

exposure through TV ads and candidates events do affect political audience (Ansolabehere &

lyengar, 1995). The geopolitical factor of information availability is a crucial predictor for

political attitudes, candidate preferences, issue awareness, and voting intentions. The

relationships between media, geopolitics, and political attitudes at multiple levels of individual,

local, and national needs further scholarly research for more concrete theoretical developments

and sophisticated empirical approaches.









choice for academic life. I was able to grow up in a happy environment due to my wonderful

parents. My final appreciation goes to my soul mate who has shared every single moment with

me for the last eleven years.

I always felt their affection and belief in me. It has given me my endless energy for my

academic progress. Due to them, I have never thought about 'giving up' in any single step of my

academic progress. I know how lucky I am having such wonderful people around me. I deeply

appreciate all of them and promise that I will try to be one who can make them happy.









CHAPTER 2
LITERATURE REVIEW

Selective Exposure and the Dynamics in Political Communication

An audience is a non-scientific social behavioral concept. Audiences develop identities

sharing cultural meaning and generate the meaning in the process of constructing and

reproducing culture, meanings, and ideas. In order to understand the complexity and diversity of

an audience in a "media dominated world" for political information, it is important to understand

"the nature of audience engagement with the practice of everyday media life." For the audience,

"the real-world meaning systems" are generated "where media information is debated, discussed,

and enacted" (Nightungale, 2004, p. 239).

There are multiple media channels and various audience interests. According to the PEW

research center (2004), audiences tend to use different media sources for their political

information. Overall, local TV news (57%), daily newspapers (41%), and radio news (41%) are

the most popular media, but network TV news (32%), the Internet (25%), and talk radio (17%)

are alternative media sources for political information. In terms of audience interests, local or

national political news (22%) ranked 5th place following general community news (31%), crime

(30%), health news (26%), and sports (25%).

Audiences are not very selective in their attention, but they are very "selective in

perception and resistant to unwanted influence" (McQuail, 1997, p.59). With relatively lower

levels of political interests, the audiences are more likely to use convenient sources for political

information search based on their predispositions rather than being exposed to the mainstream

news (Leighley, 2004, p.83). In addition, individuals' political knowledge often does not match

the available information. Audiences realize important issues based on their own interests and

pre-exposures to various life experience; thus, they gather political issues or policy positions that










Lazarsfeld, P.F., Berelson, B., & Guadet, H. (1948). The people 's Choice. NY: Columbia
University Press

Leighley, J.E. (2004). M~ass M~edia and Politics: A Social Science Perspective. Boston, MA:
Houghton Mifflin Company.

Levin, M. B., & Eden, M. (1962). Political Strategy for the Alienated Voter. Public Opinion
Quarterly, 26(1, 47-63.

Lowin, A., (1967). Approach and Avoidance: Alternative Modes of Selective Exposure to
Information. Journal ofPersonality andSocialPsychology, 6, 1-9

Mann, S. (1997). What the survey of American college freshman tells us about their interest in
politics and political science. Political Studies, 32, 262-268.

Marlantes, L. (2004). Inside red-and blue America. Christian Science M~onitor, July 14.

McClelland, G. H., & Judd, C. M. (1993). The statistical difficulties of. detecting interactions
and moderator effects. Psychological Bulletin, 114(2), 376-390.

McDonald, D. G. (1990). Media orientation and television news viewing. Journalism
Quarterly, 67(1, 11-20

McDonald, D. G. (2004). Twentieth-Century Media Effects Research. In John Downing, The
Sage Handbook of2~edia Studies. Thousand Oaks, CA: Sage Publications, Inc.

McCombs, M. (1994). News influence on our pictures of the world. In Jenings Bryant and Dolf
Zillman et al., M~edia Effects Advances in Theory and Research. UK: Lawrence Erlbaum
Associates, Publishers

McCombs, M., & Shaw, D. (1972) The Agenda-Setting Function of Mass Media. Public
Opinion Quarterly, 36(2), 176-187

McDowell, S. D., & Lee, J. (2006). Tracking "Localism" in Television Broadcasting: Utilizing
and Structuring Public Information. In Napoli, P.M., M~edia Diversity andLocalism:
Meaning and Metrics. Mahwah, NJ: Lawrence Erlbaum Associates

McGuire, W. J. (1968). Personality and attitude change: An information processing theory. In A.
G. Greenwald, T. C. Brock, and T. M. Ostrom (Eds.), Psychological foundations of
attitudes (pp. 171-196). San Diego, CA: Academic Press.

McGuire, W. J. (1985). Attitudes and attitude change. In G. Lindzey and E. Aronson (Eds.), The
handbook ofsocial psychology (vol. 2), (pp. 233-246). NY: Random House.

McQuail, D. (1997). Audience Analysis. Thousand Oaks, CA: SAGE Publications

Mendelsohn, H. (1973). Some reasons why information campaigns can succeed. Public
Opinion Quarterly, 37, 50-61









strongly committed" (Harris, 1989, p.170). Of course, such reinforcement is liable to be

translated into actual political behaviors and creates political consequences (Harris, 1989, p.171).

However, due to inconsistent measurement and validity, studies in the field still persist

with a great deal of disagreement (Semetko, 2004, p.359). In addition, limitation in approaches

at the individual level overlooked macro levels of systemic impact and constrains of politics and

media on individuals' attitudes changes. Therefore, this study intends to explore attitudinal

dynamics by different levels of political information bridging micro levels of individual

tendencies and macro levels of geopolitical media structure.

Political Information Efficacy

Political information efficacy refers to how confident an individual is that he or she has

sufficient information to engage in the political process (Kaid, McKinney, & Tedesco, 2007).

This concept is closely related to internal efficacy, which Neimi, Craig, and Mattei (1991,

p.1407) define as "beliefs about one's own competence to understand, and to participate

effectively in, politics." However, political information efficacy differs from internal efficacy, in

that it focuses solely on the voter's confidence in his or her own political knowledge and its

sufficiency to engage the political process (to vote).

In terms of demographic factor relevance, previous research found that voters who are

more engaged in politics and have stronger partisanship tend to have higher levels of political

information efficacy (Kaid, McKinney & Tedesco, 2004, 2007). Efficacy and political

participation are learned through a socialization process based on informal norms and values

and formal information exchange (Easton, 1965; Dennis, 1967, pp25-38).

Moreover, individuals' levels of information confidence are highly related to information

accessibility and exposure. Political information and the discussion of political issues increase

general confidence in individuals' own political capability that is positively related to active










Ferguson and Dr. Lynda Lee Kaid, Uvote inter-university research on U.S. elections under Dr.

Lynda Lee Kaid, and United States Election Assistance Commission' s proj ect establishing

election law database under Dr. Lynda Lee Kaid and Dr. Cliff Jones The former proj ect dealt

with media advocacy, government public information, and issue management on terrorism. The

Uvote research has focused on political advocacy and political information effects. The

development of electronic database of U. S. election laws intended to provide U. S. citizens easy

internet search function for comprehensive U.S. election law. She has worked for these proj ects

as a data analyst and proj ect manager.

She also worked as data archiving assistant for ICPSR (Inter-university Consortium for

Political and Social Research) for a part of graduate assistantship duty. She was trained in

advanced methodology through ICPSR as well as by the departments of statistics, political

science, and journalism and mass communications. Her methodological training across different

fields includes managements of data through various applications and various levels of statistics

such as linear regression, categorical analysis, multivariate analysis, maximum likelihood

analysis, game theory, content analysis, scaling, and measurement.

Her next proj ect is to collect linearly coherent multi-level data that links individual

perceptions, attitudes, and preferences with the aggregate level of media and political

predispositions in different political regions, election turnouts, policy efficiency, and other

media-politics routines in order to conduct research with theoretically reliable connections across

different levels of dynamics.

Hyun Jung Yun who has two doctorate degrees, one in political science and the other in

journalism and communications, will work as an assistant professor at Texas State University

starting from August 2007.










Using experiment data (N=2965) gathered by the Uvote research team during the 2004

presidential election at multiple locations, this study compares two groups of individuals who

have been exposed to multiple campaign information in battle-grounds and who have been

exposed to limited, no, or indirect campaigning information in non-battle grounds to measure

individuals' political attitudes based on political information availability. In addition, this study

also measures how these individuals with various levels of information accessibility evaluate

political candidates, recognize issues, and perceive media messages differently by equally given

sets of political ads and debates. The study found that individuals in battle-ground states with

multiple political information are more likely to have higher levels of information efficacy and

lower levels of cynicism, tend to evaluate candidates positively, more easily change their

previous opinions, and tend to have a broader range of issue awareness than people in non-battle

ground states with limited or no campaigning information, with some variations within

demographic factors.










especially in the situation where limited information is accessible (Fischer et al., 2005, p. 488).

Individuals who have a high orientation for political information, but have limited or indirect

political resources, tend to Eind alternative ways to obtain necessary information such as

interpersonal communication (Mutz & Marin, 2001).

Media information availability interacts with political demographic factors such as

political ideologies rather than one influencing another in communication processes. For

instance, strong partisans are more likely to be selective in their information choice (Lowin,

1967). Republicans are more selective in information filtering and more actively engaged in

assessing previously selected political information (Barlett, Drew, Fahle, & Watts, 1974, p.269).

From views of media, Republicans tend to have more dissimilar political information from

television and newspaper sources than Democrats (Mutz & Martin, 2001, p. 102). Therefore,

political attitudes are more likely to be determined by mixture of individuals' political

characteristics, purposive selectivity, and given information accessibility.

Information Trap between Media, Politics, and the Public

Media are core elements in political communication like an engine for a car. In terms of

functionality, media serve as channels for information delivery, and they also serve as a system

for information exchanges. In terms of effects, some argue that the audience perceives media

information directly without any filtering mechanism. The hypodermic needle model or magic

bullet theory assumes homogenous and significant influence of mass media on the mass public.

According to the assumptions, individuals are passive information receivers. These theoretical

approaches were very appealing in understanding media and audience dynamics during the WWI

and WWII eras, from which there were very limited information sources with highly committed

national political agendas. However, scholars began fairly quickly to recognize limited media










Cronbach's Alpha reliability score of .84 to analyze how individuals in different geopolitical

boundaries perceive the same kind of political information variously (Appendix E).

Demographic factors, such as gender, age, party ID, and ethnicity, and previous

campaign information exposure were controlled in order to examine a more transparent

relationship between audiences' political attitudes and political information availability.

Particularly, past week political information exposure was adopted to control individuals'

amount of previous political information exposure, rather than directly looking at multiple

channel exposures or total amount of media consumption time because my study intends to look

at only the dynamics of political information over overall media consumption patterns.

As we recognized in previous research, political information exposure is somewhat a

troublesome concept in the new and multi-media era. There are different levels of accessibilities

to and desire of political information. Some individuals are unintentionally exposed to political

messages, but some others intentionally search for political information. In other words, an

individual who is very interested in elections in non-battleground states still would look for

political advertising and candidates' agendas through the Internet, other cable channels, national

news, or interpersonal communication. In contrast, there could be a possibility for audiences to

be exposed unintended political spots and candidate speeches due to a great amount of political

information available while turning media channels on. A more problematic issue to track

political information consumption is that different medium cannot be comparable in terms of

amount of political information since the unit of coverage is different and the effects vary.


2 A combined scale with 1) Level of media exposure of presidential campaign in the past week and 2) Level of talk
with other people about presidential campaign in the past week (0 Never 5 A lot). There were no statistical
differences between using the combined scale of the two ways of campaign information exposures and using each
indicator separately as control variables in the main analyses. Therefore, I used the combined indicator for a more
cohesive and parsimonious measurement of the main effect of regional information availability by simplifying other
control variables.









Structurally, some argue that prevalent political cynicism is highly related to the way the

media deal with political events, candidates, and system (Cappella & Jamieson, 1997). Some

suggest political and media structures are alienating the political citizen, and others bring in the

prevailing trend of negative belief about politics as the reason of political cynicism. Political

audiences who are exposed to negative political information, such as negative advertising,

become politically more negative, and positive ads reinforce positive political expectations

(Kaid, McKinney, and Tedesco, 2000). Moreover, distorted, biased, or unfair media information

or coverage about politics turns audiences off from any further attentions to political information

(Ansolabehere, lyengar, Simon, & Valentino, 1994). Therefore, the media's negativity and

malfunction magnifies prevalent cynical attitudes among American citizens.

Recent scholars have focused on political information function as both the cause of and

solution for political cynicism. For instance, while political advertising viewing has mixed

effects on political cynicism based on types of advertising, televised debates tend to significantly

reduce audiences' cynical attitudes (Kaid, Landreville, Postelnicu, & Martin, 2005). Scholarly

research found that political cynicism is curable and treatable by efficient political

communication strategies (Berman, 1997). For instance, 'persistent, diverse, and consistent

information campaigns' are efficient ways to reduce disbelief in politics and increase awareness

of government tasks (Wheeler, 1994; Garnett, 1992; Denton &Woodward, 1990). Political

communication with a broad range of information is necessary to perceive positive political

services, issues, and processes (Stipak, 1977; Berman, 1997, p. 110). Such political information

is provided by various strategies in various ways (Berman, 1997, p.107). Berman's study (1997)

argued that "cities that use frequent information strategies experience less cynicism (p.111l).









APPENDIX C
POLITICAL INFORMATION EFFICACY

* I consider myself well-qualified to participate in politics

* I think I am better informed about government and politics than most people

* I feel that I have a pretty good understanding of the important political issues facing our
country

* If a friend asked me about the presidential election, I feel I would have enough information
to help my friend figure out who to vote for









effects. Rather than viewing media as "propaganda" producers, the researchers in the limited

media effect paradigm started to see various levels of influence on different types of individuals.

The media definitely influence the public by providing factual information about events

occurring in the current world, commentary interpretation about people, obj ects, and issues, and

sets of norms and values that are necessary to sustain societies. Likewise, the public also

influence the media by not only favoring certain information the media provide, but also by

avoiding certain information with which they do not agree. Communications are operated when

both communicators and audiences are balanced and supported by a shared outlook and common

beliefs (Mcquail, 1997, p. 116). Therefore, the media industries try to provide information the

audience prefer to attract rather than being correct, fair, and educational (Leighley, 2004, pp.13-

15). As a result, media, politics, and audiences are trapped by each other in political

communication process.

Hostile Media Phenomenon: Geopolitics and Media

At the Micro-Level

Media information is interpreted variously in different social and political contexts

(Neuman, 1991, p.96). Even when media messages are balanced, individuals believe that

perceived media are biased if the contents are against their own attitudes or beliefs (Eveland &

Shah, 2003; Vallone, et al., 1985). Attitudes are very stable, consequential, and very difficult to

change. Thus, it is very rare that individuals change their attitudes in the course of normal daily

life (Hovland, 1959; Hyman & Sheatsley, 1947). The attitudes of individuals are significantly

affected by their previous experience and emotion to relevant arguments regardless of whether

the true credibility of the source is high or not (Heesacker, Petty, & Cacioppo, 1983).

As a result, the impact of political advertising varies across different cultural and societal

structures (Kaid & Holtz-Bacha, 1995). Individuals' preexisting political preferences mediate










Milburn, M. A. (1979). A Longitudinal Test of the Selective Exposure Hypothesis. Public
Opinion Quarterly, 43(4), 507-517

Miron, D., & Bryant, J. (2007). Mass Media and Voter Turnout. In Preiss et al. (eds.), M\~a;ss
Media Effects Research. Mahwah, NJ: Lawrence Erlbaum Associates.

Mutz, D. C. (1998). Impersonal influence: How perceptions of mass collective affect political
attitudes. Cambridge, UK: Cambridge University Press

Mutz, D. C., & Martin P.S. (2001). Facilitating Communication across Lines of Political
Difference: The Role of Mass Media. The American Political Science Review, 95(1), 97-
114.

Mutz, D.C., Sniderman, P. M., & Brody, R. A. (1996). Political persuasion and attitude change.
Ann Arbor: University of Michigan Press, 1-14.

Nelson, T. E., Oxley, Z. M., & Clawson, R. A. (1997). Toward a psychology of framing effects.
Political Behavior, 19(3), 221-246

Neuman, R.W. (1991). The Future of the M~a;ss Audience. NY: Cambridge University Press

Niemi, R. G., Craig, S. C., & Mattei, F. (1991). Measuring internal political efficacy in the 1988
national election study. American Political Science Review, 85(4), 1407-1413.

Nightungale, V. (2004). Contemporary Television Audiences: Publics, Markets, Communities,
and Fans. In Downing, J. et al. (eds.), The Sage Handbook of2~edia Studies (pp. 227-250).
Thousand Oaks: Sage Publications

Payne, J. G., Golden, J. L., Marlier, J., and Ratzan, S. C. (1989). Perceptions of the 1988
Presidential and Vice Presidential Debates. The American Behavioral Scientist, 32(4),
425-435.

Perloff, R. M., (2002). Political Campaign: Persuasion and its Discounts. In J. P. Dillard & M.
Pfau (Eds.), The persuasion Handbook: Developments in theory and Practice (pp.605-
632). London: SAGE Publications,

Petrocik, J. R. (1997). Issue Ownership in Presidential Elections, with a 1980 Case Study.
American Journal of Political Science, 40(3), 825-850.

Petty R. E., and Krosnick, J. A., (1995). Attitude .Soengthrl Antecedents and Consequences.
Mahwah, NJ: Lawrence Erlbaum Associates.


PEW Research Center for the People & the Press. (2006). Public's News Habits Little Changed
by September 11. Retrieved on Nov. 14, 2006 from http://people-
press. org/reports/di splay.php3 ?PagelD=616










SD=7.82) after controlled all demographic factors such as previous political information

exposure, age, gender, party ID, and ethnicity (F[1, 2636]=7.215, P<.007) (Table 4-11).

Issue Awareness

As hypothesis 8 posited, individuals who were in information rich states (M=3.23,

SD=1.17) tended to be aware of broader range of important issues than people from information

poor states (M=3.12, SD=1.17) (F[1, 2082]=4.977, P<.026). However, as hypothesis 9

predicted, exposures to a series of political information messages reduced the gap in policy issue

awareness between those individuals from information rich and poor states; thus, the different

levels of issue awareness faded away (F [1, 2294]=.278, P<.598) (Table 4-12).

Regardless of the regional difference in political information availability, during the 2004

presidential campaigning period, people believed that "war" and "economy" were the most

important issues the country faced. However, there were some variations in importance of each

policy issue according to different regions. The rank order correlation between low and high

information regions was .958. However, after exposure to the same sets of political information,

the correlation between issue importance among individuals from information rich and poor

states became .993. In order to see whether there is a statistical difference between two different

correlations of the pre correlation between information rich and poor states and of the post

correlation between information rich and poor states, I adopted a technique of Fisher' s r-to-z

transformation that overcomes non-normality independent correlation problems (Blalock, 1972,

406-407). The combination of Spearman's rho rank order correlation tests and Fishers' r-to-z

transformation test proved that a series of political information exposures made issue importance

become more identical among individuals from limited political campaign information states and

from information rich states. According to Fishers' r-to-z transformation statistical test, the

difference between the .958 on the pre-test and the .993 on the post-test was proved to be












Procedures .............. ...............44....

A nalyses .............. ...............44....


4 RE SULT S .............. ...............47....


Political Information Effieacy .............. ...............47....
Political Cynicism .............. ...............48....
Candidate Evaluations ................. ...............50.................
Issue Awareness .............. ...............53....
M edia Bias ................. ...............54.................

Voting Intentions ................. ...............54.................


5 DISCUS SION/ INTERPRETATION................ ............6


Political Information Effieacy .............. ...............68....

Cynicism............... ...............71
Candidate Evaluations ................. ...............72.................
Issue Awareness .............. ...............74....
M edia Bias ................. ...............74.................
Behavior Intention ................. ...............75.................

Limitation/ Importance ........._.__....... .__. ...............76....


APPENDIX


A UNIVERSITIES AT WHICH DATA WERE COLLECTED................. ..............7


B SPOTS USED IN THE AD EXPERIMENT IN THE FOLLOWING ORDER ....................80


First Ad s Spots ................. ...............80........... ....
Second Ads sSpots .............. ...............8 0....


C POLITICAL INFORMATION EFFICACY .............. ...............82....


D POLITICAL CYNICISM ................. ...............83................


E MEDIA BIAS .............. ...............84....


F MEDIA CONSUMPTION .............. ...............85....


LIST OF REFERENCES ................. ...............86................


BIOGRAPHICAL SKETCH .............. ...............95....









APPENDIX D
POLITICAL CYNICISM

* Vote has no influence on what politicians do
* One never knows what politicians really think
* People like me don't have any say about what government does
* Politicians and government seem complicated for a person like me
* One can be confident politicians will always do right thing
* Politicians often forget election promises after campaign
* Politicians are more interested in power than what people think
* One cannot always trust what politicians say










ACKNOWLEDGMENTS

Completing two doctoral programs has been a fabulous life j ourney for me. I have gone

through exciting, enjoyable, surprising, blissful, but sometimes stressful, depressing, and

heartbreaking moments. I believe that I am very lucky since I have been surrounded by amazing

people for the last seven years of this j ourney. I would like to express my gratitude to all of the

people who gave me endless support, encouragement, and love from the beginning of my

graduate program through to the point of completion of my two dissertations, one in Political

Science and the other in Journalism and Communications.

First of all, I want to thank my life time advisers and chairs, Dr. Lynda Lee Kaid and Dr.

David Hedge. Dr. Kaid has always stimulated me with research ideas through a weekly research

meeting throughout my doctoral program and taught me invaluable lesson of how the research is

supposed to be processed. Dr. Hedge has guided me and corrected me whenever I need any

direction to go and added his sweetness to my journey with incredible encouragement and

compliments. I also deeply appreciate that Dr. Mary Ann Ferguson, Dr. Michael Martinez, and

Dr. Renee Johnson have been always there for me and given me precious advice at the right

moments. Dr. Spiro Kiousis and Dr Lynn Leverty's endless support for my academic projects

has accelerated my academic progress. My interdisciplinary research was able to be completed

due to all of my committee members' special academic expertise and emotional support and

dedication.

I would also likely to express my special thanks to my mentors, Dr. Goran Hyden, Dr.

Seung Ik Yoo, Dr. Mi Kyung Jin, Dr. Chul Whan Kim, Dr. Sun Joo Yoon, Dr. Sung Hwa Yoon,

Dr. Jun Han Kim, Dr. Soo Bok Lee, and Dr. Kyung Ho Lee since my undergraduate program in

Korea. Their priceless lessons and guidance fueled my academic eagerness in the early years of

my university education. Moreover, Korean faculty members at the University of Florida, Dr.









In order to understand the field of political communication, we need to deal with

interactive dynamics of media, political structure, and audiences' predispositions. As my study

observed, there are multiple interactions in the geopolitical phenomenon. As scholars, such as

Gerbner, Gross, Morgen, and Signiorelli, examined about the relationship between the amount of

TV viewing and political attitudes, overall individuals who were exposed to heavy amounts of

TV viewing were more likely to consider themselves as politically moderates, while the light

viewers tend to be either strong conservative or liberal opinion holders. Therefore, there was a

more significant gap between individuals who were exposed to limited amount of political

information compared to those who were exposed to large amount of information in different

political locations (Harris, 1989, p. 171). However, we also need to aware of facts that higher

media exposure tends to have lower quality of attention. With heavy viewing, audiences are less

alert, and the experiences are less rewarding (Kubey &, Csikszentmihalyi, 1991).

Limitation/ Importance

This research intended to build a theoretical connection between the micro-level of

individuals' political attitudes and the macro-level of the media and political structures.

Although this study showed evidence that political information levels "depend considerably on

the intensity of the campaign and the flow of information" and determine degrees of audiences'

political attitude changes, it has its own empirical and theoretical limitations.

Empirically, despite the broad range of samples across twenty two different locations in

the U.S., the samples consisted of voluntary participants in a university setting. In other words,

the data were not gathered through a random sampling of all possible sub-populations.

Therefore, the possibility exists that the results over- or under-represent certain sub-populations.

In addition, in the index scaling process, although all categories I adopted were repeatedly

proven as reliable indicators by national survey data such as NES, thus legitimizing their use,










p.586). Furthermore, individuals with less information tend to retrieve a piece of information

from "information retained from the past," but individuals with higher levels of information

"operate with a large storage of political lore;" thus, different levels of political information

differentiate individuals' political decision making processes in various degrees (Converse, 1962,

p.583).

Information processing is always involved in an attitudinal decision (Bassili, 1993, p.55)

Therefore, "if the flow of information is strong, no particular prediction is safe" because diverse

information inputs have a great potential for any direction of changes for voters, but "if the flow

of information is weak, the vote will be a pure party vote" because there are few chances of

defections and voting indecisiveness (Converse, 1962, p.586-87). Therefore, "the more remote

the respondent was from the flow of information, the more stable his vote intention" (Converse,

1962, p.590). The lack of information is directly related to decrease of the probability of voting

in presidential elections (Bartels, 1996). Overall, we can infer that although the volume of

information cannot fully determine voting tendency; at least, it magnifies or oscillates floating

voting tendencies (Converse, 1962, p.591).









In addition, people's media reach and the interaction with diverse individual traits caused

different political consequences. People tend to get certain types of information from particular

media and filter the given media information based on their various tastes, beliefs, and

preferences. Hence, although national audiences are heterogeneous, there are stable patterns of

media interaction with certain types of audiences (McQuail, 1997, p.54-55).

Dynamics of Political Attitudes in Political Information Availability

Attitudes refer to persistence, resistance, impact on information processing and judgments,

and guiding behavior (Petty & Krosnick, 1995, p.4). The media's effect can be cognitive,

attitudinal, or behavioral. Nonetheless, it is difficult to track when audiences shift across these

three different consequential lines. Thus, this study addresses only attitudinal changes resulting

from media information. Previous scholarly research has already focused on individuals'

changes in political attitudes by direct media information exposure; however, they concentrated

on psychological factors at the micro level of the individual, ignoring external factors such as

surrounding political mood and patterns of information distribution or vice versa. Therefore,

results were somewhat inconsistent based on what factor each particular study looked at.

Converse (1964) argued that citizens rarely or randomly change their attitudes when they

have lower levels of belief and information exposure, but they tend not to, or only systemically,

change their attitude if they have strong beliefs with high levels of information exposure.

Individuals with middle levels of beliefs are most likely to change their attitudes by given

information (Converse, 1964). There would be "perfect stability where information input is zero

.. .Beyond this minimum, of course stability would increase as a function of involvement"

(Converse, 1962, p.587).

Similarly, Zaller (1991 & 1992) argued a nonlinear relationship between attitude changes

and political awareness. According to Zaller (1991 & 1992), citizens with intermediate levels of

































Table 4-16. Information Availability on Pre-Vote Intention
Kerry Bush Undecided Other
Low Attention States 43.98% 44.22% 9.41% 2.40%
High Attention States 50.67% 37.19% 9.91% 2.22%
X2= 15.765, P<.001


Table 4-15. ANCOVA Test of Media Bias
Media Bias
Intercept
Media consumption
Age
Information Availability
Party ID
Ethnicity
Sex
Campaign Attention* Party ID
Campaign Attention* Race
Party ID* Race
Information Availability Party ID* Race
Information Availability* Gender
Party ID* Gender
Information Availability* Party ID *Gender
Race Gender
Information Availability* race *Gender
Party ID* Race *Gender
Information Availability*Party ID*Race *Gender


F( Sig.)
2492.537(.001)***
10.749(.001)***
24.100(.001)***
10.613(.001)***
2.596(.075)*
0.002(.965)
11.420(.001)***
0.148(.863)
0.020(.887)
2.353(.095)*
0.159(.853)
0.637(.425)
0.407(.665)
0.211(.810)
0.505(.477)
0.165(.684)
0.342(711)
0.101(.904)









After viewing a series of political ads and debates, younger people were more likely to

increase their levels of information efficacy (t=-3.590, P<.001), and females (M=0.21, SD=0.51)

were more likely to increase their confidence in their political knowledge than males (M=0.16,

SD=.45) (F[1, 2686]=4.264, P<.039). However, individuals who consumed more political

information in the previous week were less likely increased their confidence in their political

knowledge by a given set of political information compared to people who had previously

consumed less campaign information (t=-7.240, P<.001)

As the second hypothesis predicted, exposure to political campaigning information

changed individuals' previous political attitudes differently according to the geographical factor

of political information availability and other demographic factors. In terms of information

confidence, although the geographical factor solely did not have any statistical significance in

changes in political information efficacy, it had an interaction effect with other demographic

factors. After seeing political ads and debates, different ethnic groups in different levels of

political information availability changed their levels of political information efficacy variously.

Ethnic minorities in political information poor states (M=0.25, SD=0.56) tended to more

significantly increase their levels of information efficacy than any minority in the information

rich states (M=0. 18, SD=0.46) and ethnic maj orities in both information poor (M=0. 16,

SD=0.46) and rich states (M=0.19, SD=0.46), while controlling other demographic variables

(F[1, 2686]=3.756, P<.053). Thus, we can infer that political ads and campaigning messages

could be more influential for ethnic minorities who do not have enough political information in

non-battleground states (Figure 4-2 & Tables 4-2 & 4-5).

Political Cynicism

In general, ethnic minorities, such as Blacks, Asians, and Hispanics, (M=3.27, SD=.58)

tended to be more cynical than those belonging to the ethnic maj ority, Whites (M=3.09, SD=.61)










H7: After a series of political information exposures, individuals in information rich states are

more likely to reinforce their perceptions of candidates' personal traits than individuals in the

information poor states.

Issue Awareness

H8: Individuals in the political information rich states tend to recognize larger numbers of

important issues than individuals in the political information poor states, controlling for other

demographic variables.

H19: After a series of political information exposures, important issues become more comparable

between individuals in limited political campaign information states and individuals already in

information rich states.

Media Bias

H10: Individuals in information poor states are more likely to think that media are biased than

individuals in information rich states.

Behavior Intention

H1 1: Different levels of political information availability influence individuals' voting intention.

Method

Purpose

There are various concepts in measuring political attitude, such as accessibility,

ambivalence, preference, certainty, elaboration, importance, perception, and knowledge. There

are also various factors that influence attitude formation, including personality, direct personal

experience, family influence, group influence, and interpersonal/small group/mass

communication. In this study, in order to observe variability of individuals' political attitudes

depending on the level of political information individuals can possess, information availability

at the state level and the information effects on individuals bridging two different levels of



















Bush Feeling Thermometer


49.OO -


48.OO -


47.OO -


46.OO -


45.OO -


44.OO -


43.OO -


campaign attention
I Iow attention states
Mop 15 attention states


Z


C


MInrt MoIt
Minority Majorityit


Figure 4-5. Bush Evaluation & Ethnicity










Kerry Feeling Thermometer


70.OO -



60.OO -



50.OO -



40.OO -



30.OO -


campaign attention
- = ow attention states
opp 15 attention states


Democrat Independent Republican

Party ID


Figure 4-6. Kerry Evaluations









voting intentions. Voters from political information poor states were equally likely to vote for

both Bush (44%) and Kerry (44%); however, people from information rich states were more

likely to vote for Kerry (51%) over Bush (37%) (X2= 15.765, P<.001). Therefore, hypothesis 11

on different voting intentions by geopolitical variance of information availability was confirmed

(Table 4-16).










(Kaid, 2004; Shaw, 1999). Individuals who have more information available for their political

decisions are inclined to be supportive of their political leaders (Berman, 1997, p. 110).

In addition, individuals who are less informed about politics more frequently shift their

political position or preference than individuals who have more political information (Converse,

1962). However, at the same time, very detailed political information about candidates' traits

could create more complex ways of processing information, and thus lead to multiple political

interpretations (Funk, 1999, p.715).

Nonetheless, we still need to consider the strength of party affiliations as an important

factor in determining candidate evaluations. Strong partisans are distinctly supporting the same

party candidates across multiple levels of political information availability (Erikson & Tedin,

2007, pp.84, 119, & 278; Funk, 1999, p.715; Rahn, 1993). Political information about

candidates and preexisting political preferences or external political contexts interact in

determining voters' evaluations on candidates (Fiske & Taylor, 1991). "A mixture of information

available" for a given political sector and the level of individuals' political attention co-

determine and allow for the identification of like-minded candidates (Funk, 1999, p.716)

Although some empirical inconsistence exist, such as no evidence of 1976 debate

influence on candidates' image evaluations (Chaffee, 1978), but clear evidence of 1960 debate

on different evaluations toward Nixon and Kennedy (Katz & Feldman, 1962), the influence of

political information exposure has been an important factor and had been at least an interacting

indicator with other demographic in determining candidates evaluations (Kaid, 2000; Payne,

Golden, Marlier, & Ratzan, 1989).

Issue Awareness.

Individuals learn important political issues by given media, the political elite, or people

around them. Particularly, media discourse shapes public awareness of political issues and helps










Rahn, W. M. (1993). The role of partisan stereotypes in information processing about political
candidates. American Journal of Political Science, 37(2), 472-496

Rogers, E. M., & Dearing, J. W. (1994). Agenda-setting research: Where has it been, where is it
going? In Doris A. Graber (3rd Ed.), M~edia Power in Politics. Washington, DC:
Congressional Quarterly Press.

Rokeach, M., Rothman, G. (1965). The principle of belief congruence and the congruity
interaction. Psychological Review, 72, 128-142

Romero, D. W. (1996). The Case of the Missing Reciprocal Influence: Incumbent Reputation
and the Vote. The Journal ofPolitics, 58(4), 1198-1207.

Rosenberg, M. (1955). Some Determinants of Political Apathy. The Public Opinion Quarterly,
18(4), 349-366

Rosenstone, S. J., Kinder, D. R., & Miller, W. E. (1997). American National Election Study,
1996: Pre- andPost-Election Survey (computer file). Ann Arbor, MI: University of
Michigan, Center for Political Studies (producer), Inter-university Consortium for
Political and Social Research (distributor).

Rudolph T. J., Gangl, A., & Stevens, D. (2000). The Effects of Efficacy and Emotions on
Campaign Involvement. The Journal ofPolitic, 62(4), 1189-1197.

Sears, D. O., & Freedman, J. L. (1967). Selective Exposure to Information: A Critical Review.
Public Opinion Quarterly, 31(2), 194-213.

Schmitt, K. M., Gunther, A.C., & Liebhart, J.L. (2004). Why partisans see mass media as
biased. Conanunication Research, 31(6), 623-641.

Schooler, C., Chaffee, S. H., Flora, J. A., & Roser, C. (1998). Health campaign channels:
Tradeoffs among reach, specificity, and impact. Human Conanunication Research, 24(3),
410-432.

Schultz, D. A. (2000). It 's show time!: M~edia, politics, and popular culture. NY: Peter Lang.

Schumpeter, J. (1993). Capitalism, socialism, and democracy. In P.-Green (ed.), Democracy.
City. NJ: Humanities Press.

Semetko, H. A. (2004). Media, Public Opinion, ad Public Action. In J. Downing, et al. (eds.),
The Sage Handbook of2~edia Studies (pp.351-374). Thousand Oaks, CA: Sage
Publications.

Shaw, D. R. (1999). The Effect of TV Ads and Candidate Appearances on Statewide
Presidential Voters, 1988-96. The American Political Science Review, 93(2), 345-361

Shaw, D. L., & McCombs, M. (Eds.). (1977). The emergence ofAnzerican political issues. St.
Paul, MN: West.










changes an individual's perception on candidates' personal attributes. The twelve different pairs

of personal traits on a scale of seven each in qualification, sophistication, honesty, believability,

successfulness, attractiveness, friendliness, sincerity, excitement, aggressiveness, strength, and

activeness were measured. These measures achieved high Cronbach's Alpha scores for Bush's

(.866) and Kerry's (.794) multiple personal attributes; thus, a single indicator of personal traits

was created for each candidate.

Issue attentiveness has also been one important measurement in exploring political

communication effects. In order to measure individual's issue awareness in different levels of

political information availability, the survey asked participants to write down all issues they

believed to be the most important issues the country faced. In the data cleaning process, the

Uvote research team created twelve exhaustive and exclusive issue categories of health, war,

economic, crime, terror, education, environment, elderly, foreign policy, children, tax, and

welfare, and marked all categories each individual indicated. The total numbers of issues in each

individual's list were used to track how an individual who was from an information rich region is

different from an individual from an information poor state in terms of range of issue

recognition. In addition, the ranks of important issues in the information rich and poor regions

were used to explore the aggregate levels of issue awareness in different political information

circumstances.

In order to measure media bias, the survey adopted seven different questions asking

individuals' perceptions on media fairness, obj ectiveness, and normative contribution to the

society. These questions have been asked in multiple national surveys such ANES and GSS for

the last several decades. A measurement of media bias on the scale of 5 was created with









Moreover, newly received information tend to activate "information already at the

recipients' disposal, stored in long-term memory" (Nelson, Oxley, & Clawson, 1997, p.225).

High knowledge levels and personal sophistication on issues determine the levels of attitude

changes by new information. For instance, people who are exposed to previous information are

more likely to be familiar with an issue provided by media; thus, the additional information tend

to be trivial than the effect of new information for individuals with low levels of previous

information (Eagly & Chaiken, 1993; McGuire, 1968 & 1985; Zaller, 1991 & 1992). In other

words, a new piece of information for individuals who were already exposed to previous

information "should not be affected to such a degree" of less informed individuals by the new

information (Nelson, Oxley, & Clawson, 1997, p.228).

Although previous scholars studying agenda-setting and framing argued about the

relationship between the media aspects of issue presentation and the public issue salience

(McCombs and Shaw, 1972; Rogers and Dearing, 1994), these studies overlooked the

importance of individuals' political predisposition and the variations of media information

availability in the different political environments. Therefore, this study investigates the

interaction of individuals' given political characteristics within a political region and the amount

of information accessibility at the regional level in order to observe issue awareness determined

at multiple stages.

Voting Intention.

Political audiences are "sensitive to contextual cues when they make decisions,

formulated judgments, or express opinions" (Iyengar, 1991, p. 11). Individuals are vacillating in

voting decision especially during campaigning periods (Converse, 1962, p.579). A large amount

of political information supplied by media stimulates voters to be more active; however, limited

information gives voters no meaningful difference between different candidates (Converse, 1962,









Milburn, 1979). Individuals' motivation for selective exposure often reinforces the effects of the

political campaign messages that they prefer (Klapper, 1960).

Previous research on selective political campaign exposure can be broadly summarized

into three results. First, short-term media campaigns are inefficient at changing individuals'

preexisting political beliefs. Second, campaigning is primarily a diffusion effect rather than

increasing knowledge levels or changing specific attitudes. Third, various media and diverse

political individuals have dynamic interrelations; thus, research has mixed results (Milburn,

1979, p. 510). However, if researchers understand not only campaigning materials but also

individuals' beliefs, values, mass media habits, and environmental factors such as surrounding

political moods, they would have a better chance to see clearer phenomenon of media and

audience effects (Mendelsohn, 1973).

Information Accessibility in Polarized Political Communication

In an ideal world, political audiences are exposed to various pieces of information through

multiple sources. However, in practical political campaigning with numerous intervening

factors, political audiences are exposed to different types and levels of information. Political

polarization influences local media to be reflective of political interests of politicians and

audiences in their coverage. Media localism induces different types of program selection and

quantity of political information (McDowell & Lee, 2006, p. 177; Nightungale, 2004, p. 234).

Therefore, the "media have been criticized for not offering a wide range of political viewpoints"

(Semetko, 2004, p.356). For example, if an individual is placed in a political battleground, he or

she is naturally exposed to multiple sources of political information while another individual in a

strongly Democrat or Republican district would be exposed to limited and one-sided campaign

messages. Media facilitate and manipulate "communication across lines of political difference"

(Mutz, 1998, p.270), and control the flow of political information (Semetko, 2004, p.356). Thus,









Table 4-10. ANCOVA Tests of Candidate Evaluations
Bush feeling
thermometer
Source F(Sig.)
Intercept 437.780(.001)***
Political Information 5.652(.018)**


4.117(.043)**
2.855(.091)*
798.540(.001)***
11.373(.001)***
1.880(.170)


Kerry feeling
thermometer
F(Sig.)
590.181(.001)***
3.402(.065)*
11.904(.001)***
6.028(.014)**
524.947(.001)***
11.971(.001)***
0.009(.923)

3.010(.049)**

1.970(.161)
6.153(.002)**

1.265(.282)

1.366(.243)
2.647(.071)*

2.415(.090)*
2.860(.091)*

7.081(.008)***
1.555(.211)


1.999(.136)


Changes in Bush
Evolutions
F(Sig.)
3.856(.050)*
3.591(.058)*
1.578(.209)
0.014(.905)
1.006(.366)
0.321(.571)
1.285(.257)

0.263(.769)

0.104(.747)
0.260(.771)

0.290(.749)

0.414(.520)
0.111(.895)

0.748(.474)
3.578(.059)*

5.079(.024)**
0.073(.930)


0.393(.675)


Changes in Kerry
Evaluations
F( Sig.)
0.968(.325)
0.680(.410)
0.257(.612)
0.581(.446)
2.944(.053)*
1.181(.277)
1.010(.315)

0.503(.605)

0.434(.510)
0.119(.888)

0.234(.791)

0.026(.872)
2.469(.085)*

0.727(.483)
1.340(.247)

0.032(.858)
0.906(.404)


0.309(.743)


Age
Information Availability
Party ID
Ethnicity
Gender
Information Availability *
Party ID
Information Availability *
Race
Party ID* Race
o\ Information Availability *
Party ID* Race
Information Availability*
Gender
Party ID* Gender
Information Availability*
Party ID *Gender
Race Gender
Information Availability*
race *Gender
Party ID* Race *Gender
Information
Availability*Party
ID*Race *Gender
***p<.01, **p<.05, p<.10


2.323(.098)*

3.742(.053)*
7.632(.001)***

0.517(.596)

0.090(.765)
6.385(.002)**

2.584(.076)*
0.311(.577)

9.592(.002)***
0.874(.418)


0.373(.689)











LIST OF FIGURES


Figure page

4-1 Interaction of Information availability & Party ID ................ ............... ......... ...64

4-2 Changes in Political information Efficacy ................. ...............64...............

4-3 Political Cynicism ................. ...............65........... ....

4-4 Changes in Political Cynicism ................. ...............65........... ...

4-5 Bush Evaluation & Ethnicity .............. ...............66....

4-6 Kerry Evaluations .............. ...............66....

4-7 M edia Bias .............. ...............67....

4-8 M edia Bias .............. ...............67....











































10




















Information Efficacy


campaign attention
l ow attention states

Mop 15 attention states


3.70-



3.60-



3.50 -

34-





3.30-



3.20 -


dg


Z
L


Dernocrat Independent Republican

Party ID


Figure 4-1. Interaction of Information availability & Party ID


Information Exposure and Changes in Information Efficacy


0.22-



O.20 -

01--



O.18-









O.12-



O.10-


campaign attention
- = = Iow attention states
.op 15 attention states


+


I
Minority


I
Majority


IVinority vs. IVajority


Figure 4-2. Changes in Political information Efficacy






















64









APPENDIX A
UNIVERSITIES AT WHICH DATA WERE COLLECTED

* Consumnes River College
* Dominican University
* Elon University
* Emerson College
* Iowa State University
* New York University
* North East Oklahoma State University.
* Northeastern State University
* Ohio University
* Saint Cloude State University
* Texas A&M
* University of Akron
* University of Colorado at Denver
* University of Florida
* University of Kansas
* University of Missouri
* University of New Haven
* University of North Carolina-Chapel Hill
* University of Oklahoma
* University of South Dakota
* University of Texas at San Antonio
* Virginia Tech










campaign information because of their political geographic constraint. Based on media localism

and different levels of political information reach, this study compares individuals' political

attitudes in battle versus non-battleground states using political information availability.

Adopting the 2004 peak season campaign attention index (Fairvote, 2004)3, individuals in the

top 15 campaign attention states (Iowa, Ohio, New Hampshire, Wisconsin, New Mexico,

Florida, Nevada, Pennsylvania, Minnesota, Colorado, Michigan, Maine, Oregon, West Virginia,

and Missouri) are compared to individuals in the rest of the states where there was low campaign

attention. Those top 15 campaign attention states had 94. 14% of candidate visits, 98.45% of TV

political ads, and 98.48% of campaign-money expenditure during the 2004 presidential

campaigning period (Fairvote, 2004). Besides these top attention states, unfortunately, the rest

of the states obtained almost no campaign attention.

This study's sample consists of some variations of demographic groups in two different

regions of high and low political information. 1694 participants (57. 1%) in the sample were

from information poor states and 1271 participants (42.9%) were from information rich states.

Among individuals in information poor states, 46.5% (n=772) were males and 53.5% (n=889)

were females; 38.5% (n=653) were ethnic minorities and 61.5% (n=1041) were ethnic

maj orities; and 41.2% (n=659) were Democrats, 40.8%(n=653) were Republicans, and 17.9%

(n=287) were Independents. Among individuals in information rich states, 37.9% (n=481) were

males and 62. 1% (n=787) were females; 20.5%(n=261) were ethnic minorities and 79.5%

(n=1010) were ethnic maj orities; lastly, 42.5% (n=5 14) were Democrats, 3 8.8% (n=1088) were

Republicans, and 19.5% (n=546) were Independents. There were statistical differences in the



3 The document was reported by FairVote-The Center for Voting and Democracy's Presidential Elections Reform
Program. The index was calculated by total campaigning funding, total candidate visits, and total TV ads aired in
each state.









WHAT MEDIA DO TO BI-POLAR, 50-50 STATES


By

HYUN JUNG YUN














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007









APPENDIX B
SPOTS USED IN THE AD EXPERIMENT INT THE FOLLOWING ORDER

First Ads Spots

* Bush positive ad Speaks on camera, Laura seats next to him, which child
to pick first on 9/11

* Kerry positive ad I defended this country

* Kerry negative ad 10 million jobs

* Bush negative ad Taxing our economy

* Kerry positive ad Prescription drugs

* Bush negative ad Practical vs. big government

* Kerry negative ad Wrong choices

* RNC negative ad Now Kerry promises (Intelligence budget reforms)

* RNC positive ad Agenda for America

* DNC positive ad Stronger (endorsement by Gen. Merrill McPeak



Second Ads Spots


* Bush negative ad- After September 11 our world has changed.

* Kerry negative ad Clip of Bush from plane carrier with the "mission accomplished
banner"

* Bush positive ad Homeland security and fighting terrorism

* Kerry positive ad Economy (creating j obs, helping small business, tax cuts, cutting down
on dependence on middle east oil)

* Independent positive ad Pro Bush with Ashley's story

* DNC negative ad Bush' s mistakes in Iraq

* Bush negative ad Kerry's flip-flop on Iraq










* Independent (the Media Fund) negative ad Connection between Bush's family and Saudi
Arabian oil business

* Bush negative ad Kerry's issue on healthcare

* DNC negative ad Bush' s healthcare




























j


Table 4-6. Information Availability, Ethnicity & Gender on Bush Evaluation
Campaign Attention Indexa Ethnicityb Gendere Mean (SD)
Low Attention States Minority Male 41.91 (30.68)
Female 42.72 (33.44)
Total 42.34 (32.15)
Maj ority Male 60.61 (33.84)
Female 54.37 (34.64)
Total 57.15 (34.41)
Total 51.57 (34.32)
High Attention States Minority Male 39.25 (30.22)
Female 28.03 (27.97)
Total 31.97 (29.21)
Maj ority Male 49.30 (34.56)
Female 53.33 (35.52)
Total 51.77 (35.19)
Total 47.97 (34.99)


a. F 1, 2657 =2.855, P1.091, a~b. F 1, 2657 =3.742, P1.053,
a~b~c F[1, 2657]=9.592, P<.002, c. Min=0, Max=100


Table 4-7. Information Availability & Party Affiliations on Kerry Evaluations
Campaign Attention Indexa Party IDb Mean(SD)
Low Attention States Democrat 69.41 (20.85)
Independent 50.02 (21.70)
Republican 26.77 (23.35)
Total 48.35 (29.37)
High Attention States Democrat 73.78 (18.72)
Independent 53.01 (23.99)
Republican 23.76 (23.51)
Total 51.36 (31.03)
a. F[1, 2596]=6.028, P1.014, a~b. F[2, 2596]=3.010, P1.049, c. Min=0, Max=100









knowledge are most likely to change their previous attitudes by information. In contrast, people

who have low or high levels of knowledge are more likely to be stable in their attitudes.

Moreover, individuals with higher levels of exposure and interests tend to associate with

more active political evaluations with "a greater ability to interpret, encode, store, and retrieve

new information" (Semetko, 2004, p.361). Benoit et al. (2004) also found that individuals who

live in battleground states had greater levels of understanding on issues and stronger standpoints

compared to people in non-battleground states.

Atkin and Heald (1976) found that voters who were highly exposed to media messages

were somewhat more likely to attach higher agenda priorities. The frequency of political

information exposure was associated with positive candidate evaluations. As expected,

individuals who were exposed to more political media information were more likely to have

higher or more accurate political knowledge. Liking and disliking candidates were strongly

associated with individuals' political predispositions but media exposure tended to help positive

evaluations towards targeting political candidates. Interestingly, individuals who were less

exposed to campaigning political information and have stronger motivation to obtain information

tend to be politically more polarized by attaining increased advertising exposure (Atkin & Heald,

1976, pp.217, 220, & 227).

Although there are great levels of controversy with the effects of content types, timing, and

strategies, political ads definitely influence political audiences. Campaign information obtained

through political ads or candidate visits often "help crystallize existing attitudes by sharpening

and elaborating them" by providing more information and intangible impressions about political

candidates and stimulating emotional reaction. In addition, "political advertising reinforce

existing attitudes in voters to 'keep in the fold' a voter who is leaning toward a candidate but not









Democrats in information low states (M=69.41, SD=20.85). Interestingly, Independents showed

similar tendency. Political Independents in information rich states (M=53.01, SD=23.99)

evaluated Kerry more positively compared to other Independents in information poor states

(M=50.02, SD=21.70) (F[2, 2596]=3.010, P<.049) (Figure 4-7 & Tables 4-7 & 4-8).

Moreover, females who belonged to ethnic minorities in information rich states

(M=66.52, SD=22.77) were more likely to be Kerry supporters and males who belonged to

ethnic maj orities in information poor states (M=39.72, SD=28.78) were less likely to be Kerry

supporters (F[1, 2596]=7.081, P<.008) (Tables 4-8 & 4-10)

However, direct political information exposure had no effect on changes in candidate

evaluations between individuals who were from low and high political information availability

states. Again, however, there were interaction effects regarding information availability and

other demographic factors. After viewing political ads and debates, females who belonged to

ethnic minorities in information rich states (M=2.09, SD=13.73) were most likely to increase

their positive evaluation toward Bush, but males who belonged to ethnic minorities in

information rich regions were most likely to reduce their positive attitudes toward Bush (M=-

2.36, SD=17.07) compared to individuals with some other combinations of those three

demographic factors (F[1, 2646]=5.079, P1.024) (Tables 4-9 & 4-10).


Despite no influence of direct information exposure in changes of individuals' overall

feeling thermometer toward candidates in different regions, the information exposure changed

individuals' perceptions on candidate personal traits especially for challenger, Kerry. When the

survey asked series of personal attributes of Bush and Kerry, individuals in battleground states

with high levels of information (M=1.94, SD=6.91) were more likely to increase positive

perception about Kerry's personal traits than people in information poor states (M=1.18,









CHAPTER 4
RESULTS

Political Information Efficacy

Individuals who were older (t=4.508, p<.001) and consumed more political information

in the previous week (t=33.599, p<.001) had higher levels of political information efficacy. In

general, females (M=3.16, SD=.60) tended to be more confident about their levels of political

knowledge than males (M=3.12, SD=.63) (F[1, 2686]=81.610, P<.001).

On average, individuals' levels of confidence on political information were similar

between information rich states (M=3.53, SD=1.01) and information poor states (M=3.3 8,

SD=1.01), while controlling other demographic and media variables such as campaigning

information attention, gender, party ID, age, and ethnicity. Thus, the first hypothesis was not

confirmed.

When the geopolitical factor was combined with political party identification, however,

the interaction effect had a statistical significance. For instance, Republicans in information rich

states (M=3.66, SD=0.98) tended to have higher levels of political information efficacy than

Republicans in information poor states (M=3.48, SD=1.00). However, political Independents in

information poor states (M=3.41, SD=0.96) had higher levels of political information efficacy

than the Independents in information rich states (M=3.14, SD=1.08) (F[2, 2708]=7.635,

P<0.001 .

More interestingly, individuals who had strong political identification, regardless of their

direction of partisanship, were more likely to have higher levels of confidence on their political

information efficacy than individuals with low or no degrees of political leaning in both

information rich and poor states (F[2, 2708]=14.831, P<0.001) (Figure 4-1, Tables 4-1 & 4-5).











Shearer, E. (2005). Reflections on civic engagement. Nonvoters: Who are America's No-
Shows? In Mitchell S. McKinney, Lynda L. Kaid, Dianne G. Bystrom, and Diana B.
Carlin (Eds.), Conanunicating politics: Engaging the public in democratic life (pp. 87 -
91). New York, NY: Peter Lang Publishing.

Sigelman, L., & Gant, M. M. (1989). Anticandiate voting in the 1984 presidential election.
Political Behavior, 11(1, 8 1-92

Source Watch (2005). U.S. Presidential Election, 2004: Camnpaign Issues. A Proj ect of the
Center for Media and Democracy. Retrieved on June 11, 2007 from
http://www. sourcewatch. org/index.php?title=U. S._presidential_election%2C_2004:_Camp
aignIssues.

Steenbergen, M. R. & Jones, B.S. (2002). Modeling multiple date structures. American Journal
ofPolitical Science, 46(1, 218-237

Stipak, Brian (1977). Attitudes and Belief Systems Concerning Urban Services. Political
Opinion Quarterly, 41(Spring), 41-55.

The Center for Voting and Democracy's Presidential Elections Reform Program (2005). Fairvote.
Retrieved on Feb. 26, 2005 from http://www.fairvote. org/presidential/

Third Millenium (1999, December). Young adults largely ignoring presidential campaign. Press
release. Retrieved on March 10, 2005 from
http://www.vanishingvoter.org/VV2000/PressRlae/22-9shtml.

USA Today. 2004 Election Campaign. Retrieved on Nov. 24, 2005 from
http://www.usatoday .com/news/politicselections/vote2004/NJ.hm

Vallone, R., Ross, L., & Lepper, M. R. 1985. The Hostile Media Phenomenon: Biased
Perception and Perceptions of Media Bias in Coverage of the Beirut Massacre. Journal of
Personality and Social Psychology, 49, 5 77-5 85.

Wheeler, K. M., (Ed.) (1994). Effective Conanunication. Municipal Management Series.
Washington, D.C.: International City Management Association.

Wright, G. C. (1998). Level-of-analysis effects on explanations of voting: The case of the 1982
U.S. Senate elections. British Journal of Political Science, 19, 381-398

Zaller, J. (1991). Information, values and opinion. American Political Science Review, 85(4),
1215-1238

Zaller, J. (1992). The nature and origins of mass opinion. NY: Cambridge University Press.










LIST OF REFERENCES

Agger, R.E., Goldstein, M.N., & Pearl, S.A. (1961). Political cynicism: Measurement and
meaning. The Journal ofPolitics, 23(3), 477-506.

Almond, G. A., & Verba, S. (1963). The civic culture. Princeton: Princeton University Press.

Ansolabehere, S., & lyengar, S. (1995). Going negative: How attack ads shrink and polarize the
electorate. NY: Free Press.

Ansolabehere, S., lyengar, S., Simon, A., & Valentino, N. (1994). Does attack advertising
demobilize the electorate? American Political Science Review, 88(4), 829-838.

Arensberg, C. M. (1961). The community as object and as sample. American Anthropologist,
63(2), 241-264.

Atkin, C., & Heald, G. (1976). Effects of political advertising. The Public Opinion Quarterly,
40(2), 216-228.

Barlett, D. L., Drew, P. B., Fahle, E. G., & Watts, W. A. (1974). Selective exposure to a
Presidential campaign appeal. The Public Opinion Quarterly, 38(2), 264-270.

Barnes, B. E., & Thomson, L. M. (1994). Power to the people (meter): Audience measurement
technology and media specialization. In J.S. Ettema and D.C. Whitney (eds.), Audience
making: How the media create the Audience (pp.75-94). Thousand Oaks, CA: Sage
Publications.

Bartels, L. M. (1996). Uniformed votes: Information effects in presidential elections. American
Journal of Political Science, 40(1, 194-230.

Barwise, P., & Ehrenberg, A. (1988). Television and its audience. London: Sage.

Bassili, J. N. (1993). Response latency versus certainty as indexes of the strength if voting
intentions in a cati survey. Public Opinion Quarterly, 57(1, 54-61.

Benoit, W.L., McKinney, M.S., & Holbert, R.L. (2001). Beyond learning and persona:
Extending the scope of presidential debate effects. Communication M~onoguraphs, 68(3),
259-273.

Benoit, W. L., Hansen, G. J., & Holbert, R. L. (2004). Presidential campaigns and democracy.
M~ass Communication & Society, 7(2), 177-190.

Berman, E. M. (1997). Dealing with cynical citizens. Public Administration Review, 57(2), 105-
112.

Blalock, H. (1972). Socialstatistics. NY: McGraw-Hill.

Burden, B. C., & Kimball, D.C., (2000). Why Americans split their tickets: Camnpaigns,
competition, and divided government. Ann Arbor, MI: The University of Michigan Press.










Campbell, A, Converse P. E., Miller, W. E., & Stokes, D. E. (1960). The American Voter. NY:
John Wiley.

Cappella, J. N., & Jamieson, K. H. (1997). Spiral of cynicism: The press and the public good.
NY: Oxford University Press.

Chaffee, S. (1978). Presidential debates Are they useful to voters? Conanunication
Monographs, 45.

Chaffee, S. H., Saphir, M.N.; Graf, J., Sandvig, C., & Hahn K.S. (2001). Attention to counter-
attitudinal messages in a state election campaign. Political Conanunication, 18, 247-272.

Circle (2004). Youth voting in the 2004 election. Fact Sheet. Retrieved on March 10, 2005 from
http://www. civi cy outh. org/

Clarke, H. D., & Acock, A. C. (1989). National elections and political attitudes: The case of
political efficacy. British Journal of Political Science, 19(4), 551-562.

Coleman, J. J. (1996). Party Organizational Strength and Public Support for Parties. American
Journal of Political Science, 40(3), 805-824.

Converse, P. E. (1962). Information flow and the stability of partisan attitudes. Public Opinion
Quarterly, 26 (4), 578-599.

Converse, P. E. (1964). The nature of belief systems in mass publics. In D.E. Apter (ed.),
Ideology and discontent (pp.206-261). NY: Free Press

Dahl, R. A. (1961). Who Governs? Democracy and Power in an American City. New Haven:
Yale University Press.

Dalton, R. J., Beck, P. A., & Huckfeldt, R. (1998). Partisan cues and the media: information
flows in the 1992 Presidential election. American Political Science Review, 92(1), 111-126.

Damore, D. F. (2004). The Dynamics of Issue Ownership in Presidential Campaigns. Political
Research Quarterly, 57(3), 391-397.

Declare Yourself (2003, November 13). Major national survey shows increased promise, new
strategies for increasing youth vote. Youth vote 2004 survey shows narrow window of
opportunity to engage young voters; points to "political incompetence" as maj or barrier to
voting. Retrieved on June 20, 2005 from
http ://www.declareyourself.org/press/pressroomhm

Deli Carpini, M. X. (2000). Gen.com: youth, civic engagement, and the new information
environment. Political Conanunication, 17, 341-349.

Dennis, J. (1967, March) The child's acquisition of regime norms: political efficacy. American.
Political Science Review.




Full Text

PAGE 1

1 WHAT MEDIA DO TO BI-POLAR, 50-50 STATES By HYUN JUNG YUN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Hyun Jung Yun

PAGE 3

3 To all who nurtured my intellectual curiosity, academic interests, and sense of scholarship throughout my lifetime, maki ng this milestone possible

PAGE 4

4 ACKNOWLEDGMENTS Completing two doctoral programs has been a fabulous life journey for me. I have gone through exciting, enjoyable, su rprising, blissful, but sometim es stressful, depressing, and heartbreaking moments. I beli eve that I am very lucky since I have been surrounded by amazing people for the last seven years of this journey. I would like to express my gratitude to all of the people who gave me endless support, encourag ement, and love from the beginning of my graduate program through to the point of comple tion of my two disserta tions, one in Political Science and the other in Jour nalism and Communications. First of all, I want to thank my life time advisers and chai rs, Dr. Lynda Lee Kaid and Dr. David Hedge. Dr. Kaid has always stimulated me with research ideas th rough a weekly research meeting throughout my doctoral program and taught me invaluable lesson of how the research is supposed to be processed. Dr. Hedge has guide d me and corrected me whenever I need any direction to go and added his sweetness to my journey with incredible encouragement and compliments. I also deeply appreciate that Dr. Mary Ann Ferguson, Dr. Michael Martinez, and Dr. Renee Johnson have been always there for me and given me precious advice at the right moments. Dr. Spiro Kiousis and Dr Lynn Leve rtys endless support for my academic projects has accelerated my academic progress. My interd isciplinary research was able to be completed due to all of my committee members special academic expertise and emotional support and dedication. I would also likely to express my special th anks to my mentors, Dr. Goran Hyden, Dr. Seung Ik Yoo, Dr. Mi Kyung Jin, Dr. Chul Whan Kim, Dr. Sun Joo Yoon, Dr. Sung Hwa Yoon, Dr. Jun Han Kim, Dr. Soo Bok Lee, and Dr. Kyung Ho Lee since my undergraduate program in Korea. Their priceless lessons and guidance fuel ed my academic eagerness in the early years of my university education. Moreover, Korean f aculty members at the University of Florida, Dr.

PAGE 5

5 Won-Ho Park, Dr. Chang-Hoan Cho, Dr. YooJin C hoi, and Dr. Hyojin Kim, have always offered to lend their hands to me. I al so would like to give my special thanks to Dr. Joo Myung Song, Dr. Do Kyung Ahn, and Dr. Yoo Hy ang Kim who have shared their hearts even giving me luck money for my job interviews. In addition, Dr. Richard Scher and Dr. Badredine Arfis compliments and support on my teaching experience have built me up with positive confidence, thus make me be able to overcome all unnecessary self-consciousness as an international student My great colleagues and best friends, David Conklin, Dr. Mo nica Postelnicu, Dr. Jun Soo Lim, Dr. Seung Eun Lee, Dr. Jong Hoon Lee, Michele Kim, Dr. Hyung Koo Kang, Dr. Eyun Jung Ki, Soo Yeon Km, Hyung Suk Lee, Dr. Nadi a Ramutar, Sarah Urriste, and Shari Kwon, deserve to have my whole heart. I was able to stay human due to their sense of humor, faith, love, and trust. In addition, I would like to thank to Mrs. Jody Hedge, Mrs. Sue LawlessYanchisin, and Mrs. Debbie Wallen who have been my personal life advisers beyond academia by giving me important information and sharing our stories. They always shared their great smiles and open-minds with me. They are the core people for the progr ams of Journalism & Mass Communications and Political science. They made all this process easier, smoother, and enjoyable. My special thanks also goes to Gordon Tapper, who has a wonderful heart and cultural insights. Sharing my cultural backgr ound, he has observed my pattern of English speaking to guide me in how a foreign speaker at the advanced level can reach to the level of native English speaker, even dedi cating his free Friday for me. Lastly, I would like to express my special than ks to my parents. My daddy, who is very protective and would do anything he can do for me, thus it is not easy for him in the beginning to let me go away even for an academic program, has been an important reason for me to be a good person. My mom has dedicated every single breath to me and has been the best supporter for my

PAGE 6

6 choice for academic life. I was able to grow up in a happy environment due to my wonderful parents. My final appreciation goes to my soul mate who has sh ared every single moment with me for the last eleven years. I always felt their affection and belief in me It has given me my endless energy for my academic progress. Due to them, I have never th ought about giving up in any single step of my academic progress. I know how lucky I am having such wonderful people around me. I deeply appreciate all of them and promise that I will try to be one who can make them happy.

PAGE 7

7 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............11 CHAPTER 1 INTRODUCTION..................................................................................................................13 Geopolitics.................................................................................................................... ..........13 Media Politics................................................................................................................. ........14 2 LITERATURE REVIEW.......................................................................................................18 Selective Exposure and the Dynami cs in Political Communication......................................18 Information Accessibility in Po larized Political Communication..........................................20 Information Trap between Medi a, Politics, and the Public....................................................22 Hostile Media Phenomenon: Geopolitics and Media.............................................................23 At the Micro-Level..........................................................................................................23 At the Macro-Level.........................................................................................................25 Demographics and Media Geopolitics....................................................................................26 Dynamics of Political Attitudes in Political Information Availability...................................27 Political Information Efficacy.........................................................................................29 Political Cynicism...........................................................................................................30 Candidate Evaluations.....................................................................................................32 Issue Awareness..............................................................................................................33 Voting Intention...............................................................................................................35 3 HYPOTHESES & METHOD.................................................................................................37 Hypotheses..................................................................................................................... .........37 Political Information Efficacy.........................................................................................37 Cynicism....................................................................................................................... ...37 Candidate Evaluations.....................................................................................................37 Issue Awareness..............................................................................................................38 Media Bias..................................................................................................................... ..38 Behavior Intention...........................................................................................................38 Method......................................................................................................................... ...........38 Purpose........................................................................................................................ ....38 Samples........................................................................................................................ ....39 Measurement...................................................................................................................40

PAGE 8

8 Procedures..................................................................................................................... ..44 Analyses....................................................................................................................... ...44 4 RESULTS........................................................................................................................ .......47 Political Information Efficacy.........................................................................................47 Political Cynicism...........................................................................................................48 Candidate Evaluations.....................................................................................................50 Issue Awareness..............................................................................................................53 Media Bias..................................................................................................................... ..54 Voting Intentions.............................................................................................................54 5 DISCUSSION/ INTERPRETATION.....................................................................................68 Political Information Efficacy.........................................................................................68 Cynicism....................................................................................................................... ...71 Candidate Evaluations.....................................................................................................72 Issue Awareness..............................................................................................................74 Media Bias..................................................................................................................... ..74 Behavior Intention...........................................................................................................75 Limitation/ Importance....................................................................................................76 APPENDIX A UNIVERSITIES AT WHICH DATA WERE COLLECTED................................................79 B SPOTS USED IN THE AD EXPERIME NT IN THE FOLLOWING ORDER....................80 First Ads Spots................................................................................................................ ........80 Second Ads Spots............................................................................................................... ....80 C POLITICAL INFORMATION EFFICACY..........................................................................82 D POLITICAL CYNICISM.......................................................................................................83 E MEDIA BIAS..................................................................................................................... ....84 F MEDIA CONSUMPTION.....................................................................................................85 LIST OF REFERENCES............................................................................................................. ..86 BIOGRAPHICAL SKETCH.........................................................................................................95

PAGE 9

9 LIST OF TABLES Table page 4-1 Information Availability & Party Aff iliation on Political Information Efficacy...............56 4-2 Information Availability & Ethnicity on Changes in Political Information Efficacy........56 4-3 Information Availability & Part y Affiliation on Political Cynicism.................................56 4-4 Information availability & Party aff iliation on Changes in Political Cynicism.................56 4-5 ANCOVA Tests of Political Info rmation Efficacy and Cynicism....................................57 4-6 Information Availability, Ethnicity & Gender on Bush Evaluation..................................58 4-7 Information Availability & Part y Affiliations on Kerry Evaluations................................58 4-8 Information Availability, Ethnicity & Gender on Kerry Evaluation.................................59 4-9 Information Availability, Ethnicity & Gender on Changes in Bush Evaluation...............59 4-10 ANCOVA Tests of Ca ndidate Evaluations........................................................................60 4-11 Information Availability on Personal Attribute Evaluation...............................................61 4-12 Information Availability on the Range of Issue Awareness..............................................62 4-13 Information Availability on Issue Salience........................................................................62 4-14 Information Availability on Media Bias............................................................................62 4-15 ANCOVA Test of Media Bias...........................................................................................63 4-16 Information Availability on Pre-Vote Intention................................................................63

PAGE 10

10 LIST OF FIGURES Figure page 4-1 Interaction of Informati on availability & Party ID............................................................64 4-2 Changes in Political information Efficacy.........................................................................64 4-3 Political Cynicism......................................................................................................... .....65 4-4 Changes in Political Cynicism...........................................................................................65 4-5 Bush Evaluation & Ethnicity.............................................................................................66 4-6 Kerry Evaluations.......................................................................................................... ....66 4-7 Media Bias................................................................................................................. ........67 4-8 Media Bias................................................................................................................. ........67

PAGE 11

11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy WHAT MEDIA DO TO BI-POLAR, 50-50 STATES By Hyun Jung Yun August 2007 Chair: Lynda Lee Kaid Major: Mass Communication As Geopolitics assumes, people in different places have different ways of thinking, personalities, and behaviors because they experi ence and are exposed to different social and political environment. In addition, as sele ctive exposure and hostile media phenomena assume, individuals have a tendency to prefer information correspond ing to and avoid information inconsistent with their own beliefs. For thes e reasons, individuals pe rceive even the same political information differently based on where they are placed. If individuals are placed in a political battle ground, they are naturally exposed to multiple sources and types of polit ical information, while other individuals in strong Democrat or Republican districts would have smaller amounts of campaign information. Liking and disliking candidates is strongly associated with individuals political pr edispositions, but media exposure tends to reinforce or decrease th eir political beliefs towards ta rgeting political candidates and issues they are interested in. Therefore, by knowing the media dynamics on political preferences among people within certain political boundaries, we are able to predict with a better explanatory power. This study particularly looks at the d ynamics between the amounts of political information available and individual s political attitudes in battle and non-battle ground states.

PAGE 12

12 Using experiment data (N=2965) gathered by the Uvote research team during the 2004 presidential election at multiple locations, this study compares two groups of individuals who have been exposed to multiple campaign inform ation in battle-grounds and who have been exposed to limited, no, or indirect campaigning information in non-battle grounds to measure individuals political at titudes based on political information availability. In addition, this study also measures how these individuals with various levels of information accessibility evaluate political candidates, recognize issues, and percei ve media messages differently by equally given sets of political ads and debate s. The study found that individuals in battle-ground states with multiple political information are more likely to have higher levels of information efficacy and lower levels of cynicism, tend to evaluate candidates positively, more easily change their previous opinions, and tend to have a broader ra nge of issue awareness th an people in non-battle ground states with limited or no campaigning in formation, with some variations within demographic factors.

PAGE 13

13 CHAPTER 1 INTRODUCTION Geopolitics The Republican candidates red swept across th e South, the Great Plains, and most of the Rocky Mountain West, while the Democratic ca ndidates blue covered almost all of New England, the Mid-Atlantic, and the West Coast in the 2004 presidential election (Marlantes, 2004). There are various scholarly approaches to exploring the causes of the divided nation. A number of scholars explain the re d and blue division through the pr eexisting bipartisan political system. Other people argue that it is because of the voters strategic political tendency toward balancing a two party system. Some other schola rs argue that it is the sum of local political characteristics and decisions (Burden & Kimball, 2000, p.13). Although different scholars argue with different explanations, they ad mit that territorial and geographical units are the basi c concept of societal and polit ical transmission. Geographical boundaries determine characteristic s of members within the group, societal form, and functional capability based on given economic, sociological, or political assets and the patterns (Arensberg, 1961, p.248). Especially, during campaigning s eason, hundreds of millions of dollars are poured into political communica tion; however, these amounts of money are spent unevenly across different political regions (Ansolabehere & Iyengar, 1995). Current scholars have started to realize the implication of geographic location as one of crucial variable to measure political phenomena (Miron & Bryant, 2007, p.399). Political ideology and political region of a physical boundary are interchangeable in U.S. politics. As long as political boundaries remain fixed and popul ation movements are modest, we expect that the political char acteristics will stay nearly constant and it is also unlikely voters preferences [will] shift markedly between elections (Burden & Kimball, 2000, pp.39,

PAGE 14

14 96). As Geopolitics assumes, people in different places have different ways of perceiving, interpreting, and behaving (Schultz, 2000, pp. 85-87). Value appeals that resonate with individuals predispositi ons are more likely to have influence on political audience as seen in their attitude and behavior changes (Gordon & Miller 2004). Differences in societal status in wealth, occupation, education, and political experien ce also shape individual s traits and beliefs among a particular group in partic ular places (Leighley, 2004, p.148). Media Politics Media and politics interact for special inte rests especially duri ng campaigning season. Politicians want to allocate part icular information in particular locations. Political parties intentionally present themselves through media to gain supports from various audiences. Campaigners strategically choose to access different political market to persuade various types of national audiences. Media often reflect thes e demands in the amount of information and programming and thus provide an environment wh ere political adverting and propaganda can be advantageously deployed (Nightungale, 2004, p. 233). The media provide citizens both objective fact ual and subjective commentary information, and citizens make political decisions based on gi ven information that is more accessible and available for them. Information about politics is mediated or learned through the press and we look to them to inform us. The media have se t current agendas, interpret the cause of events, and predict possible results (S chultz, 2000, pp.15-16). Therefore, media information is never pure and simple, especially in po litics (Miron & Bryant, 2007, p.392). There are a significant number of scholarly findings about the dynamics of political information exchanges between voters and the me dia. First, political audiences consume political information through various but accessi ble media sources. Second, citizens are simple minded if they do not have a str ong preference or ideology when they enter the poli tical arena.

PAGE 15

15 Political audiences perceive eith er pure or manifested political information through the medias view (Schultz, 2000, p.53; Schumpeter, 1993, p.85). Therefore, citizens need the media to provide a broader range of viewpoi nts, an awareness of the comple xity of politics, and a richer contextual understanding of politics in order to appropriately grasp current politics (Schultz, 2000, p.68). However, mass media today do not deliver a ma ss audience. There are also particular media structures that represent the relatively constant array of channe ls, choices, and contents in different places, given individual s affinity for political choice, preference, interests, habits, and expectation (McQuail, 1997, p.67; McDonal d, 1990). Politicians tend to offer piecemeal bits of policy proposal a little bit for everyone (Leighley, 2004, p.32) Media themselves select specific types of contents and narrow the scope of information for local audiences based on the political interests and par tially audience agreements (M cDowell & Lee, 2006, pp.179, 185). Political information seen through the media, thus, is often a set of differentiated messages that are more complex and [that] force receivers to sear ch for cues about which messages to believe. Therefore, as an alternative, individuals ofte n use their preexisting polit ical cues to filter the external information (Dalton, Beck, & Huck feldt, 1998, p112). In addition, individuals are exposed to different levels of political inform ation and opinion based on individuals patterns of social interaction and political predisposition (Huckfel dt, et al., 1995). Therefore, factors like who you are and where you belong determine the act ual level of information exposure and the ways of perceiving or interpreting given inform ation. For instance, people are much more vulnerable to political exposure when they do no t have preexisting politic al characteristics, but individuals who are occupied by st rong political beliefs are less lik ely to be influenced by given political information. The media may not direc tly influence individuals perceptions for those

PAGE 16

16 whose political identities were pr eviously established, thus less likely to change those previous attitudes; but, at least in part media exposure still stimulates voters interests and involvement in issues (McCombs, 1994, p.9) and interact with audiences political char acteristics in political communication. There should be a combined political outcome of geopolitical factors, media, and political audiences (Mcquail, 1997, p.68). However, the c onnection between media information, politics, and the impact on citizens has not been explor ed with any conceptual clarification (Miron & Bryant, 2007, p.398). A number of previous scholars have failed to provide a concrete way of understanding geopolitical pa tterns either by only taking a snap shot of a cert ain political event at a particular time at the state level, looking at particular group s of individuals in a specific location or by considering one type of media effect on an audience. For instance, although Benoit and his collea gues (2004) found different audience issue salience by different levels of information e xposure in battleground and non-battleground states, they have not looked at how media information a nd audience political pr edispositions interact, thus even the same amount of political inform ation may induce differently individuals attitude changes. In addition, the majority of research in the macro perspectives of political systems and governments often ignore media as the primar y connection between an individual and the political system in the political comm unication processes (Semetko, 2004, p.355). In order to overcome such limitations, we need to provide a connection between the macro-level of the bipolar poli tical system and the micro-le vel of individuals political preferences and ideological pattern and then consider mediati ng factors, such as media, bridging macro and micro levels of political dynamics (Burden & Kimball, 2000, pp.32, 34). Multilevel

PAGE 17

17 analyses allow researcher to investigate how th e dynamics of each layer interact and influence political consequences in a more transp arent way (Steenbergen & Jones, 2002). Therefore, this study intends to answer th e question of why do i ndividuals who are in different geopolitical locations pe rceive and interpret political in formation variously? Thus, it tests the interaction between i ndividuals political predisposi tions and political information availability in different polit ical geography of states as th e cause of various political campaigning effects. For a more concrete finding of the theoretical c onnection of geopolitics, we should understand different levels of units not only at individuals and the nation but also at state and local districts. Such an approach is theoretically logical since election outcomes are determined by aggregating votes within those geogr aphical electoral units. Voters within each geographic boundary are critical political units for red vs. blue and battleground vs. nonbattleground state compositions in the U.S. (Wright, 1998). Therefore, this study adopts geographical regions as a given factor for campai gning information availabi lity that influences individuals political att itudes and differentiates the effects of incoming political information.

PAGE 18

18 CHAPTER 2 LITERATURE REVIEW Selective Exposure and the Dynami cs in Political Communication An audience is a non-scientific social behavi oral concept. Audiences develop identities sharing cultural meaning and generate the meaning in the process of constructing and reproducing culture, meanings, and ideas. In orde r to understand the comple xity and diversity of an audience in a media dominated world for political information, it is important to understand the nature of audience engagement with the pr actice of everyday media life. For the audience, the real-world meaning systems are generated where media information is debated, discussed, and enacted (Nightungale, 2004, p. 239). There are multiple media channels and various audience interests. According to the PEW research center (2004), audiences tend to us e different media sources for their political information. Overall, local TV news (57%), da ily newspapers (41%), an d radio news (41%) are the most popular media, but network TV news (32 %), the Internet (25%), and talk radio (17%) are alternative media sources for political information. In terms of audience interests, local or national political ne ws (22%) ranked 5th place following general community news (31%), crime (30%), health news (26%), and sports (25%). Audiences are not very selective in their attention, but they are very selective in perception and resistant to unwanted influen ce (McQuail, 1997, p.59). With relatively lower levels of political interests, th e audiences are more likely to us e convenient sources for political information search based on their predispositions rather than being exposed to the mainstream news (Leighley, 2004, p.83). In addition, individu als political knowledg e often does not match the available information. Audi ences realize importan t issues based on thei r own interests and pre-exposures to various life experience; thus, they gather political issues or policy positions that

PAGE 19

19 stand for their own perspectives and releva nce (Leighley, 2004, p.143; McQuail, 1997, p.118). For that reason, media often reflectively enco urage individuality by offering ideas that are congenial to a persons self-interests rather than transmitting all messages they deliver into random audiences (McDonald, 2004, p.184). Individuals have tendencies to expose themselv es to messages and retain information that is consistent with their own beliefs, and they are likely to interpret messages in a way they already believe (Klapper, 1960). Individuals are more likely to prefer information corresponding to, and avoid information inconsistent with, th eir own beliefs and attitudes (Festinger, 1957). Audiences watch what they like and like what they watch (Barwise & Ehrenberg, 1988). Therefore, individuals who are exposed to mu ltiple viewpoints presumably select a piece of information that fits their point of view and r eact to the same information differently based on their own interests and perceptions. Although some scholars argue that there are limited effects of se lective exposure in political attitude or behavior changes (Chaffee; Saphir; Sandv ig, &; Hahn, 2001; Klapper, 1960; Lazarsfeld, Berelson, & Gaudet, 1948), individuals who have different levels of political information and different selective tendencies ha ve different political perspectives (Klapper, 1960; Milburn, 1979). Moreover, indi viduals decode the meaning of media messages in order to fit their own perspectives and va lues (Hall, 1980). In other words, altho ugh political audiences may not be completely free from unwanted info rmation and block themselves from changes in their attitudes and behaviors from such informati on, they have particular tendencies to focus on political appeals from their prefe rred candidates or parties. If i ndividuals are exposed to counterattitudinal information, they consider the messa ge as distorted information, and they only remember information that supports their initia l political beliefs (Iyengar, Hahn, & Prior, 2001;

PAGE 20

20 Milburn, 1979). Individuals motivation for selec tive exposure often reinforc es the effects of the political campaign messages that they prefer (Klapper, 1960). Previous research on selective political campaign exposure can be broadly summarized into three results. First, short-term media ca mpaigns are inefficient at changing individuals preexisting political beliefs. Second, campaigning is primarily a diffusion effect rather than increasing knowledge levels or changing specific att itudes. Third, various media and diverse political individuals have dynamic interrelations ; thus, research has mixed results (Milburn, 1979, p. 510). However, if researchers understa nd not only campaigning materials but also individuals beliefs, values, ma ss media habits, and environmen tal factors such as surrounding political moods, they would have a better ch ance to see clearer phenomenon of media and audience effects (Mendelsohn, 1973). Information Accessibility in Polarized Political Communication In an ideal world, political audiences are e xposed to various pieces of information through multiple sources. However, in practical pol itical campaigning with numerous intervening factors, political audiences are exposed to different types and le vels of information. Political polarization influences local media to be refl ective of political interests of politicians and audiences in their coverage. Media localism induces different types of program selection and quantity of political information (McDowe ll & Lee, 2006, p.177; Nightungale, 2004, p. 234). Therefore, the media have been criticized for not offering a wide range of political viewpoints (Semetko, 2004, p.356). For example, if an individua l is placed in a politi cal battleground, he or she is naturally exposed to multip le sources of political informa tion while another individual in a strongly Democrat or Republican district woul d be exposed to limited and one-sided campaign messages. Media facilitate and manipulate co mmunication across lines of political difference (Mutz, 1998, p.270), and control the flow of political information (Semetko, 2004, p.356). Thus,

PAGE 21

21 it is necessary to understand how media make certain political information available for particular audiences and how those different levels of information availability influence individuals political at titudes in order to measure the theo retical connection be tween selectivity and given accessibility. Media distribute information based on the cost-e fficiency calculations, and the levels and types of information are determined by the demands of information providers and preference of the target audiences (McQuail, 1997, pp. 52-55). One of the most popular strategies of media planning, presentation, and editorial orientation is to respond to a specific social-demographic category or to a taste of culture (McQuail, 1997, p. 85). Politicians, media, and audiences want to maximize their own utilities in the process of political comm unication. Whereas politicians wish to gain further support by distributing their agendas, the me dia want to satisfy information sources and attract more audiences, and audiences want to obtain the type of information they need and prefer (Nightungale, 2004, pp. 234-35). Therefore, the leve ls and types of information in a particular place are determined by these three contributors to the point that meets all efficiency and preferences. In other words, a different political envi ronment or geographical location has different information available by the interaction of political interests, media efficiency, and audience predispositions. According to an experimental study on restri cted versus unrestrict ed availability of information, individuals with limited information are more selective than people without such restriction. If the information search is restricted, people expe rience scarcity and increase their cognitive activity in order to achieve their goal of finding the best pieces of information, either by testing information more criti cally or by being more strongly oriented towards information quality (Fischer et al., 2005, p.482). Therefor e, information quality is very important,

PAGE 22

22 especially in the situation wher e limited information is accessible (Fischer et al., 2005, p. 488). Individuals who have a high orie ntation for political informati on, but have limited or indirect political resources, tend to find alternative wa ys to obtain necessary information such as interpersonal communicati on (Mutz & Marin, 2001). Media information availability interacts w ith political demographic factors such as political ideologies rather than one influenc ing another in communica tion processes. For instance, strong partisans are more likely to be selective in their information choice (Lowin, 1967). Republicans are more selective in inform ation filtering and more actively engaged in assessing previously selected political information (Barlett, Drew, Fahle, & Watts, 1974, p.269). From views of media, Republicans tend to have more dissimilar political information from television and newspaper sour ces than Democrats (Mutz & Martin, 2001, p.102). Therefore, political attitudes are more likely to be de termined by mixture of individuals political characteristics, purposive selectivity, and given information accessibility. Information Trap between Medi a, Politics, and the Public Media are core elements in political communica tion like an engine for a car. In terms of functionality, media serve as channels for information delivery, and they also serve as a system for information exchanges. In terms of effect s, some argue that the audience perceives media information directly without any filtering mech anism. The hypodermic needle model or magic bullet theory assumes homogenous and significant influence of mass media on the mass public. According to the assumptions, individuals are passi ve information receivers. These theoretical approaches were very appealing in understand ing media and audience dynamics during the WWI and WWII eras, from which there were very lim ited information sources with highly committed national political agendas. However, scholars began fairly quickly to recognize limited media

PAGE 23

23 effects. Rather than viewing media as pr opaganda producers, the re searchers in the limited media effect paradigm started to s ee various levels of influence on di fferent types of individuals. The media definitely influence the public by providing factual info rmation about events occurring in the current world, commentary inte rpretation about people, objects, and issues, and sets of norms and values that are necessary to sustain societies. Likewise, the public also influence the media by not only favoring certai n information the media provide, but also by avoiding certain information with which they do not agree. Communications are operated when both communicators and audiences are balanced and supported by a shared outlook and common beliefs (Mcquail, 1997, p.116). Ther efore, the media industries tr y to provide information the audience prefer to attract rather than being correct, fair, and educational (Leighley, 2004, pp.1315). As a result, media, politics, and audi ences are trapped by each other in political communication process. Hostile Media Phenomenon: Geopolitics and Media At the Micro-Level Media information is interpreted variously in different social a nd political contexts (Neuman, 1991, p.96). Even when media messages are balanced, individuals believe that perceived media are biased if the contents are against their own attitude s or beliefs (Eveland & Shah, 2003; Vallone, et al., 1985). Attitudes are very stable, consequential, and very difficult to change. Thus, it is very rare th at individuals change their attitudes in the course of normal daily life (Hovland, 1959; Hyman & Sheats ley, 1947). The attitudes of i ndividuals are significantly affected by their previous experience and emoti on to relevant arguments regardless of whether the true credibility of the source is high or not (Heesacker, Petty, & Cacioppo, 1983). As a result, the impact of political advertisi ng varies across different cultural and societal structures (Kaid & Holtz-Bacha, 1995). Individuals preexisti ng political preferences mediate

PAGE 24

24 individuals perceptions of me dia and even distinguish betwee n different media outlets, and these are not easily changed (Gussin & Baum, 2004). Individuals assessments of source credibility are determined by their internal factor s such as political awareness or partisanship. Individuals who have a particular political tendency are likely to evaluate or interpret media coverage accordingly. For instan ce, strong partisans believe that media tend to be biased and perceive different implication from the same so rt of messages; thus, they tend to be more selective for information and strengthen pro a nd con attitudes than weak partisans (Schmitt, Gunther, & Liebhart, 2004, pp.623-41). The more committed an individual is to an atti tude, the greater he or she is to resist attempts to change it (Hovland, Campbell, & Brock, 1957). The more knowledgeable a person becomes about an attitude abject, the harder it will likely be to change his or her attitude toward the object because there is so much support fo r the existing viewpoint (Petty & Krosnick, 1995, p.4). Political audiences who have strong attitudes are actively in volved in the processing of political information. If they perceive counte r-attitudinal messages from media, they tend to think the media message is biased. It is not ju st a difference of the opi nion but a difference of perception (http://www.answers.com/topic/ hos tile-media-effect). Thus, either multiple political messages or a limited single piece of in formation can be faulty facts as long as the media messages contain information that discomfo rt audiences preexisti ng political beliefs and preferences. Audiences politic al beliefs do influence their perception of media messages. Vallone, Ross and Lepper (1985) i llustrated that Arab and Israel i students both felt broadcast news on the Middle East conflic t was biased against them.

PAGE 25

25 According to an empirical study, Republican audiences are more likely to see news coverage as leaning to Democra tic candidates, but Democratic id entifiers are more likely to see the coverage as leaning to the Republican candi dates. As the hostile media effects theory assumes, partisans may see even neutral political information as hostile to their own viewpoint (Gunther & Schmitt, 2004; Schmitt, Gunther, &L iebhart, 2004). Another empirical study found that each individual reacts to the same media messages differently based on their preexisting political preferences and issue relevance. When two groups on opposite points of the Middle East issue were exposed to an identical news story about the c onflict, both groups evaluated the content as biased against to their vi ews (Dalton, Beck, & Huckfeldt, 1998). According to belief congruence theory by Rokeach and Rothman (1965), audiences previously built values, attitude, and belief syst em serve as protection layers for external influence. Krosnick et al.s (1993) estimates of correlations be tween latent attitude dimensions, talking and thinking (R=.84), previous knowledge and talking (R=.79), and previous knowledge and thinking (R=.76) are highly correlated. Thus, hostile media effect at the level of the individual needs to be c onsidered as a precondition fo r political communication. At the Macro-Level Hovland, Irving, and Kelley ( 1953) argued that individuals conforming tendencies are highly influenced by expectations or perspectiv es of committed groups with whom they are associated. Consequently, political candidates use various strategies to appeal their positions in order to get political support from different grou ps of people. Different communication channels have different effects on contra sting types of people. Various political audiences find different types of media to acquire desirable informati on (Schooler, Chaffee, Flora, & Roser, 1998). Hence, effective campaigns use more than one type of media, combin e mass media with other types of activities, such as interpersonal interac tions, and represent their view through experts of

PAGE 26

26 various fields for more effectiv e political appeals to various po litical audiences (Hofstetter, Shultze, & Mulvihill, 1992). Eveland and Shah (2003) found th at perception of media bias was more likely related to consistent information exchange among ideologica lly similar individuals rather than amount of discussion. Group preferences are another important factor for e xploring media effects. For instance, their study found that such a phenomen on is more prominent among Republicans than Democrats (Eveland & Shah, 2003). Therefore, by understanding dissimilarity between groups of people who share different po litical characteristics in differe nt political boundaries, we are able predict media effects with a better explanatory power. Thus this study particularly looks at the dynamics between the amount of political information availa bility and possible political perspectives of individuals acr oss different political boundaries. Demographics and Media Geopolitics Political audiences exist in different geogra phical spaces, and this factor interacts with various individuals predispositions such as gender, social cont ext, and subcu lture (McQuail, 1997, pp.50-52). Where an individual belongs is a crucial indicator in predicting political behavior. According to USA TODAYs survey results in the 2004 pres idential election, party identification and ethnicity as well as previous voting behavior were important factors for potential political behaviors. In fact, 89% of Democrat s voted for Kerry, and 93% of Republicans voted for Bush. White voters (57% ) selected Bush while Hispanics (55%) and Asians (57%) voted for Kerry. Considering 90% of people who voted for a certain party in the previous presidential election vo ted for the same party for the 2004 presidential election, political demographics should be considere d, or at least be c ontrolled, to understand media geopolitics (USA Today, 2004).

PAGE 27

27 In addition, peoples media reach and the interaction with dive rse individual traits caused different political consequences. People tend to ge t certain types of information from particular media and filter the given media information based on their various tastes, beliefs, and preferences. Hence, although na tional audiences are heterogeneous there are stable patterns of media interaction with certain type s of audiences (McQuail, 1997, p.54-55). Dynamics of Political Attitudes in Political Information Availability Attitudes refer to persistence, resistance, im pact on information processing and judgments, and guiding behavior (Petty & Krosnick, 1995, p.4). The medias effect can be cognitive, attitudinal, or behavioral. None theless, it is difficult to track when audiences shift across these three different consequential lines. Thus, this study addresses only attit udinal changes resulting from media information. Previous scholarly research has already fo cused on individuals changes in political attitudes by direct media information exposur e; however, they concentrated on psychological factors at the micro level of th e individual, ignoring exte rnal factors such as surrounding political mood and patterns of informa tion distribution or vice versa. Therefore, results were somewhat inconsis tent based on what factor each particular study looked at. Converse (1964) argued that citizen s rarely or randomly change their attitudes when they have lower levels of belief and information expos ure, but they tend not to or only systemically, change their attitude if they have strong belie fs with high levels of information exposure. Individuals with middle levels of beliefs are most likely to change their attitudes by given information (Converse, 1964). There would be per fect stability where information input is zero .Beyond this minimum, of course stabilit y would increase as a function of involvement (Converse, 1962, p.587). Similarly, Zaller (1991 & 1992) argued a nonlinear relationship between attitude changes and political awareness. Accord ing to Zaller (1991 & 1992), citizens with intermediate levels of

PAGE 28

28 knowledge are most likely to change their previous att itudes by information. In contrast, people who have low or high levels of knowledge are mo re likely to be stable in their attitudes. Moreover, individuals with highe r levels of exposure and inte rests tend to associate with more active political evaluations wi th a greater ability to interpre t, encode, store, and retrieve new information (Semetko, 2004, p.361). Benoit et al. (2004) also found th at individuals who live in battleground states had greater levels of understanding on issues and stronger standpoints compared to people in non-battleground states. Atkin and Heald (1976) found that voters w ho were highly exposed to media messages were somewhat more likely to attach higher agenda priorities. The frequency of political information exposure was associated with pos itive candidate evalua tions. As expected, individuals who were exposed to more political media information were more likely to have higher or more accurate political knowledge. Liking and disliking candidates were strongly associated with individuals political predispos itions but media exposure tended to help positive evaluations towards targeting po litical candidates. Interestin gly, individuals who were less exposed to campaigning political information and have stronger motivation to obtain information tend to be politically more polarized by attaini ng increased advertising exposure (Atkin & Heald, 1976, pp.217, 220, & 227). Although there are great levels of controversy wi th the effects of content types, timing, and strategies, political ads definitely influence pol itical audiences. Campaign information obtained through political ads or candidate visits often help crystallize existi ng attitudes by sharpening and elaborating them by providing more informa tion and intangible impressions about political candidates and stimulating emotional reaction. In addition, political advertising reinforce existing attitudes in voters to keep in the fold a voter who is leaning to ward a candidate but not

PAGE 29

29 strongly committed (Harris, 1989, p.170). Of c ourse, such reinforcement is liable to be translated into actual political be haviors and creates political cons equences (Harris, 1989, p.171). However, due to inconsistent measurement a nd validity, studies in th e field still persist with a great deal of disagreement (Semetko, 2004 p.359). In addition, limitation in approaches at the individual level overlooked macro levels of systemic impact and constrains of politics and media on individuals attitudes changes. Therefor e, this study intends to explore attitudinal dynamics by different levels of political inform ation bridging micro le vels of individual tendencies and macro levels of geopolitical media structure. Political Information Efficacy Political information efficacy re fers to how confident an individual is that he or she has sufficient information to engage in the poli tical process (Kaid, Mc Kinney, & Tedesco, 2007). This concept is closely related to internal efficacy, which Neimi, Craig, and Mattei (1991, p.1407) define as beliefs about ones own co mpetence to understand, and to participate effectively in, politics. However, political inform ation efficacy differs from internal efficacy, in that it focuses solely on the voter's confiden ce in his or her own po litical knowledge and its sufficiency to engage the political process (to vote). In terms of demographic factor relevance, previous research found that voters who are more engaged in politics and have stronger partisan ship tend to have higher levels of political information efficacy (Kaid, McKinney & Te desco, 2004, 2007). Efficacy and political participations are learned through a socializati on process based on informal norms and values and formal information exchange (Easton, 1965; Dennis, 1967, pp25-38). Moreover, individuals levels of information confidence are highly related to information accessibility and exposure. Political information a nd the discussion of political issues increase general confidence in individuals own political cap ability that is positively related to active

PAGE 30

30 political participations (Lane, 1969, p. 152; Almond and Verba, 1963, p257; Dahl, 1961, p. 286; Campbell et al., 1960, p 105). Specifically, polit ical campaigning information is positively correlated to political information efficacy, wh ich possibly leads to active voting behaviors (Clarke & Acock, 1989, pp.551-55). According to another experimental study, both political advertising and televised debate exposure incr ease individuals confid ence in their level of political knowledge (Kaid, Landrevill e, Postelnicu, and Martin, 2005) People reported the lack of information about candidates as one of the top reasons for not participating in politics (Declare Yourself, 2003). Regarding this rationale, this study assumes that the factors of information variability a nd accessibility are important indicators that determine individuals level of confidence in th eir political knowledge and capability, and thus finally leading them to be good citizens. Political Cynicism Political Cynicism is disbelief in politics and mistrust of political actors (Levin & Eden, 1962). Contrasted to political in formation efficacy, political cyni cism has negative implication in democracy. Political cynicism could lead to lower voter turnout and to a lessening of information seeking about political issues and candidates (Kaid, McKinney, & Tedesco, 2000, p.680). Voters feel powerlessness and insignifica nce (Miron & Bryant, 2007, p. 395). The level of cynicism has been quite high in American politics. Previous scholars have argued about the undesirable phenomenon with different theoretical perspectives. Cognitiv ely, American publics believe that political participation such as voting does not make any significant difference (Agger, Goldstein, & Pearl, 1961). Systemically, even more scholars discuss the un satisfactory outcomes of the political process as the cause of American citizens politi cal cynicism (Levin & Eden, 1962).

PAGE 31

31 Structurally, some argue that prevalent politic al cynicism is highly related to the way the media deal with political events, candidates, and system (Cappella & Jamieson, 1997). Some suggest political and media struct ures are alienating th e political citizen, and others bring in the prevailing trend of negative belief about politics as the reason of political cynicism. Political audiences who are exposed to ne gative political information, such as negative advertising, become politically more negative, and positive ads reinforce positive political expectations (Kaid, McKinney, and Tedesco, 2000). Moreover, dist orted, biased, or unfair media information or coverage about politics turns au diences off from any further atte ntions to political information (Ansolabehere, Iyengar, Simon, & Valentino, 19 94). Therefore, the medias negativity and malfunction magnifies prevalent cynical attitudes among American citizens. Recent scholars have focused on political information function as both the cause of and solution for political cynicism. For instance, while political advertis ing viewing has mixed effects on political cynicism based on types of a dvertising, televised deba tes tend to significantly reduce audiences cynical attitudes (Kaid, Landrev ille, Postelnicu, & Martin, 2005). Scholarly research found that political cynicism is curable and treatable by efficient political communication strategies (Berman, 1997). For inst ance, persistent, diverse, and consistent information campaigns are efficient ways to redu ce disbelief in politics and increase awareness of government tasks (Wheeler 1994; Garnett, 1992; Denton &Woodward, 1990). Political communication with a broad range of information is necessary to perceive positive political services, issues, and processes (Stipak, 1977; Be rman, 1997, p.110). Such political information is provided by various strategi es in various ways (Berman, 1997, p.107). Bermans study (1997) argued that cities that use fre quent information strategies ex perience less cynicism (p.111).

PAGE 32

32 Therefore, the political information should be explored in order to determine an appropriate explanation related to political cynicism. Therefore, my study looks particularly at the interaction between information accessibility and constraints and the preexisting political characteristics crossing diffe rent political locations. Two counter-attitudinal measurements of polit ical information efficacy and cynicism are highly related to the leve l of individuals information expos ure, information accessibility, and external information constr aints (Cappella & Jamieson, 1997; Kaid, McKinney, & Tedesco, 2000; Shearer, 2005). Therefore, rather than looki ng at types of media or specific contents to predict a single or limited political attitude, this research investigates the amount of political information that is available for individuals in different geopolitical regions and its multiple functions across contradictory and procedur ally different political attitudes. Candidate Evaluations Previous research argues that political in formation exposures through political ads or debates influence audiences candida te evaluations. Although there is a mixture of evidence, the valence of information influence candidate evaluations (Funk, 1999, p.701). According to Kaid, political information exposure, such as poli tical advertising, shapes audiences perceptions of targeted candidates image, inform them ab out candidates issues, and influences general political attitudes about political systems and voting choice (Kaid, 1995). Benoit, McKinney, and Holbert (2001) argue that expos ure to direct political information, such as debates, changes viewers evaluations of the candi dates and strengthens confidence in preferred candidates, as well as the factors that determine vote choice. These studies have also found a positive correlation between political information and ca ndidate evaluations. Exposures to political advertising, TV debates, and candi date visits increase positive ev aluations of political candidates

PAGE 33

33 (Kaid, 2004; Shaw, 1999). Individuals who have mo re information availabl e for their political decisions are inclined to be supportive of their political leaders (Berman, 1997, p.110). In addition, individuals who are less informed about politics more frequently shift their political position or preference than individuals who have more political information (Converse, 1962). However, at the same time, very detailed political information a bout candidates traits could create more complex ways of processing in formation, and thus lead to multiple political interpretations (Funk, 1999, p.715). Nonetheless, we still need to consider the strength of party affiliations as an important factor in determining candidate evaluations. St rong partisans are distinct ly supporting the same party candidates across multiple levels of poli tical information availability (Erikson & Tedin, 2007, pp.84, 119, & 278; Funk, 1999, p.715; Rahn, 1993). Political information about candidates and preexisting political preferences or external polit ical contexts interact in determining voters evaluations on candidates (Fiske & Taylor, 1991) A mixture of information available for a given political sector and the level of individuals political attention codetermine and allow for th e identification of like-mi nded candidates (Funk, 1999, p.716) Although some empirical inc onsistence exist, such as no evidence of 1976 debate influence on candidates image evaluations (C haffee, 1978), but clear evidence of 1960 debate on different evaluations toward Nixon and Kennedy (Katz & Feld man, 1962), the influence of political information exposure has been an important factor and had been at least an interacting indicator with other demographic in determ ining candidates evalua tions (Kaid, 2000; Payne, Golden, Marlier, & Ratzan, 1989). Issue Awareness. Individuals learn important poli tical issues by given media, the political elite, or people around them. Particularly, media di scourse shapes public awareness of political issues and helps

PAGE 34

34 individuals interpretation of issue thus reducing the gap between individuals diverse attitudes toward a public issue (Gamson & Modigliani, 19 89; Shaw & Maxwell, 1977). The concept of issue awareness here is consistent with the theoretical arguments of MaCombs and Shaws agenda setting (1972) that l ooked at the interaction between media coverage on issues and audiences issue salience. However, rather than investigating the direct correlations between the amount of media coverage on partic ular issues and audiences issu es, this study intends to look at how the range of mixture of political information during the campaigning period determines individuals perceptions of important issues varied in di fferent political locations. Such media information interacts with individu als levels of information in determining public issue awareness. According to empirical research, the public tend to easily become aware of currently prominent issues ove r a relatively short period of ti me. Instant issue awareness is more likely to occur among people who have low and medium levels of political information. Moderately aware citizens are more flexible in th eir attitudes toward part icular political issues (Koch, 1998, pp.224-27; Zaller, 1992). Citizens awareness of impor tant political issues has changed substantially by given pol itical attentions to certain political issues. Therefore, the relationship between the politic al elites, the media, and public opinion are more critically interacted in determining overall issue awaren ess by given political information than for any other political consequence. The medias interpretative package of a given range of scope and symbols on fluctuating and time-sensitive issues, such as the nuclear weapons issue, tend to reduce different schemas and ambivalence am ong individuals and i nduce more homogenous tendency of issue awareness and simultaneous re actions of people who experience different life events (Gamson & Modigliani, 1989, pp.3, 33-36).

PAGE 35

35 Moreover, newly received information tend to activate information already at the recipients disposal, stored in long-term memory (Nel son, Oxley, & Clawson, 1997, p.225). High knowledge levels and personal sophistication on issues determ ine the levels of attitude changes by new information. For instance, people who are exposed to previous information are more likely to be familiar with an issue provide d by media; thus, the additional information tend to be trivial than the effect of new informati on for individuals with lo w levels of previous information (Eagly & Chaiken, 1993; McGuire, 1968 & 1985; Zaller, 199 1 & 1992). In other words, a new piece of information for individua ls who were already exposed to previous information should not be affected to such a degree of less informed individuals by the new information (Nelson, Oxley, & Clawson, 1997, p.228). Although previous scholars studying age nda-setting and framing argued about the relationship between the media aspects of issu e presentation and the public issue salience (McCombs and Shaw, 1972; Rogers and Dear ing, 1994), these studies overlooked the importance of individuals politic al predisposition and the vari ations of media information availability in the different pol itical environments. Therefor e, this study investigates the interaction of individuals given political characteristics within a political region and the amount of information accessibility at the regional level in order to observe issue awareness determined at multiple stages. Voting Intention. Political audiences are sensitive to cont extual cues when they make decisions, formulated judgments, or express opinions (Iye ngar, 1991, p.11). Individuals are vacillating in voting decision especially dur ing campaigning periods (Convers e, 1962, p.579). A large amount of political information supplied by media stimulat es voters to be more active; however, limited information gives voters no meaningful differe nce between different candidates (Converse, 1962,

PAGE 36

36 p.586). Furthermore, individuals with less inform ation tend to retrieve a piece of information from information retained from the past, but individuals with higher levels of information operate with a large storage of political lore; thus, different levels of political information differentiate individuals polit ical decision making processes in various degrees (Converse, 1962, p.583). Information processing is always involved in an attitudinal decision (Bassili, 1993, p.55) Therefore, if the flow of information is strong, no particular prediction is safe because diverse information inputs have a great potential for any di rection of changes for voters, but if the flow of information is weak, the vote will be a pur e party vote because ther e are few chances of defections and voting indecisiveness (Converse, 1962, p.586-87). Therefore, the more remote the respondent was from the flow of information, the more stable his vote intention (Converse, 1962, p.590). The lack of information is directly related to decrease of the probability of voting in presidential elections (Bartels, 1996). Over all, we can infer that although the volume of information cannot fully determine voting tendency; at least, it magnifies or oscillates floating voting tendencies (Converse, 1962, p.591).

PAGE 37

37 CHAPTER 3 HYPOTHESES & METHOD Hypotheses Political Information Efficacy H1: Individuals in the political information rich states tend to have highe r levels of political information efficacy than individuals in the pol itical information poor states, controlling other demographic variables. H2: After a series of political information exposu res, individuals in information poor states are more likely to increase their levels of political information efficacy than individuals already in the information rich states. Cynicism H3: Individuals in the political information rich states tend to have lowe r levels of political cynicism than individuals in th e political information poor states controlling other demographic variables. H4: After a series of political information exposure s, individuals in the political information poor states are more likely to reduce their levels of cynicism than i ndividuals already in information rich states. Candidate Evaluations H5: Different levels of political information av ailability influence indi viduals evaluations of different candidates. H6: After a series of political information expo sures, individuals in limited political campaign information states are more likely to increa se their positive evalua tion towards political candidates than individuals alrea dy in information rich states.

PAGE 38

38 H7: After a series of political information exposure s, individuals in information rich states are more likely to reinforce their per ceptions of candidates personal traits than individuals in the information poor states. Issue Awareness H8: Individuals in the political information ri ch states tend to recognize larger numbers of important issues than individuals in the politi cal information poor states, controlling for other demographic variables. H9: After a series of political information exposures, important issues become more comparable between individuals in limited political campaign information states and individuals already in information rich states. Media Bias H10: Individuals in information poor states are more likely to th ink that media are biased than individuals in information rich states. Behavior Intention H11: Different levels of political information av ailability influence indi viduals voting intention. Method Purpose There are various concepts in measuring political attitude, su ch as accessibility, ambivalence, preference, certa inty, elaboration, importance, perception, and knowledge. There are also various factors that influence attitude formation, including personality, direct personal experience, family influence, group infl uence, and interpersonal/small group/mass communication. In this study, in order to observe variability of individua ls political attitudes depending on the level of political information individuals can possess, information availability at the state level and the information effects on individuals bridging tw o different levels of

PAGE 39

39 approaches were investigated. In other words, levels of political information accessibility and availability in each state that constrains the amount of info rmation available for political audiences within the political regions were consid ered at the macro level. At the micro level, individuals political attitude s and evaluations under different information circumstance and degrees of attitude changes after exposure to the same set of political information in different political regions were measured. A theoretical connection between micro levels of individual attitudes and macro levels of mass communication was established by looking at how individuals who are in information rich states differ from individuals who are in informa tion poor states in their political attitudes and information perceptions. These connections of micro and macro level approaches are recently recognized by many mass communication scholars and are useful ways to deal with a more concrete pattern of political comm unication (Mutz et al., 1996). Samples This research used experimental data c onducted and collected by the Uvote 2004 research group, a non-partisan, inter-univers ity research team, during the peak campaigning season of the 2004 presidential election from Sept ember to November across 22 different locations in the U.S. (Note A). Respondents (n=2965) we re enrolled at the participating univer sities in various courses and voluntarily participated in the study. The average age of participants is 20.78 years. The sample consists of 43% males and 57% females and about equal numbers of Republicans (37%) and Democrats (40%), but lesser numbers for Independents/Others (23.2%) in their party ID. Since the majority of participants were young adults, there could be particular political tendencies reflective of younger peop les political characteristics. Considering the noticeable increase of young adults participation in the 200 4 presidential election, the sample would be

PAGE 40

40 more meaningful and representative of the part icular political tendenc ies and outcomes of the election year. According to th e 2004 presidential election resu lt, a large majority (85%) of young people declared that they were paying close attention to the campaign, 80% registered to vote, and 42% actually showed up and voted. Th ese figures were significantly higher among the college student population, a group that reached a 66% voter turnout ra te (Circle, 2004). However, there could still be possibilities for excessive presenting of cynical attitudes and insufficient presenting of political inform ation efficacy using young Americans who have been known as politically ignorant, highly cyni cal, and highly apathe tic throughout American history (Deli Carpini, 2000; Mann, 1997; Third Millenium, 1999). Thus, this study may have some limitation to track age variations in di fferent political tendenc ies and the effects of information exposure in di fferent political regions. Measurement The survey asked a series of questions about participants demogra phic factors, such as age, gender, party ID, race, and political and me dia tendencies, such as levels of cynicism, information efficacy, candidate feeling thermo meters, perceptions on candidates personal attributes, issue awareness, media consumption pa tterns, media bias, and voting intentions. To measure individuals cynical attitudes in a mo re consistent way, eight different questions measuring individuals levels of trust on govern ment, political leaders, and general political systems were adopted from reliabl e previous research, such as the American National Election Panel Studies (ANES) and many other journa l articles (Kaid, Mc Kinney, & Tedesco, 2000; Rosenstone et al, 1997). These eight categories were moderately correlated based on Cronbachs Alpha reliability score1 of .66. Based on the general accepta nce level of .70 for Cronbach Alpha 1 Cronbachs alpha reliability score indicates a coeffici ent of consistency among different variables. High Cronbachs alpha reliability refers to how well variou s different items can be constructed into a single

PAGE 41

41 test, it did not have a very satisfactory level of reliability. However, this study still created a measurement of cynicism on a scale of 5 by averag ing those eight variable s giving high credits to previous studies repeat ed reliable tests on thes e categories (Appendix D). In measuring political information efficacy, f our items asking the levels of respondents confidence on their political know ledge and information were also adopted from previous studies, and these items have been asked in the series of ANES (Kaid, McKinney, & Tedesco, 2000; Rosenstone et al, 1997). They are also measured on 5-point agree-disagree scales and comprised the political information efficacy scal e, which achieved a high Cronbachs alpha level of .85 (Appendix C). For candidate evaluation measurement, the su rvey asked respondents to give any number of degrees between 0 and 100 for each presidential candidate. This measure was adopted from a feeling thermometer traditionally used by ANE S studies (Rosenstone et al.,1997). A higher number is referenced to a more positive evaluation towards the candidate. The measurement of candidate favorability was also proved as a very reliable indicator for a comparison of relative evaluations for different candidates (Kaid, 2004b). Besides overall feeling thermometers toward candidates, the survey also measured individuals perception about vari ous personal traits of each candida te. These personal attribute scales were adopted from Lynda Lee Kaid and her colleagues series of national election surveys and have been proved as excellent measuremen ts with multiple reliabi lity tests (Kaid, 1995 & 2004). These scales were well inco rporated into this experimental research in order to measure how each individual in multiple locations with multiple levels of information perceives candidates personal attributes in various ways and how much direct information exposure unidimensional scale. A reliability coefficient of .70 or higher is considered "acce ptable" in most social science research (http://www.ats.ucla. edu/STAT/SPSS/faq/alpha.html).

PAGE 42

42 changes an individuals perception on candidates personal attributes The twelve different pairs of personal traits on a scale of seven each in qualification, sophisticati on, honesty, believability, successfulness, attractiveness, friendliness, sin cerity, excitement, aggressiveness, strength, and activeness were measured. These measures achieved high Cronbachs Alpha scores for Bushs (.866) and Kerrys (.794) multiple personal attributes ; thus, a single indicator of personal traits was created for each candidate. Issue attentiveness has also been one impor tant measurement in exploring political communication effects. In order to measure individuals issue awar eness in different levels of political information availability, the survey as ked participants to write down all issues they believed to be the most important issues the co untry faced. In the data cleaning process, the Uvote research team created tw elve exhaustive and exclusive i ssue categories of health, war, economic, crime, terror, education, environmen t, elderly, foreign policy, children, tax, and welfare, and marked all categories each individual indicated. The to tal numbers of issues in each individuals list were used to tr ack how an individual who was from an information rich region is different from an individual from an information poor state in terms of range of issue recognitions. In addition, the ranks of important issues in the informati on rich and poor regions were used to explore the aggreg ate levels of issue awareness in different political information circumstances. In order to measure media bias, the survey adopted seven different questions asking individuals perceptions on me dia fairness, objectiveness, and normative contribution to the society. These questions have been asked in multiple national surveys such ANES and GSS for the last several decades. A m easurement of media bias on the scale of 5 was created with

PAGE 43

43 Cronbachs Alpha reliability score of .84 to anal yze how individuals in different geopolitical boundaries perceive the same kind of political information variously (Appendix E). Demographic factors, such as gender, ag e, party ID, and ethnicity, and previous campaign information exposure were controlled in order to examine a more transparent relationship between audiences pol itical attitudes and political information availability. Particularly, past week po litical information exposure2 was adopted to control individuals amount of previous political information exposur e, rather than directly looking at multiple channel exposures or total amount of media c onsumption time because my study intends to look at only the dynamics of political informati on over overall media consumption patterns. As we recognized in previous research, pol itical information exposure is somewhat a troublesome concept in the new and multi-media era. There are different leve ls of accessibilities to and desire of political information. Some indi viduals are unintentionally exposed to political messages, but some others intentionally search for political information. In other words, an individual who is very interest ed in elections in non-battlegr ound states still would look for political advertisings and candida tes agendas through the Internet other cable ch annels, national news, or interpersonal communication. In contrast there could be a possibi lity for audiences to be exposed unintended political spots and candida te speeches due to a great amount of political information available while turning media cha nnels on. A more problem atic issue to track political information consumption is that differe nt medium cannot be comparable in terms of amount of political information since the unit of coverage is different and the effects vary. 2 A combined scale with 1) Level of me dia exposure of presidential campaign in the past week and 2) Level of talk with other people about presidential campaign in the past week (0 Never 5 A lot). There were no statistical differences between using the combined scale of the two ways of campaign information exposures and using each indicator separately as control variable s in the main analyses. Therefore, I used the combined indicator for a more cohesive and parsimonious measurement of the main effect of regional information availability by simplifying other control variables.

PAGE 44

44 Therefore, it is more appropriate to control each individuals previous political information exposure regardless of channels or sources than controlling particular media consumption patterns or sum of media time span. Procedures Participants were asked to fill out a pre-test asking about general political tendencies such as party ID, ideology, cynicism, information effi cacy, and candidate evaluations. Following the pretest, respondents were then exposed to sets of political advertisements or one of the presidential TV debates. The ad spots shown consisted of a series of Bush and Kerry spots, consisting of a mixture of positive and negative s pots for each candidate (Appendix B). The first advertising experiments took plac e during September 28-30, just prio r to the first presidential debate, and the second advertising experiment s were held from October 28-29. Presidential debates were shown in real time as the events occurred. The participants completed a pre-test questionnaire and then viewed first, second, and third presidential debates in real time as it occurred in Miami, Missouri, and Arizona on September 30, October 8, and October 13 respectively. After viewing, respondents completed post-te st questionnaires asking about the same categories of political characteristics, attitudes, and candidate evaluations. The experimental setting that allowed both single or multiple information exposure(s) were more appropriate to reflect the reality of individuals political information receptions since general audiences were exposed to some mixture of these political information. Analyses The study mainly measured media exposure di versity and the attitudinal diversity by different geopolitical regions. This study compares two groups of individuals who are able to directly access multiple campaign information and who can obtain only limited, indirect, or no

PAGE 45

45 campaign information because of their political geographic constraint. Based on media localism and different levels of political information reach, this study compares individuals political attitudes in battle versus nonbattleground states us ing political information availability. Adopting the 2004 peak season campai gn attention index (Fairvote, 2004)3, individuals in the top 15 campaign attention states (Iowa, Ohi o, New Hampshire, Wisconsin, New Mexico, Florida, Nevada, Pennsylvania, Minnesota, Co lorado, Michigan, Maine, Oregon, West Virginia, and Missouri) are compared to individuals in the rest of the states where there was low campaign attention. Those top 15 campaign attention states had 94.14% of candidate visits, 98.45% of TV political ads, and 98.48% of campaign money expenditure during the 2004 presidential campaigning period (Fairvote, 2004). Besides thes e top attention states, unfortunately, the rest of the states obtained almost no campaign attention. This studys sample consists of some varia tions of demographic gr oups in two different regions of high and low political information. 1694 participants (57.1%) in the sample were from information poor states and 1271 participants (42.9%) were from information rich states. Among individuals in information poor states, 46.5% (n=772) were males and 53.5% (n=889) were females; 38.5% (n=653) were ethnic minorities and 61.5% (n=1041) were ethnic majorities; and 41.2% (n=659) were Democrat s, 40.8%(n=653) were Republicans, and 17.9% (n=287) were Independents. Am ong individuals in information rich states, 37.9% (n=481) were males and 62.1% (n=787) were females; 20.5% (n=261) were ethnic minorities and 79.5% (n=1010) were ethnic majorities; lastly, 42.5% (n=514) were Democrats, 38.8% (n=1088) were Republicans, and 19.5% (n=546) were Independents. There were st atistical differences in the 3 The document was reported by FairVote-The Center fo r Voting and Democracys Pres idential Elections Reform Program. The index was calculated by total campaigning fund ing, total candidate visits, and total TV ads aired in each state.

PAGE 46

46 composition of sub-populations of gender ( 2=21.445, p .001), ethnicity ( 2=110.493, p .001), and partisans ( 2=8.749, p .013) between information rich and poor states. In other words, there were higher proportion of female, ethnic majority and political Independents in battle-ground, information rich states. In order to avoid statis tical noise from these su b-population variations, I controlled for them in each set of main analyses. In the first set of analyses, individuals polit ical attitudes, such as political cynicism, information efficacy, candidate ev aluations, and issue awareness, were compared in battleground and non-battleground states. In the second set of analyses, individua ls changes in their political cynicism, information efficacy, candidate eval uations, issue awareness, perception on media bias, and voting intention were compared by determining equally given sets of political advertising and debates to see how people interpret the same message in various ways based on their preexisting information leve ls, controlling for all demogr aphic variables and political interests and attentions.

PAGE 47

47 CHAPTER 4 RESULTS Political Information Efficacy Individuals who were older (t=4.508, p .001) and consumed more political information in the previous week (t=33.599, p .001) had higher levels of polit ical information efficacy. In general, females (M=3.16, SD=.60) tended to be mo re confident about thei r levels of political knowledge than males (M=3.12, SD=.63) (F[1, 2686]=81.610, P .001). On average, individuals le vels of confidence on political information were similar between information rich states (M=3.53, SD =1.01) and information poor states (M=3.38, SD=1.01), while controlling other demographic and media variables such as campaigning information attention, gender, party ID, age, a nd ethnicity. Thus, the first hypothesis was not confirmed. When the geopolitical factor was combined with political party identification, however, the interaction effect had a sta tistical significance. For instance, Republicans in information rich states (M=3.66, SD=0.98) tended to have higher le vels of political information efficacy than Republicans in information poor states (M=3.48, SD=1.00). However, political Independents in information poor states (M=3.41, SD=0.96) had highe r levels of political information efficacy than the Independents in information ri ch states (M=3.14, SD=1.08) (F[2, 2708]=7.635, P 0.001). More interestingly, individuals who had strong political identi fication, regardless of their direction of partisanship, were mo re likely to have higher levels of confidence on their political information efficacy than individuals with lo w or no degrees of political leaning in both information rich and poor states (F[2, 2708]=14.831, P 0.001) (Figure 4-1, Tables 4-1 & 4-5).

PAGE 48

48 After viewing a series of po litical ads and debates, younger people were more likely to increase their levels of information efficacy (t=-3.590, P .001), and females (M=0.21, SD=0.51) were more likely to increase their confidence in their political knowle dge than males (M=0.16, SD=.45) (F[1, 2686]=4.264, P .039). However, individuals who consumed more political information in the previous week were less likel y increased their confidence in their political knowledge by a given set of political information compared to people who had previously consumed less campaign information (t=-7.240, P .001) As the second hypothesis predicted, exposur e to political campaigning information changed individuals previous polit ical attitudes differently accord ing to the geographical factor of political information availability and other de mographic factors. In terms of information confidence, although the geographica l factor solely did not have any statistical significance in changes in political information efficacy, it had an interaction effect with other demographic factors. After seeing political ad s and debates, different ethnic groups in different levels of political information availability changed their le vels of political inform ation efficacy variously. Ethnic minorities in political information poor states (M=0.25, SD=0.56) tended to more significantly increase their levels of informati on efficacy than any minority in the information rich states (M=0.18, SD=0.46) and ethnic ma jorities in both information poor (M=0.16, SD=0.46) and rich states (M=0.19, SD=0.46), wh ile controlling other demographic variables (F[1, 2686]=3.756, P .053). Thus, we can infer that pol itical ads and campaigning messages could be more influential for ethnic minorities who do not have enough political information in non-battleground states (Figure 42 & Tables 4-2 & 4-5). Political Cynicism In general, ethnic minorities, such as Black s, Asians, and Hispanics, (M=3.27, SD=.58) tended to be more cynical than those belonging to the ethnic majorit y, Whites (M=3.09, SD=.61)

PAGE 49

49 (F[1, 2686]=6.013, P .014). Political Independents were more likely to be cynical (M=3.26, SD=.58) than Democrats (M=3.18, SD=.61) and Republicans (M=3.06, SD=.62). (F [1, 2686] = 16.704, P .001). People who had higher levels of information about the campaigns from previous weeks tended to be less cynical than those that had lower levels of previous political information exposure (t=-13.398, P .001). The factor of information availability by different regions ha d a strong effect on individuals political cynici sm. Individuals in informa tion poor states (M=3.22, SD=0.60) tended to be more cynical than those in poli tical information rich states (M=3.05, SD=0.62), while controlling all individual media consumpti on, party ID, age, gender, and race (F[1, 2620] =12.237, P 0.001). Thus, this result confirmed the third hypothesis. At the same time, information availability is also interacted with other demographic factors, such as party ID, in determining the cynicism level. Although political Independents te nded to be more cynical than Democrats and Republicans in both information poor and rich states, the Independents in information poor states (M=3.32, SD=0.57) were more likely to be cynical than Independents in information rich states (M=3.19, SD=0.58) (F[2, 2620] =4.038, P 0.018) (Figure 4-3 & Tables 4-3 & 4-5). After exposure to a mixture of positive and ne gative political spots and candidate debates in an experiment setting, participants, who already consumed certain levels of campaigning information from the previous week, increased thei r levels of cynicism mo re significantly than people with lower previous campaign information (t=2.772, P .006). In addition, ethnic minorities reduced more significantly their prev ious levels of cynicism (M=-.07, SD=.44) compared to ethnic majority after viewing sets of political information (.M=0, SD=.39) (F[1, 2580] =7.939, P 0.005)

PAGE 50

50 However, unlike the fourth hypothesis predic ted, direct information exposure did not make any difference in change of political cynicism between people from information rich and poor states. Nonetheless, the information availabili ty is again interacted with individuals party identification in changing indivi duals cynical attitudes. Afte r political information exposure, Democrats (M=-0.05, SD=0.44) and Independe nts (M=-0.05, SD=0.39) in the political information rich states reduced their levels of cynicism, but Democrats (M=-0.01, SD=0.40) and Independents (M=0.01, SD=0.39) in the information poor states remained at their previous level of political cynicism, while controlling ot her demographic variables (F[2, 2580]=3.076, P .046) (Figure 4-4 & Tables 4-4 & 4-5). Candidate Evaluations Overall, older adults were mo re supportive of Bush (t=2.914, P .004), and younger adults were more supportive of Kerry (t=-4.398, P .001). Ethnic majorities (M=54.45, SD=34.89) evaluated Bush more positively th an ethnic minorities (M=39.40, SD=31.67) (F[1, 2657]=11.373, P .001), however, ethnic minorities (M=58.6 9, SD=26.84) evaluated Kerry more positively than ethnic majorities (M=45.96, SD=30.64) (F[1, 2596]=11.971, P .001). As expected, Republicans were strong supporters of Bush (M=81.50, SD=18.57), but Democrats evaluated Bush relatively negativ ely (M=24.11, SD=23.68) (F[2, 2657]=798.540, P .001). In contrasts, Democrats were strong supporters of Kerry (M=71.39, SD=20.03), but Republicans evaluated Kerry significantly more negatively (M=25.54, SD=23.45) (F[2, 2596]=524.947, P .001). In addition, previous pol itical information consumpti on induced the tendency of negative evaluation toward Bush (t=-3.253, P .001). Hypothesis 5 arguing that different levels of political inform ation availability influence individuals candidate evaluations was partially confirmed, in that it appeared true for the Democrat candidate, Challenger Kerry, but not fo r the Republican candidate, incumbent Bush.

PAGE 51

51 Although information availability was not str ongly influential on evaluation towards Bush, individuals who were in information low states (M=51.57, SD=34.32) seemed to evaluate Bush a little more positively than people in inform ation rich states (M=47.97, SD=34.99) (F[1, 2657]=2.855, P .091). However, the factor of informa tion availability, once again, interacted with other demographic factors, such as et hnicity and gender, in influencing candidate evaluations. Ethnic minorities in informa tion rich states (M=31.97, SD=29.21) tended to evaluate Bush more negatively than ethnic minorities in information poor states (M=42.34, SD=32.15). However, ethnic majorities in either information rich (M=51.77, SD=35.19) or poor states (M=57.15, SD=34.41) evaluated Bush in similar degrees (F[1, 2657]=3.742, P .053) (Figure 4-5 & Tables 4-6 & 4-8) In addition, males who belonged to ethnic majorities in information poor states (M=60.61, SD=33.84) evalua ted Bush most positively and females who belonged to ethnic minorities in information rich states tended to have the least positive Bush evaluations (M=28.03, SD=27.97) compared to other people with different co mbinations of these demographic factors (Tables 4-6 & 4-8). Geographic variances in political informati on availability significantly influenced evaluations of Democratic Part y candidate, Challenger Kerry. Individuals in the political information rich states (M=51.36, SD=31.03) were more likely to evaluate Kerry positively than individuals in political inform ation poor states (M=48.35, SD=29.37) while controlling for other demographic factors (F[1, 2596]=6.028, P .014). In addition, even under unequal information av ailability, Democrats were always more supportive of Democratic Party candidate, Ke rry. However, the geopolitical factor of information availability influenced the effect of party ID for candidate support. Democrats in information rich states (M=73.78, SD=18.72) mo re positively evaluated Kerry than other

PAGE 52

52 Democrats in information low states (M=69.41, SD =20.85). Interestingly, Independents showed similar tendency. Political Independents in information rich states (M=53.01, SD=23.99) evaluated Kerry more positively compared to other Independents in information poor states (M=50.02, SD=21.70) (F[2, 2596]=3.010, P .049) (Figure 4-7 & Tables 4-7 & 4-8). Moreover, females who belonged to ethnic minorities in information rich states (M=66.52, SD=22.77) were more likely to be Ke rry supporters and males who belonged to ethnic majorities in information poor states (M=39.72, SD=28.78) were less likely to be Kerry supporters (F[1, 2596]=7.081, P .008) (Tables 4-8 & 4-10) However, direct political information expos ure had no effect on changes in candidate evaluations between individuals who were from lo w and high political information availability states. Again, however, there were interaction effects regarding information availability and other demographic factors. After viewing polit ical ads and debates, females who belonged to ethnic minorities in information rich states (M=2.09, SD=13.73) were mo st likely to increase their positive evaluation toward Bush, but males who belonged to ethnic minorities in information rich regions were most likely to reduce their positive att itudes toward Bush (M=2.36, SD=17.07) compared to individuals with some other combinations of those three demographic factors (F[1, 2646]=5.079, P .024) (Tables 4-9 & 4-10). Despite no influence of direct information exposure in changes of individuals overall feeling thermometer toward candi dates in different regions, th e information exposure changed individuals perceptions on candidate personal traits especially for challenger, Kerry. When the survey asked series of personal attributes of Bush and Kerry, i ndividuals in battleground states with high levels of information (M=1.94, SD=6.91) were more likely to increase positive perception about Kerrys persona l traits than peop le in information poor states (M=1.18,

PAGE 53

53 SD=7.82) after controlled all demographic factor s such as previous political information exposure, age, gender, party ID and ethnicity (F[1, 2636]=7.215, P .007) (Table 4-11). Issue Awareness As hypothesis 8 posited, indivi duals who were in information rich states (M=3.23, SD=1.17) tended to be aware of broader range of important issues than people from information poor states (M=3.12, SD=1.17) (F[1, 2082]=4.977, P .026). However, as hypothesis 9 predicted, exposures to a series of political information messages reduced the gap in policy issue awareness between those individuals from inform ation rich and poor stat es; thus, the different levels of issue awareness faded away (F[1, 2294]=.278, P .598) (Table 4-12). Regardless of the regional difference in poli tical information availa bility, during the 2004 presidential campaigning period, people believed that war and economy were the most important issues the country faced. However, ther e were some variations in importance of each policy issue according to different regions. The rank order correlation between low and high information regions was .958. However, after expos ure to the same sets of political information, the correlation between issue importance among individuals from inform ation rich and poor states became .993. In order to se e whether there is a statistical difference between two different correlations of the pre correlation between info rmation rich and poor st ates and of the post correlation between information rich and poor stat es, I adopted a techniqu e of Fishers r-to-z transformation that overcomes non-normality independent correlation problems (Blalock, 1972, 406-407). The combination of Spearmans r ho rank order correlation te sts and Fishers r-to-z transformation test proved that a series of political information exposures made issue importance become more identical among individuals from limited political campaign information states and from information rich states. According to Fishers r-to-z transforma tion statistical test, the difference between the .958 on the pre-test a nd the .993 on the post-test was proved to be

PAGE 54

54 significant. In other words, th e post test correlation is statistically higher than the pre-test correlation (Z'=-2.2172, P .001). 4 Therefore, we can infer that political information do change the perceptions of important polic y issues and the sets of politic al information make individuals issue recognition more identical (Table 4-13). Media Bias As the hypothesis 10 predicted, information availability determined individuals perception about media. After seeing a series of balanced political ads or a candidate TV debate, individuals who were in information poor stat es (M=3.62, SD=0.64) were more likely to think that media were biased compared to individuals who were originally in information rich states (M=3.53, SD=0.68) (F[1, 2726]=10.613, P .001) (Figure 4-7 & 4-8 & Table 4-14 & 4-15). People who have limited previous information w ith strong political ch aracteristics in nonbattleground states tended to be lieve that counter-attitudinal info rmation was biased, unfair, and exaggerated, although the inform ation were correct and balan ced. Although there is no systematic and pervasive ideological bias in the media, the preexisting media pattern and the audiences routine in interpreting media messa ges deviate the percepti on of given political information (Erikson & Tedin, 2007, pp.236-7) Voting Intentions When the survey asked participants for wh om they planned to vote in the coming 2004 presidential election, par ticipants in different levels of in formation availability had differing 4 Fishers z' (called r-to-z transformation) is used for computing confidence intervals on correlation and difference between correlations. This technique transfers r to z-score in order different sets of correlations to be compared statistically. In the first step, z scor es are calculated based on the formula, Z' = ln[|(r+1)/r-1)|]/2 using r scores (Z'1=1/2 ln (1+.958/ 1-.958)=1.921, Z'2=1/2 ln (1+.993/ 1-.993)=2.826). In the second step, the standard error of difference between the two correlations is estimated based on formula, SE = SQRT[(1/(n1 3) + (1/(n2 3)].( .SQRT (.0833+.0833)=.408). In the third step, the difference between two z-scores is divided by the standard error (Z'=1.921-2.826/.408)=-2.218). Since the absolute z-score is higher 1.96. Therefore, it is statistically significant at =.05. In other words, the post test correlation between information rich and poor states is statically higher than the pre test correlation betw een information rich and poor states.

PAGE 55

55 voting intentions. Voters from political informa tion poor states were equally likely to vote for both Bush (44%) and Kerry (44%); however, peopl e from information rich states were more likely to vote for Kerry (51%) over Bush (37%) ( 2= 15.765, P .001). Therefore, hypothesis 11 on different voting intentions by ge opolitical variance of informa tion availability was confirmed (Table 4-16).

PAGE 56

56 Table 4-1. Information Availability & Party Affiliation on Political Information Efficacy Campaign Attention Indexa Party IDb Mean (SD) Low Attention States Democrat 3.53 (1.03) Independent 3.41 (0.96) Republican 3.48 (1.00) High Attention States Democrat 3.49 (0.99) Independent 3.14 (1.08) Republican 3.66 (0.98) a. Not significant, a*b. F[2, 2708]=7.635, P 0.001, c. Min=1, Max=5 Table 4-2. Information Availability & Ethnicity on Changes in Political Information Efficacy Campaign Attention Indexa Ethnicityb Mean (SD) Low Attention States Minority 0.25 (0.56) Majority 0.16 (0.46) High Attention States Minority 0.18 (0.46) Majority 0.19 (0.46) a. Not significant, a*b. F[1, 2686]=3.756, P .053 Table 4-3. Information Availability & Party Affiliation on Political Cynicism Campaign Attention Indexa Party IDb Mean (SD) Low Attention States Democrat 3.24 (0.59) Independent 3.32 (0.57) Republican 3.16 (0.61) total 3.22 (0.60) High Attention States Democrat 3.11 (0.63) Independent 3.19 (0.58) Republican 2.92 (0.61) total 3.05 (0.62) a. F[1, 2620] =12.237, P 0.001, a*b. F[2, 2620] =4.038, P 0.018, c. Min=1, Max=5 Table 4-4. Information availability & Party affiliation on Changes in Political Cynicism Campaign Attention Indexa Party IDb Mean(SD) Low Attention States Democrat -0.01 (0.40) Independent 0.01 (0.39) Republican -0.02 (0.39) High Attention States Democrat -0.05 (0.44) Independent -0.05 (0.39) Republican -0.02 (0.42) a. Not significant, a*b. F[2, 2580]=3.076, P .046

PAGE 57

57Table 4-5. ANCOVA Tests of Political Information Efficacy and Cynicism Political Information Efficacy Political Information Efficacy Changes Cynicism Cynicism Changes F(Sig.) F(Sig.) F(Sig.) F(Sig.) Intercept 369.099(.001)***104.752(.001)***3878.160(.001)***1.064(.302) Political Information 1050.172(.001)***40.350(.001)***118.437(.001)***4.221(.040)** Age 16.152(.001)***10.389(.001)*** 1.150(.273) 1.217(.270) Information Availability 0.886(.347) 0.001(.971) 12.237(.001)***1.348(.246) Party ID 14.831(.001)***0.001(.999) 16.704(.001)***0.437(.646) Ethnicity 1.036(.309) 0.903(.342) 6.013(.014)** 7.939(.005)*** Gender 81.610(.001)***4.264(.039)** 0.644(.422) 0.550(.459) Information Availability Party ID 7.635(.001)***1.649(.192) 4.038 (.018)** 3.076(.046)** Information Availability Race 0.008(.930) 3.756(.053)* 0.816(.366) 1.241(.265) Party ID* Race 6.037(.002)***1.886(.152) 4.096(.017)** 2.299(.101) Information Availability Party ID* Race 5.122(.006)***0.815(.443) 1.737(.176) 1.701(.183) Information Availability* Gender 0.191(.662) 0.019(.890) 0.886(.347) 0.279(.598) Party ID* Gender 1.550(.213) 0.901(.406) 0.205(.815) 0.463(.629) Information Availability* Party ID *Gender 1.152(.316) 0.386(.680) 0.912(.402) 0.413(.662) Race Gender 0.085(.771) 0.256(.613) 3.269(.071)* 0.504(.478) Information Availability* race *Gender 0.749(.387) 0.074(.786) 2.658(.103) 1.652(.199) Party ID* Race *Gender 0.065(.937) 2.121(.120) 0.342(.710) 1.237(.290) Information Availability*Party ID*Race *Gender 0.691(.501) 2.069(.126) 1.836(.160) 1.432(.239) ***p .01, **p .05, p .10

PAGE 58

58 Table 4-6. Information Availability, Ethnicity & Gender on Bush Evaluation Campaign Attention Indexa Ethnicityb Genderc Mean (SD) Low Attention States Minority Male 41.91 (30.68) Female 42.72 (33.44) Total 42.34 (32.15) Majority Male 60.61 (33.84) Female 54.37 (34.64) Total 57.15 (34.41) Total 51.57 (34.32) High Attention States Minority Male 39.25 (30.22) Female 28.03 (27.97) Total 31.97 (29.21) Majority Male 49.30 (34.56) Female 53.33 (35.52) Total 51.77 (35.19) Total 47.97 (34.99) a. F[1, 2657]=2.855, P .091, a*b. F[1, 2657]=3.742, P .053, a*b*c F[1, 2657]=9.592, P .002, c. Min=0, Max=100 Table 4-7. Information Availability & Party Affiliations on Kerry Evaluations Campaign Attention Indexa Party IDb Mean(SD) Low Attention States Democrat 69.41 (20.85) Independent 50.02 (21.70) Republican 26.77 (23.35) Total 48.35 (29.37) High Attention States Democrat 73.78 (18.72) Independent 53.01 (23.99) Republican 23.76 (23.51) Total 51.36 (31.03) a. F[1, 2596]=6.028, P .014, a*b. F[2, 2596]=3.010, P .049, c. Min=0, Max=100

PAGE 59

59 Table 4-8. Information Availability, Ethnicity & Gender on Kerry Evaluation Campaign Attention Indexa Ethnicityb Genderc Mean (SD) Low Attention States Minority Male 56.69 (25.33) Female 56.58 (29.25) Total 56.63 (27.41) Majority Male 39.72 (28.78) Female 46.50 (29.57) Total 43.48 (29.40) Total 48.35 (29.37) High Attention States Minority Male 58.48 (27.44) Female 66.52 (22.77) Total 63.73 (24.73) Majority Male 48.57 (30.49) Female 49.32 (32.40) Total 48.42 (31.65) Total 51.36 (31.03) a. F[1, 2596]=6.028, P .014, a*b. Not Significant, a*b*c F[1, 2596]=7.081, P .009, c. Min=0, Max=100 Table 4-9. Information Availability, Ethnici ty & Gender on Changes in Bush Evaluation Campaign Attention Indexa Ethnicityb Genderc Mean (SD) Low Attention States Minority Male 0.48 (15.50) Female -0.08 (15.12) Total 0.18 (15.29) Majority Male -0.42 (14.89) Female -0.12 (12.72) Total -0.26 (13.22) Total -0.09 (14.03) High Attention States Minority Male -2.36 (17.07) Female 2.09 (13.73) Total 0.55 (15.08) Majority Male 0.30 (14.06) Female -1.27 (14.18) Total -0.77 (14.14) Total -0.52 (14.33) a. Not Significant, a*b. Not Signi ficant, a*b*c F[1, 2646]=5.079, P .024, c. Min=0, Max=100

PAGE 60

60Table 4-10. ANCOVA Tests of Candidate Evaluations Bush feeling thermometer Kerry feeling thermometer Changes in Bush Evolutions Changes in Kerry Evaluations Source F(Sig.) F(Sig.) F(Sig.) F(Sig.) Intercept 437.780(.001)***590.181(.001)*** 3.856(.050)* 0.968(.325) Political Information 5.652(.018)** 3.402(.065)* 3.591(.058)* 0.680(.410) Age 4.117(.043)** 11.904(.001)*** 1.578(.209) 0.257(.612) Information Availability 2.855(.091)* 6.028(.014)** 0.014(.905) 0.581(.446) Party ID 798.540(.001)***524.947(.001)*** 1.006(.366) 2.944(.053)* Ethnicity 11.373(.001)***11.971(.001)*** 0.321(.571) 1.181(.277) Gender 1.880(.170) 0.009(.923) 1.285(.257) 1.010(.315) Information Availability Party ID 2.323(.098)* 3.010(.049)** 0.263(.769) 0.503(.605) Information Availability Race 3.742(.053)* 1.970(.161) 0.104(.747) 0.434(.510) Party ID* Race 7.632(.001)***6.153(.002)** 0.260(.771) 0.119(.888) Information Availability Party ID* Race 0.517(.596) 1.265(.282) 0.290(.749) 0.234(.791) Information Availability* Gender 0.090(.765) 1.366(.243) 0.414(.520) 0.026(.872) Party ID* Gender 6.385(.002)** 2.647(.071)* 0.111(.895) 2.469(.085)* Information Availability* Party ID *Gender 2.584(.076)* 2.415(.090)* 0.748(.474) 0.727(.483) Race Gender 0.311(.577) 2.860(.091)* 3.578(.059)* 1.340(.247) Information Availability* race *Gender 9.592(.002)***7.081(.008)*** 5.079(.024)**0.032(.858) Party ID* Race *Gender 0.874(.418) 1.555(.211) 0.073(.930) 0.906(.404) Information Availability*Party ID*Race *Gender 0.373(.689) 1.999(.136) 0.393(.675) 0.309(.743) ***p .01, **p .05, p .10

PAGE 61

61Table 4-11. Information Availabili ty on Personal Attribute Evaluation Bush Kerry Change in Bush's Trait Change in Kerry Trait Low Attention States M=51.90 (SD=13.74)M=52.30 (SD=12.49)M=0.49 (SD=6.59)M=1.18 (SD=7.82) High Attention States M=53.04 (SD=14.89)M=53.49 (SD=11.56)M=0.54 (SD=5.92)M=1.94 (SD=6.91) Not significantNot significa ntNot significantF[1, 2636]=7.215, P .007

PAGE 62

62 Table 4-12. Information Availability on the Range of Issue Awareness Campaign Attention Index Information Exposure Mean (SD) Low Attention States Pre 3.12 (1.17) Post 3.26 (1.22) High Attention States Pre 3.23 (1.17) Post 3.20 (1.91) Pre issue comparison: F[1, 2082]=4.977, P .026, Post issue comparis on: Not significant Table 4-13. Information Availability on Issue Salience Low Attention High Attention Rank Prea Postb Prec Postd 1 War War Economy War 2 Economy Economy War Economy 3 Terrorism Health Education Health 4 Education Terrorism Health Education 5 Health Education Terrorism Terrorism 6 Tax Tax Foreign policy Tax 7 Foreign policy Foreign policy Education Foreign policy 8 Elderly Elderly Elderly Elderly 9 Environment Environment Environment Environment 10 Welfare Welfare Crime Welfare 11 Crime Crime Welfare Crime 12 Children Children Children Children Spearman's rho R (a*b)=.979, Spearman' s rho R (c*d)=.972, Spearman's rho R (a*c)=.958, Spearman's rho R (b*d)=.993 Table 4-14. Information Availability on Media Bias Campaign Attention Index Mean(SD) Low Attention States 3.62 (0.64) High Attention States 3.53 (0.68) F[1, 2726]=10.613, P .001

PAGE 63

63 Table 4-15. ANCOVA Test of Media Bias Media Bias F(Sig.) Intercept 2492.537(.001)*** Media consumption 10.749(.001)*** Age 24.100(.001)*** Information Availability 10.613(.001)*** Party ID 2.596(.075)* Ethnicity 0.002(.965) Sex 11.420(.001)*** Campaign Attention* Party ID 0.148(.863) Campaign Attention* Race 0.020(.887) Party ID* Race 2.353(.095)* Information Availability Party ID* Race 0.159(.853) Information Availability* Gender 0.637(.425) Party ID* Gender 0.407(.665) Information Availability* Party ID *Gender 0.211(.810) Race Gender 0.505(.477) Information Availability* race *Gender 0.165(.684) Party ID* Race *Gender 0.342(711) Information Availability*Party ID*Race *Gender 0.101(.904) Table 4-16. Information Availabi lity on Pre-Vote Intention Kerry Bush UndecidedOther Low Attention States 43.98%44.22%9.41%2.40% High Attention States 50.67%37.19%9.91%2.22% 2= 15.765, P .001

PAGE 64

64 Figure 4-1. Interaction of Inform ation availability & Party ID Figure 4-2. Changes in Polit ical information Efficacy

PAGE 65

65 Figure 4-3. Political Cynicism Figure 4-4. Changes in Political Cynicism

PAGE 66

66 Figure 4-5. Bush Evaluation & Ethnicity Figure 4-6. Kerry Evaluations

PAGE 67

67 Figure 4-7. Media Bias Figure 4-8. Media Bias

PAGE 68

68 CHAPTER.5 DISCUSSION/ INTERPRETATION My research examined the co-effects of me dia information availability and political predisposition within different geopolitical boundaries in predicti ng individuals political attitudes and media perceptions. Individuals prev ious beliefs or attitude and external political information circumstances co-affect their political attitudes, preferences, and changes in their previously existing political characteristics. As expected, individuals cynical attitudes and perception of media bias were more prevalent in information poor states, and the level of political information efficacy for strong partisans was much higher in information rich states. Geopolitical constraint over information availab ility or accessibility even influences candidate evaluations, issue awareness, a nd voting intentions. High inform ation availability increased positive evaluations and voting intention for th e challenger and more comprehensive issue awareness. Various demographic factors were also influenced by information availability in influencing political attitudes a nd preferences. Certain demographic groups, such as political Democrats and Independents in battleground states with high levels of information availability and ethnic minorities in non-battleground states w ith low campaigning information, were more vulnerable to newly received political informa tion. The implication of each finding will be discussed in the following sections. Political Information Efficacy When the relationships between informati on availability and political information efficacy were measured, different levels of info rmation availability in batterand non-battle ground states solely did not infl uence individuals polit ical information efficacy and the changes after direct information exposure. However, the combined effect with pa rty identification, one of the strong indicators in political characteristic s of individuals, diffe rentiated individuals

PAGE 69

69 confidence on their political knowledge For instance, Republicans in information rich states had higher levels of political information efficacy th an Republicans in information poor states, but Democrats and Independents in information poor st ates had higher levels of efficacy than other Democrats and Independents in information rich regions. Overall, strong partisanship increased the levels of individuals confidence on th eir levels of political information. In addition, after seeing political ads and deba tes, the geographical fa ctor solely did not have any statistical significance in changes in political inform ation efficacy. In other words, although direct political information exposure in creased overall audiences information efficacy, the differences between the level of change in people who were in information rich and poor states was very marginal. However, it had an in teraction effect with ethnicity. Ethnic minorities in political information poor st ates tended to more significan tly increase their levels of information efficacy than any minority in the inform ation rich states and ethnic majorities in both information poor and rich states, while controlling other demographic variables Throughout previous research, individuals self-perception th at they are capable of understating politics was not easily changeable and of ten turned out as a given factor in political communication processes (Rudolph, Gangl, & Stevens, 2000). There could be both theoretical and methodological reasons for the constant levels of political information efficacy. Theoretically, individuals politi cal dispositions and characteri stics are established by primary groups of people, such as family, school, and peer group. These primary groups develop solid levels of individuals political knowledge and e fficacy; thus, political efficacy is not easily changeable by temporally available levels of political information (Langton & Karns, 1969, pp.813-14). Methodologically, experi mental studies theorizing stimulus effects have difficulties in finding statistically significant results because of complicating compon ents of main effects

PAGE 70

70 and peripheral interaction with multiple social properties of individuals (McClelland & Judd, 1993). One of the critical intervening factors in the relationship between political information availability in different political regions and political attitudes was party identification. It has been well documented that the level of political information was strongly related to party identification in determining individuals political tendencies (Erikson & Tedin, 2007, pp.21617). One of the interesting result s is that Democrats and Independe nts in information rich states somehow had lower levels of political inform ation efficacy than the same partisans in information poor states could be interpreted by un easiness with available information. As issue ownership assumes, Republicans can easily handle hard issues, a nd Democrats are perceived as more capable of dealing with soft issues (Cam pbell et al. 1960; Petr ocik, 1997; Damore, 2004). Recalling that the 2004 presidential election was a war-related election, political information on hard issues such as security, war, and terrorism were less familiar to or discomforted people who lean to the left or distinguish themselves from either side of the political tendencies (Source Watch, 2005; Layman & Carsey, 2002). Theref ore, Democrats and Independents who were exposed to a great degree of uncomfortable di scussions felt less confid ent of their knowledge levels on such issues. In addition, in terms of party ID strength and information efficacy, party identifiers recognize the levels of political knowledge that are consistent w ith their political disciplines (Layman & Carsey, 2002). Party loyalists tend to be interested in and pay attention to campaigns, follow government and public affair s, possess knowledge a bout politics, watch debated between candidates on tele vision, register to vote, and tu rn out to vote (Kamieniecki, 1988, p.373). Therefore, there is a hi gher correlation between strength of partisan affiliation and

PAGE 71

71 the level of information (Erikson & Tedin, 2007, pp.87-8), and strong partisans tend to believe that they have higher levels of political information than politically independent people (Coleman, 1996). Ethnicity was another important interacti ng factor. Ethnic minorities who are often isolated from the mainstream media coverage an d political attention tend to be more vulnerable to newly incoming information (Erikson & Tedi n, 2007, pp.204-5). Therefore, ethnic minorities, such as Asians, Hispanics, Blacks, and other mixed groups in information poor states changed their attitudes more significantly af ter exposure to new political information compared to ethnic majority or other ethnic minority in information rich states. It tells us that it would be more efficient for politicians and campaigners to pr ovide campaigning information for minority people in low of political information environments. Cynicism As predicted, the factor of information avai lability determined individuals levels of political cynicism. Individuals in information poor states were more cynical than those in political information rich states regardless of other demographic factors. Less informed individuals tend to be cynical than individuals who have access to multiple political information (Rosenberg, 1955). Frequent information exposu res decease citizens cynical attitudes and increase trust toward general political pr ocess and outcomes (Berman, 1997, p.111; Cappella & Jamieson, 1997, p.83). Therefore, one of the best st rategies to reduce public cynicism is to provide persistent, divers and consistent information campaigns (Berman, 1997, p.106). However, information availability still interact ed with other demographic factors, such as party identification, in determining the cynicism level. Interestingly, political Independents tended to be more cynical than Democrats a nd Republicans in both information poor and rich states and Independents in information poor st ates were more likely to be cynical than

PAGE 72

72 Independents in information rich states. Political Independents tend to be cynical about political views of both sides of Democratic and Republican parties. Independents distinguish themselves by rejecting the major parties. Recalling Am erican history, unrest among youth, civil right groups, and disengaged people from strong party ties, especially re lated to issues of war and human rights, have been political Independents. Therefore, these groups tend to be naturally more cynical, but somewhat flexible in their pol itical choice since they use policy issues that are given for each particular elec tion in order to make political decisions (Erikson & Tedin, 2007, pp.84, 119, & 270). Again, although direct information exposure di d not make any difference in change of political cynicism between people from information rich and poor states, in formation availability interacts with individuals part y identification in cha nging individuals cynical attitudes. After political information exposure, Democrats and I ndependents in the political information rich states reduced their levels of cynicism, but Democrats and Inde pendents in the information poor states remained with their previous level of political cynicism. This tendency suggests that although the high level of political information always tends to reduce political cynicism, individuals with low political in formation often have no particul ar tendency or motivation to change their previous att itudes (Erikson & Tedin, 2007, p.86). Candidate Evaluations This study found that different levels of po litical information availability influence individuals candidate evaluations for Democrat candidate, Kerry, but not for the Republican candidate, Bush. More specificall y, individuals in the political info rmation rich states were more likely to evaluate Kerry positively than individua ls in political information poor states. In terms of candidate evaluation, although po litical information tends to increase candidates evaluations (Berma n, 1997; Shaw, 1999), there have been somewhat consistent

PAGE 73

73 political outcomes throughout American presiden tial elections that i ndividuals having higher level of political information were more likely to support the challengers but people with lower levels of information tended to support incumb ents who were often better known. Incumbents are often well known to even poli tically ignorant people through mu ltiple media coverage, but at the same time, those incumbents are also more vulnerable to negative attacks on their previous political history. Therefore, more negative issu e ads targeting incumbents failing policies and other mistakes are prevalent in media cove rage. The campaigning of the 2004 presidential election repeated the tendency (Kaid, et al, 2007 ). As a result, individuals who obtained more political information could be more likely to be exposed to negative messages about Bush, thus evaluating him more negatively and less likely to vote for him than individuals who were in low political information states. This tendency is more apparent for ethnic minorities who were originally more political Democratic and isolat ed from mainstream information (Jasperson & Yun, 2007). Interestingly, individuals did not change their feeling thermometers toward both candidates, Bush & Kerry, by a series of polit ical information; however, the information exposure changed their perceptions on candidate personal traits, especially for challenger. Individuals in battleground states with high levels of informati on were more likely to increase positive perception about Kerrys personal traits th an people in information poor states. This result is well supported by previous research arguing that the political audience tend to learn about the candidates personal tra its and issue, but they are less likely to change their attitudes toward candidates by campaigning informa tion (Kaid, McKinney, & Tedesco, 2007; Shaw, 1999; Kaid & Chanslor, 1995; Funk, 1999; Kenney & Rice 1988;). Moreover, Kerrys advanced

PAGE 74

74 personal trait appeal can also be explained by a new and fresh image of a challenger above heavily negative tags on the incumbent (Romero, 1996). Issue Awareness As expected, information availability influenced the range of political issues recognition. Individuals who were in informati on rich states listed larger number s of issues that they believed important than people from information poor state. Direct information expos ures to a series of political information reduced the gap in range of policy issues awareness and the rank of issue importance among individuals in different poli tical regions (Shaw, 1999; Funk, 1999; Atkin & Gary Heald,1976). There have been consistent findings that the political a udience learned candidates substantive policy issues and image through paid campaign ads (Kaid, McKinney, & Tedesco, 2007; Freedman, Franz, & Goldstein, 2004). Less informed individuals are more likely to be susceptible to new information than well-informed individuals, thus easily perceiving new issues (Nelson, Oxley, & Clawson, 1997, p.233). With highe r levels of information, individuals who are in battleground states are more likely to understand current issues comprehensively (Benoit et al., 2004). In other words, opinion consistency on political issues is higher among more politically informed individuals than the le ss knowledgeable (Erikson & Tedin, 2007, p.81). The result proved that media and the interaction with geopolitical cons traints on level of information shape individuals perception of curren t issues (Iyengar & Kinder, 1987). Media Bias As hostile media phenomena assume, individuals with strong politic al preference in nonbattleground states are more likely to think that media are biased than individuals in flexible and moderate political tendencies in battleground states. Even wh en media messages are neutral, individuals believe that perceived media contents are biased if the contents discomfort them

PAGE 75

75 (Eveland & Shah, 2003; Vallone et al., 1985). Strong partisans tend to be more selective for information and strengthen pro and con attitudes than weak partisans; therefore, the hostile media phenomena is more preval ent in non-battleground states with limited political information availability and strong political tendenci es (Schmitt, Gunther, & Liebhart, 2004). Behavior Intention Interestingly, voters from political information poor states were equally likely to vote for both Bush and Kerry, while people from informati on rich states were more likely to vote for Kerry over Bush. This tendency relies on info rmation flow. With low information flow, individuals are more likely to be stable in their political choices. Therefore, individuals in nonbattleground states with low politi cal information tend to stick with their previous preference. However, individuals with high information flow ar e more likely to be flexible in their political attitudes (Converse, 1962). Excessive negativity directed at an incumbent could again increase positive reaction to the opposite challenger (R omero, 1996). And, it leads to anti-candidate voting from a desire to vote against one of th e two presidential candidates (Sigelman & Gant, 1989, p.84). Thus, we can infer that political inform ation reinforcement is liable to be translated into actual political behaviors (Ha rris, 1989, p.171; Clarke & Acock, 1989). The level of political information does lead to active participation in election and makes the American democracy better functional. According to 2004 NES data, 92% of highly informed voters voted while only 65% of less informed vot ers voted (Erikson & Tedin, 2007, p.270-72, & 336). Therefore, higher political info rmation tends to reduce political audiences cynical attitudes and behaviors. In addition, geopolitics reflect both media discriminations and audiences ideological critics. Thus, political information is the necessary condition for working democracy.

PAGE 76

76 In order to understand the fi eld of political communication, we need to deal with interactive dynamics of media, political structure, and audiences predispositions. As my study observed, there are multiple interactions in the geopolitical phenomenon. As scholars, such as Gerbner, Gross, Morgen, and Signiorelli, examin ed about the relationship between the amount of TV viewing and political attitudes, overall in dividuals who were exposed to heavy amounts of TV viewing were more likely to consider themse lves as politically moderates, while the light viewers tend to be either strong conservative or liberal opinion holders. Therefore, there was a more significant gap between individuals who were exposed to limited amount of political information compared to those who were exposed to large amount of information in different political locations (Harris, 1989, p.171). However, we also need to aware of facts that higher media exposure tends to have lower quality of at tention. With heavy viewing, audiences are less alert, and the experiences are less reward ing (Kubey &, Csikszentmihalyi, 1991). Limitation/ Importance This research intended to build a theoreti cal connection between the micro-level of individuals political attitude s and the macro-level of the me dia and political structures. Although this study showed evidence that political information levels depend considerably on the intensity of the campaign and the flow of in formation and determine degrees of audiences political attitude changes, it has its ow n empirical and theoretical limitations. Empirically, despite the broad range of samples across twenty two different locations in the U.S., the samples consisted of voluntary partic ipants in a university setting. In other words, the data were not gathered through a random sampling of all possible sub-populations. Therefore, the possibility exists that the result s overor under-represent certain sub-populations. In addition, in the index scali ng process, although all categori es I adopted were repeatedly proven as reliable indicators by national survey data such as NES, thus legitimizing their use,

PAGE 77

77 one of the indexes, cynicism, had a relatively low level of Cronbachs scor e (.66); this could lead to less significant results. Theoretically, political psychol ogists stress the mechanisms that shape political opinions by the way information is processed (Eriks on & Tedin, 2007, p.66). With multiple media channel choices and specialized media contents political audiences could be both homogeneous and heterogeneous. The one obvious phenomenon is segmentation and fragmentation of audiences. In the results, me dia tend to be much more diverse to manage different consumers with appropriate products and provide more i ndividualized contents. Such segmentation and fragmentation create difficulties for media with extensiveness and unpredictability to reach both the larger general public and smaller target groups (Mcquail, 1997, p.133; Barnes & Thomson, 1994, p.89). Hence, political information become s more partisan and sensationalistic. Interpretative journalism is unavoidable phenomenon. As a mediator, the media is in the middle of political communication between politicians and public, and control the content types and size of political information for the public (Erikson & Tedin, 2007, pp.229-35). Therefore, the theoretical generalization of this study faces some limitation by inherent elements of media information and individual dive rsity. In the multimedia era, if there is no direct political information available, citizens use other channels of informal political information such interpersonal political discussio n. In other words, indi viduals who are in lack of political information somehow fill the vacuum by increasing informal communication about politics (Converse, 1962, p.595-97) In addition, there might be a great dimension of individual differences that any theoretical research cannot completely track. As we age, individuals exposed to di fferent components of media, symbolic roles, and life patterns adopt different proces ses of information perception.

PAGE 78

78 Therefore, many academic generalizations may not work perfectly for each single individual (Mcquail, 1997, p.121). For that reason, further re search can be strengthened with a more qualitative approach by investigating pe rsonal motivation and life experience. Although there has been evidence that regiona l impact on politics has diminished because of easier mobility, universal media penetration, c itizenship delusion and the arrival of new media era, the long-term trend of dis tinctive political geographic phe nomenon has persisted (Erikson & Tedin, 2007, pp.216-7). Despite the fact that political audience s immunize themselves from political information influence from media using personal beliefs (Druckman & Nelson, 2003), previous research has confirme d that political information exposure influences political audiences (Erikson & Tedin, 2007, p.251). Political audiences, politicians, and media ar e within triangular relations in political communication processes. Although there are evidence of minimal effects, political information exposure through TV ads and candidates events do affect political audience (Ansolabehere & Iyengar, 1995). The geopolitical factor of information availability is a crucial predictor for political attitudes, candidate preferences, issue awareness, and voting intentions. The relationships between media, geopolitics, and politic al attitudes at multiple levels of individual, local, and national needs further sc holarly research for more conc rete theoretical developments and sophisticated empirical approaches.

PAGE 79

79 APPENDIX A UNIVERSITIES AT WHICH DATA WERE COLLECTED Consumnes River College Dominican University Elon University Emerson College Iowa State University New York University North East Oklahoma State University. Northeastern State University Ohio University Saint Cloude State University Texas A&M University of Akron University of Colorado at Denver University of Florida University of Kansas University of Missouri University of New Haven University of North Carolina-Chapel Hill University of Oklahoma University of South Dakota University of Texas at San Antonio Virginia Tech

PAGE 80

80 APPENDIX B SPOTS USED IN THE AD EXPERIME NT IN THE FOLLOWING ORDER First Ads Spots Bush positive ad Speaks on camera, Laura seats next to him, which child to pick first on 9/11 Kerry positive ad I defended this country Kerry negative ad 10 million jobs Bush negative ad Taxing our economy Kerry positive ad Prescription drugs Bush negative ad Practical vs. big government Kerry negative ad Wrong choices RNC negative ad Now Kerry promis es (Intelligence budget reforms) RNC positive ad Agenda for America DNC positive ad Stronger (endorsement by Gen. Merrill McPeak Second Ads Spots Bush negative adAfter Septem ber 11 our world has changed. Kerry negative ad Clip of Bush from plane carrier with the mission accomplished banner Bush positive ad Homeland security and fighting terrorism Kerry positive ad Economy (creating jobs, helping small business, tax cuts, cutting down on dependence on middle east oil) Independent positive ad Pro Bush with Ashleys story DNC negative ad Bushs mistakes in Iraq Bush negative ad Kerrys flip-flop on Iraq

PAGE 81

81 Independent (the Media Fund) negative ad Connection between Bushs family and Saudi Arabian oil business Bush negative ad Ke rrys issue on healthcare DNC negative ad Bushs healthcare

PAGE 82

82 APPENDIX C POLITICAL INFORMATION EFFICACY I consider myself well-qualifie d to participate in politics I think I am better informed about gove rnment and politics than most people I feel that I have a pretty good understanding of the important political issues facing our country If a friend asked me about the presidential el ection, I feel I would have enough information to help my friend figure out who to vote for

PAGE 83

83 APPENDIX D POLITICAL CYNICISM Vote has no influence on what politicians do One never knows what politicians really think People like me dont have any sa y about what government does Politicians and government seem co mplicated for a person like me One can be confident politicia ns will always do right thing Politicians often forget elect ion promises after campaign Politicians are more interested in power than what people think One cannot always trust what politicians say

PAGE 84

84 APPENDIX E MEDIA BIAS News organizations, such as newspapers a nd television news, try to manipulate public opinion News organizations often fail to get all of the facts straight. News organizations often don't deal fairly with all sides of a politic al or social issue. News organizations do a poor job of separating facts from opinions. News organizations are concerned with the community's well-being. News organizations watch out for my interests. News organizations are concerned mainly about the public welfare.

PAGE 85

85 APPENDIX F MEDIA CONSUMPTION Level of media exposure of presidential campaign in past week Level of talk with othe r people about presidential campaign in past week

PAGE 86

86 LIST OF REFERENCES Agger, R.E., Goldstein, M.N., & Pearl, S.A. (1961). Political cynicism: Measurement and meaning. The Journal of Politics, 23(3), 477-506. Almond, G. A., & Verba, S. (1963). The civic culture Princeton: Princeton University Press. Ansolabehere, S., & Iyengar, S. (1995). Going negative: How attack ads shrink and polarize the electorate NY: Free Press. Ansolabehere, S., Iyengar, S., Simon, A., & Va lentino, N. (1994). Does attack advertising demobilize the electorate? American Political Sc ience Review, 88(4), 829-838. Arensberg, C. M. (1961). The commun ity as object and as sample. American Anthropologist 63(2), 241-264. Atkin, C., & Heald, G. (1976). Ef fects of political advertising. The Public Opinion Quarterly 40(2), 216-228. Barlett, D. L., Drew, P. B., Fahle, E. G., & Watts, W. A. (1974). Selective exposure to a Presidential campaign appeal. The Public Opinion Quarterly, 38(2), 264-270. Barnes, B. E., & Thomson, L. M. (1994). Power to the people (meter): Audience measurement technology and media specialization. In J.S. Ettema and D.C. Whitney (eds.), Audience making: How the media create the Audience (pp.75-94). Thousand Oaks, CA: Sage Publications. Bartels, L. M. (1996). Uniformed votes: Info rmation effects in presidential elections. American Journal of Political Science, 40(1), 194-230. Barwise, P., & Ehrenberg, A. (1988). Television and its audience London: Sage. Bassili, J. N. (1993). Response latency versus cer tainty as indexes of the strength if voting intentions in a cati survey. Public Opinion Quarterly, 57(1), 54-61. Benoit, W.L., McKinney, M.S., & Holbert, R.L. (2001). Beyond learning and persona: Extending the scope of presid ential debate effects. Communication Monographs, 68(3) 259. Benoit, W. L., Hansen, G. J., & Holbert, R. L. (2004). Presidential campaigns and democracy. Mass Communication & Society, 7(2), 177-190. Berman, E. M. (1997). Dea ling with cynical citizens. Public Administration Review, 57(2), 105112. Blalock, H. (1972). Social statistics NY: McGraw-Hill. Burden, B. C., & Kimball, D.C., (2000). Why Americans split their tickets: Campaigns, competition, and divided government Ann Arbor, MI: The Univer sity of Michigan Press.

PAGE 87

87 Campbell, A, Converse P. E., Miller, W. E., & Stokes, D. E. (1960). The American Voter NY: John Wiley. Cappella, J. N., & Jamieson, K. H. (1997). Spiral of cynicism: The press and the public good. NY: Oxford University Press. Chaffee, S. (1978). Presidential deba tes Are they useful to voters? Communication Monographs, 45. Chaffee, S. H., Saphir, M.N.; Graf, J., Sandvig, C., & Hahn K.S. (2001). Attention to counterattitudinal messages in a state election campaign. Political Communication, 18, 247-272. Circle (2004). Youth voting in the 2004 election Fact Sheet. Retrieved on March 10, 2005 from http://www.civicyouth.org/ Clarke, H. D., & Acock, A. C. (1989). National el ections and political at titudes: The case of political efficacy. British Journal of Polit ical Science, 19(4), 551-562. Coleman, J. J. (1996). Party Organizational Strength and Public Support for Parties. American Journal of Political Science, 40(3), 805-824. Converse, P. E. (1962). Information flow a nd the stability of pa rtisan attitudes. Public Opinion Quarterly, 26 (4), 578-599. Converse, P. E. (1964). The nature of belief systems in mass publics. In D.E. Apter (ed.), Ideology and discontent (pp.206-261). NY: Free Press Dahl, R. A. (1961). Who Governs? Democracy and Power in an American City. New Haven: Yale University Press. Dalton, R. J., Beck, P. A., & Huckfeldt, R. ( 1998). Partisan cues and the media: information flows in the 1992 Presidential election. American Political Science Review, 92(1), 111-126. Damore, D. F. (2004). The Dynamics of Issue Ownership in Presidential Campaigns. Political Research Quarterly, 57(3) 391-397. Declare Yourself (2003, November 13). Major na tional survey shows increased promise, new strategies for increasing yout h vote. Youth vote 2004 surv ey shows narrow window of opportunity to engage young voter s; points to political incompetence as major barrier to voting. Retrieved on June 20, 2005 from http://www.declareyourself.org/press/pressroom.htm. Deli Carpini, M. X. (2000). Gen.com: youth, ci vic engagement, and the new information environment. Political Communication, 17, 341-349. Dennis, J. (1967, March) The child's acquis ition of regime norms: political efficacy. American. Political Science Review.

PAGE 88

88 Denton, R. E., & Woodward, G. C. (1990). Political Communication in America Westport, CT: Praeger. Druckman, J. N., & Nelson, K. R. (2003). Framing and deliberation: How Citizens Conversations Limit Elite Influence. American Journal of Politic al Science, 47 (Oct.), 729-745 Eagly, A., & Chaiken, S. (1993). Psychology of Attitudes NY: Harcourt, Brace Jovanovich. Easton, D. (1965). A Systems Analysis of Political Life NY: John Wiley. Fairvote (2004). Who picks the president? Takoma Park, MD: The Center for Voting and Democracys Presidential Elections Reform Program. Retrieved on March 15, 2005 from http://www.fairvote.org/presidential/ Freedman, P., Franz, M., & Goldstein, K. (2004). Campaign Advertising and Political Citizenship. American Journal of Politic al Science, 48 (Oct.), 723-741 Erikson, R. S., & Tedin, K. L. (2007). American Public Opinion: Its Origins, Content, and Impact NY: Pearson Longman. Eveland, W.P., & Shah, D. V. (2003). The impact of individual and in terpersonal factors on perceived news media bias. Political Psychology, 24(1), 101-117 Festinger, L. (1957). A theory of cognitive dissonance Stanford, CA: Stanford University Press Fischer, T., Jonas, E., Fray, D., and Schulz-Hardt, S. (2005). Selective exposure to information: The impact of information limits. European Journal of Social Psychology, 35, 469-492 Fiske, S. T., & Taylor, S. E. (1991). Social Cognition (2nd Ed.). New York: McGraw-Hill. Funk, C. L., (1999). Bringing the candidate into models of candidate evolutions. The Journal of Politics, 61(3), 700-720 Gamson, W. A., & Modigliani, A. (1989). Medi a Discourse and Public Opinion on Nuclear Power: A Constructionist Approach. The American Journal of Sociology, 95(1), 1-37. Garnett, J. L. (1992). Communicating for results in Governm ent: A Strategic approach for public managers San Francisco, CA: Jossey-Bass. Gordon, A., & Miller, J. L. (2004). Values and persuasi on during the first Bush-Gore Presidential debate. Political Communication, 21, 71-92 Gunther, A.C., & Schmitt, K. M. (2004). Mapping boundaries of the hostile media effect. Journal of Communication, 54, 55-70. Gussin, P., & Baum, M. (2004). In the Eye of the Beholder: An Experimental Investigation into the Foundation of the Hostile Media Phenomenon. Presented at the American Political Science Association, Chicago.

PAGE 89

89 Hall, S. (1980). Coding and encoding in the te levision discourse. In S, Hall et al. (eds.), Culture, media, language London: Hutchinson. Harris, R. J., (1989). A cognitive Psychology of Mass Communication NJ: Lawrence Erlbaum Associates. Heesacker, M., Petty, R.E., & Caci oppo, J. T. (1983). Field depe ndence and attitude change: Course dependence and att itude change: affecting me ssage-relevant thinking. Journal of personality and social Psychology, 51, 653-666 Hofstetter, C. R., Shultze, W. A., & Mulvihill, M. M. (1992). Communications media, public health, and public affaires: Expos ure in a multimedia communality. Health Communication, 4, 259-271 Hovland, C. I. (1959). Reconciling conflicting re sults derived from experimental and survey studies of attitude change. American Psychologist, 14, 8-17 Hovland, C. I., Campbell, E.H., & Brock, T. C. (1957). The effects of commitment on opinion change following communication In C. I. Hovland, et al., The order of presentation on persuasion (pp.22-32). New Ha ven, CT: Yale University Press. Hovland, C. I., Irving, J., and Kelley, H.H. (1953). Communication and Persuasion New Haven, CT: Yale University Huckfeldt, R., Beck, P. A., Dalton, R. J., & Levi ne, J. (1995). Political Environments, Cohesive Social-Groups, and the Communi cation of Public-Opinion. American Journal of Political Science, 39(4), 1025-1054. Hyman, H. H., & Sheatsley, P.B. (1947). Some reason why information campaigns fail. Public Opinion Quarterly, 11 412-423. Iyengar, S., (1991). Is Anyone Responsible?: How Television Frames Political Issues MA: University of Chicago Press Iyengar, S., Hahn, K., & Prior, M. (2001). Has Technology Made Attention to Political Campaigns More Selective? An Experimental Study of the 2000 Presidential Campaign Prepared for Presentation at the Annual Me eting of the American Political Science Association, San Francisco. Iyengar, S., & Kinder, D. R. (1987). News That Matters Chicago: University of Chicago Press. Jasperson, A., & Yun, H. J. (2007). Political Advertising Effects and Americas Racially Diverse Newest Voting Generation. American Behavioral Scientist, 50(9), 1112-1123 Kaid, L. L. (1995). Measuring candidate images with semantic differentials. In K. Hacker (Ed.), Candidate images in Presid ential election campaigns (pp. 131-134). Westport, CT: Praeger.

PAGE 90

90 Kaid, L. L.(Ed.), (2004). The Handbook of Political Communication Research Mahway, NJ: Lawrence Erlbaum Associates, Publishers Kaid, L.L., & Holtz-Bacha, C. (1995). Political advertising in Western democracies. Thousand Oaks, CA: Sage. Kaid, L.L., Landreville, K., Postelnic u, M., and Martin, J. (2005, May). Enhancing information efficacy for young voters: The effects of political advertising and debates. Paper presented at the International Communication Asso ciation Conference, New York City, NY Kaid, L. L., McKinney, M. S. & Tedesco, J. C. (2000). Civic dialogue in th e 1996 presidential campaign: Candidate, media, and public voices Cresskill, NJ: Hampton Kaid, L. L., McKinney, M. S., & Tedesco, J. C. (2007). Political information efficacy theory and younger voters. American Behavioral Scientist 50(9), 1093-1111 Kaid, L.L., Postenicu, M., Landrevill, K., Yun, H. J., and Hendren, A. G. (2007). The Effects of Political Advertising on Young Voters. American Behavioral Scientists, 50(9),1137-1151 Kamieniecki, S. (1988). The Dimensionality of Partisan Strength and Political Independence. Political Behavior 10(4), 364-376. Katz, E., & Feldman, J.J. (1962). The debates in light of research: A survey of surveys. In S. Kraus (ed.), The great debates: Kennedy vs. Nixon, 1960 (pp.173-223). Bloomington: Indiana University Press. Klapper, J. T. (1960). The Effects of Mass Communication New York: Free Press. Koch, J. W. (1998). Political Rhetoric and Po litical Persuasion: The Changing Structure of Citizens Preferences on Health Insurance During Policy Debate. Public Opinion Quarterly, 62, 209-229 Krosnick, J. A., Boninger, D.S., Chuang, Y. C., Berent, M.K., & Carnot, C. G. (1993). Attitude Strength: One Construct or many related constructs? Journal of Personality and Social Psychology, 65 1132-1151 Kubey, R., & Csikszentmihalyi, M. (1991). Television and the Quality of Life Mahwah, NJ: Lawrence. Lane, R. E. (1969). Political Life: Why People Get Involved in Politics NY: Free Press. Langton, K. P., & Karns, D. A. (1969). The Rela tive Influence of the Family, Peer Group, and School in the Development of Political Efficacy The Western Political Quarterly, 22(4), 813-826. Layman, G. C., & Carsey, T. M. (2002). Party Polarization and Party Structuring of Policy Attitudes: A Comparison of Three NES Panel Studies. Political Behavior, 24 (3), 199-236.

PAGE 91

91 Lazarsfeld, P.F., Berelson, B., & Guadet, H. (1948). The peoples Choice NY: Columbia University Press Leighley, J.E. (2004). Mass Media and Politics: A Social Science Perspective Boston, MA: Houghton Mifflin Company. Levin, M. B., & Eden, M. (1962). Politi cal Strategy for the Alienated Voter. Public Opinion Quarterly, 26(1), 47-63. Lowin, A., (1967). Approach and Avoidance: Alternative Modes of Selective Exposure to Information. Journal of Personality and Social Psychology, 6, 1-9 Mann, S. (1997). What the survey of American colle ge freshman tells us a bout their interest in politics and political science. Political Studies, 32 262-268. Marlantes, L. (2004). Inside red-and blue America. Christian Science Monitor July 14. McClelland, G. H., & Judd, C. M. (1993). The sta tistical difficulties of. detecting interactions and moderator effects. Psychological Bulletin 114(2), 376-390. McDonald, D. G. (1990). Media orient ation and television news viewing. Journalism Quarterly 67(1), 11-20 McDonald, D. G. (2004). Twentieth-Century Media Effects Research. In John Downing, The Sage Handbook of Media Studie s. Thousand Oaks, CA: Sage Publications, Inc. McCombs, M. (1994). News influe nce on our pictures of the worl d. In Jenings Bryant and Dolf Zillman et al., Media Effects Advances in Theory and Research UK: Lawrence Erlbaum Associates, Publishers McCombs, M., & Shaw, D. (1972) The Ag enda-Setting Function of Mass Media. Public Opinion Quarterly 36(2), 176-187 McDowell, S. D., & Lee, J. (2006). Tracking L ocalism in Television Broadcasting: Utilizing and Structuring Public Information. In Napoli, P.M., Media Diversity and Localism: Meaning and Metrics Mahwah, NJ: Lawrence Erlbaum Associates McGuire, W. J. (1968). Personality and attitude change: An information processing theory. In A. G. Greenwald, T. C. Brock, and T. M. Ostrom (Eds.), Psychological foundations of attitudes (pp. 171-196). San Diego, CA: Academic Press. McGuire, W. J. (1985). Attitudes and attitude change. In G. Lindzey and E. Aronson (Eds.), The handbook of social psychology (vol. 2), (pp. 233-246). NY: Random House. McQuail, D. (1997). Audience Analysis Thousand Oaks, CA: SAGE Publications Mendelsohn, H. (1973). Some reasons why information campaigns can succeed. Public Opinion Quarterly 37, 50-61

PAGE 92

92 Milburn, M. A. (1979). A Longitudinal Test of the Selective Exposure Hypothesis. Public Opinion Quarterly 43(4), 507-517 Miron, D., & Bryant, J. (2007). Mass Media and Voter Turnout In Preiss et al. (eds.), Mass Media Effects Research Mahwah, NJ: Lawrence Erlbaum Associates. Mutz, D. C. (1998). Impersonal influence: How perceptions of mass collective affect political attitudes Cambridge, UK: Cambridge University Press Mutz, D. C., & Martin P.S. (2001). Facil itating Communication across Lines of Political Difference: The Role of Mass Media. The American Political Science Review 95(1), 97114. Mutz, D.C., Sniderman, P. M., & Brody, R. A. (1996). Political persuasion and attitude change. Ann Arbor: University of Michigan Press, 1-14. Nelson, T. E., Oxley, Z. M., & Clawson, R. A. ( 1997). Toward a psychology of framing effects. Political Behavior 19(3), 221-246 Neuman, R.W. (1991). The Future of the Mass Audience NY: Cambridge University Press Niemi, R. G., Craig, S. C., & Mattei, F. (1991). Measuring internal political efficacy in the 1988 national election study. American Political Science Review 85(4), 1407-1413. Nightungale, V. (2004). Contemporary Television Audiences: Publics, Markets, Communities, and Fans. In Downing, J. et al. (eds.), The Sage Handbook of Media Studies (pp. 227-250). Thousand Oaks: Sage Publications Payne, J. G., Golden, J. L., Marlier, J., and Ratzan, S. C. (1989). Perceptions of the 1988 Presidential and Vice Presidential Debates. The American Behavioral Scientist 32(4), 425-435. Perloff, R. M., (2002). Political Campaign: Persuasion and its Disc ounts. In J. P. Dillard & M. Pfau (Eds.), The persuasion Handbook: Developments in theory and Practice (pp.605632). London: SAGE Publications, Petrocik, J. R. (1997). Issue Ownership in Pr esidential Elections, w ith a 1980 Case Study. American Journal of Political Science, 40(3), 825-850. Petty R. E., and Krosnick, J. A., (1995). Attitude Strength: Antecedents and Consequences Mahwah, NJ: Lawrence Erlbaum Associates. PEW Research Center for the Peop le & the Press. (2006). Publics News Habits Little Changed by September 11. Retrieved on Nov. 14, 2006 from http://peoplepress.org/reports/display.php3?PageID=616

PAGE 93

93 Rahn, W. M. (1993). The role of partisan stereo types in information processing about political candidates. American Journal of Political Science, 37(2), 472-496 Rogers, E. M., & Dearing, J. W. (1994). Agenda-set ting research: Where has it been, where is it going? In Doris A. Graber (3rd Ed.), Media Power in Politics Washington, DC: Congressional Quarterly Press. Rokeach, M., Rothman, G. (1965). The princi ple of belief congruence and the congruity interaction. Psychological Review, 72, 128-142 Romero D. W. (1996). The Case of the Missing R eciprocal Influence: Incumbent Reputation and the Vote. The Journal of Politics, 58(4), 1198-1207. Rosenberg, M. (1955). Some Dete rminants of Political Apathy. The Public Opinion Quarterly 18(4), 349-366 Rosenstone, S. J., Kinder, D. R., & Miller, W. E. (1997). American National Election Study, 1996: Preand Post-Election Survey (computer file). Ann Arbor, MI: University of Michigan, Center for Political Studies (producer), Inte r-university Consortium for Political and Social Research (distributor). Rudolph T. J., Gangl, A., & Stevens, D. (2000) The Effects of Efficacy and Emotions on Campaign Involvement. The Journal of Politic, 62(4), 1189-1197. Sears, D. O., & Freedman, J. L. (1967). Selectiv e Exposure to Information: A Critical Review. Public Opinion Quarterly, 31(2), 194-213. Schmitt, K. M., Gunther, A.C., & Liebhart, J. L. (2004). Why partisans see mass media as biased. Communication Research, 31(6), 623-641. Schooler, C., Chaffee, S. H., Flora, J. A., & Roser, C. (1998). Health campaign channels: Tradeoffs among reach, specificity, and impact. Human Communication Re search, 24(3), 410-432. Schultz, D. A. (2000). Its show time!: Media, politics, and popular culture NY: Peter Lang. Schumpeter, J. (1993). Capitalism, socialism, and democracy. In P. Green (ed.), Democracy City. NJ: Humanities Press. Semetko, H. A. (2004). Media, Public Opinion, ad Public Action. In J. Downing, et al. (eds.), The Sage Handbook of Media Studies (pp.351-374). Thousand Oaks, CA: Sage Publications. Shaw, D. R. (1999). The Effect of TV Ad s and Candidate Appearances on Statewide Presidential Voters, 1988-96. The American Political Science Review, 93(2), 345-361 Shaw, D. L., & McCombs, M. (Eds.). (1977). The emergence of American political issues St. Paul, MN: West.

PAGE 94

94 Shearer, E. (2005). Reflecti ons on civic engagement. Nonv oters: Who are Americas NoShows? In Mitchell S. McKinney, Lynda L. Ka id, Dianne G. Bystrom, and Diana B. Carlin (Eds.), Communicating politics: Engaging the public in democratic life (pp. 87 91). New York, NY: Peter Lang Publishing. Sigelman, L., & Gant, M. M. (1989). Anticandi ate voting in the 1984 presidential election. Political Behavior, 11(1), 81-92 Source Watch (2005). U.S. Presidential Election, 2004: Campaign Issues A Project of the Center for Media and Democracy. Retrieved on June 11, 2007 from http://www.sourcewatch.org /index.php?title=U.S._presiden tial_election%2C_2004:_Camp aign_Issues. Steenbergen, M. R. & Jones, B.S. (2002). Modeling multiple date structures. American Journal of Political Science, 46(1), 218-237 Stipak, Brian (1977). Attitudes and Belief Systems Concerning Urban Services. Political Opinion Quarterly, 41(Spring) 41-55. The Center for Voting and Democracys Presidentia l Elections Reform Program (2005). Fairvote. Retrieved on Feb. 26, 2005 from http://www.fairvote.org/presidential/ Third Millenium (1999, December). Young adults largely ignori ng presidential campaign Press release. Retrieved on March 10, 2005 from http://www.vanishingvoter.org/VV2000/ Press_Releases/12-28-99.shtml. USA Today. 2004 Election Campai gn. Retrieved on Nov. 24, 2005 from http://www.usatoday.com/news/politicselections/vote2004/NJ.htm. Vallone, R., Ross, L., & Lepper, M. R. 1985. The Hostile Media Phenomenon: Biased Perception and Perceptions of Media Bias in Coverage of the Beirut Massacre. Journal of Personality and Social Psychology, 49 577. Wheeler, K. M., (Ed.) (1994). Effective Communication. Muni cipal Management Series. Washington, D.C.: International City Management Association. Wright, G. C. (1998). Level-of-analysis effect s on explanations of voting: The case of the 1982 U.S. Senate elections. British Journal of Political Science 19, 381-398 Zaller, J. (1991). Inform ation, values and opinion. American Political Sc ience Review, 85(4), 1215-1238 Zaller, J. (1992). The nature and origins of mass opinion NY: Cambridge University Press.

PAGE 95

95 BIOGRAPHICAL SKETCH Hyun Jung Yun grew up in South Korea and completed her undergraduate work majoring in Political Science at Ajou University. Starti ng with her undergraduate senior year as an exchange student, she continued her masters de gree studying in political science and doctorate in the fields of political sc ience and journalism and communi cations at the University of Florida in the United States of America Her research interests are in political perception, the poli tical communication process, policy attitudes, persuasion, and geopolitics across different levels of the individual, small group, and aggregate group. More specifically, her rese arch in the field of political communication explores the relationship between political in formation process and individuals political attitudes in different geopolitical circumstance. In the same line of interdisciplinary research, her research in political science investigates how individuals beliefs a bout various policies are influenced by varying levels of multi-dimensi onal social capital and communication networks. Her research in journal publications demons trates how individuals political perceptions and attitudes are influenced by pol itical predispositions within a group and by political resources within a given political and media system at the aggregate level. In addition, she had coauthored several book chapters examining news coverage of policy issues and political candidates across different political regions to obs erve the relationship between di fferent political characteristics and political information effects. She is cu rrently working on analyz ing the dual spirals of silence in policy opinion form ation between issue minority an d issue majority, effects of relationship between media and pol itics on voter perceptions, as we ll as political cynicism and information efficacy in young voters. She also has participated in grant-suppor ted research projects including the Florida Department of Healths 2004 Project in Medi a Terrorism Preparation under Dr. Mary Ann

PAGE 96

96 Ferguson and Dr. Lynda Lee Kaid, Uvote inter-un iversity research on U. S. elections under Dr. Lynda Lee Kaid, and United States Election Assistance Commissions project establishing election law database under Dr. Lynda Lee Kaid a nd Dr. Cliff Jones The former project dealt with media advocacy, government public informa tion, and issue management on terrorism. The Uvote research has focused on political advocac y and political information effects. The development of electronic database of U.S. elect ion laws intended to provide U.S. citizens easy internet search function for comprehensive U.S. el ection law. She has worked for these projects as a data analyst and project manager. She also worked as data archiving assist ant for ICPSR (Inter-university Consortium for Political and Social Research) for a part of gr aduate assistantship duty. She was trained in advanced methodology through ICPSR as well as by the departments of statistics, political science, and journalism and mass communications Her methodological training across different fields includes managements of da ta through various applications a nd various levels of statistics such as linear regression, categorical anal ysis, multivariate analysis, maximum likelihood analysis, game theory, content anal ysis, scaling, and measurement. Her next project is to collect linearly coherent multi-level data that li nks individual perceptions, attitudes, and pr eferences with the aggregate level of media and political predispositions in different pol itical regions, election turnouts, policy efficiency, and other media-politics routines in order to conduct research with theoreti cally reliable connections across different levels of dynamics. Hyun Jung Yun who has two doctorate degrees, one in political scien ce and the other in journalism and communications, will work as an assistant professor at Texas State University starting from August 2007.