Israel Palestine Conflict, The issue and historical context!
News consumption and the 2012 presidential election
1. News Consumption and the 2012
Presidential Election
Brad Forsythe
Communication 9520
Seminar in Media Effects
2. AIM & SIGNIFICANCE
• The aim of this research project is to determine if media news
consumption impacted voting behavior in the 2012
presidential election. Did those who consumed more news
vote at higher rates? If it did, did news consumption have a
role in deciding which presidential candidate a voter
supported?
• Previous research has found news exposure has an effect on
political behavior. But, what role did the media play in 2012?
This study is significant because it examines how media
consumption and potential bias affected voting behavior and
possibly the outcome of the 2012 presidential election.
3. PEW RESEARCH STUDY
• http://www.journalism.org/2012/11/02/winni
ng-media-campaign-2012/
4. LITERATURE REVIEW
• Aquino et al. (2009) & Latimer & Cotter (1985) found
news consumption is a significant predictor of voting.
• Eveland & Scheufele (2000) found news consumption
shrinks the “knowledge gap” between those with
high & low levels of education.
• Shaw (1999) found that media bias does have an
impact on voter preferences.
5. HYPOTHESES
& DESIGN
• H1: The more exposure one had to news sources during the
2012 campaign, the more likely he or she voted in the
election, controlling for partisanship and demographics.
• H2: Given Shaw’s (1999) research, as well as the Pew findings
of a slight pro-Obama media bias in 2012, increasing media
exposure should slightly increase the likelihood of an Obama
vote.
• I test these hypotheses using data from the 2012 American
National Election Studies survey.
• Using binary logistic regression, I test the media’s
impact on voting behavior in 2012.
6. H1: FINDINGS
Independent Variables
B
S.E.
Party Identification
Ideology
Race
Age
Religiosity
Education
Income
Gender
# of Children in Home
.006
-.077
.303
-.120
.101
-.575
-.187
-.191
-.007
.045
.027
.202
.026
.068
.096
.05
.171
.086
.886
.005**
.133
.000***
.138
.000***
.000***
.265
.933
1.006
.926
1.354
.887
1.106
.563
.829
.826
.993
Internet News Use
Television News Use
Newspaper Use
Radio News Use
-.098
-.100
.018
-.130
.039
.034
.037
.034
.011*
.003**
.622
.000***
.906
.904
1.018
.878
Interaction Term:
Age * Internet Use
.017
.007
.012*
1.017
Pseudo R2 = .26
N=1197
-2 Log Likelihood= 923.273
Significance
*p< .05
**p< .01
***p< .001
Exp(B)
7. H2: FINDINGS
Independent Variables
B
S.E.
Party Identification
Ideology
Race
Age
Religiosity
Education
Income
1.070
.152
1.693
.099
-.249
-.073
.019
.077
.047
.307
.039
.105
.126
.068
.000***
.001***
.000***
.011*
.018*
.563
.774
2.916
1.164
5.436
1.104
.780
.930
1.020
Gender
# of Children in Home
-.347
.504
.251
.136
.166
.000***
.707
1.656
Internet News Use
Television News Use
Newspaper Use
Radio News Use
-.173
.046
-.136
-.069
.052
.049
.05
.044
.001***
.351
.006**
.118
.841
1.047
.873
.933
Interaction Term:
Age * Internet Use
.006
.009
.479
1.006
Pseudo R2= .75
N=894
-2 Log Likelihood=494.686
Significance
*p< .05
**p< .01
***p< .001
Exp(B)
8. CONCLUSION &
IMPLICATIONS
• I found statistical evidence that supports both H1 and
H2 to varying degrees.
• The more people utilized internet, television, and radio
news in 2012, the more likely they were to vote,
controlling for partisanship and demographics.
• The more people used the internet and print sources
for news in 2012, the more likely they were to support
Barack Obama for president, with the same control
variables used to test H1.
• These findings add to the previous literature on how
news exposure and coverage impacts political behavior
and potentially election outcomes.