원문정보
초록
영어
This study focuses on importance of perceptions of election polls published by mass media and their effects on the support for election poll censorship. Two hypothetical models using SEM technique are tested against survey data from a sample of 422 residents of Ulsan metropolitan city in S. Korea. One model predicts that the biased media perception affects the support for censorship through mediating third person effect and sincere vote, while the other model predicts that horse race reporting affects the support for censorship through mediating third person effect and strategic vote. Two models fit well with the data. In both models the third person effects were found to play key roles in the mediation processes. We discussed several unexpected findings from the data analysis and their implications for the Korean Election Act, Article 108 by which the media publication of election poll results is prohibited during the last 6 days before the election date.
목차
1. THEORETICAL BACKGROUNDS
1.1 Election poll censorship
1.2 Biased vs. Horse race perceptions of election poll reporting
1.3 The third person effect: perceptual differences
1.4 Behavioral consequences and protectionism
2. HYPOTHETICAL MODELS
3. RESEARCH METHODS
3.1 Perceived election poll reporting: biased vs. horse race reporting perception
3.2 The third person effect of biased vs. horse race reporting perception
3.3 Sincere vote vs. strategic vote
3.4 The importance of electability
3.5 Support for election poll censorship
4. RESULTS
4.1 Results for data analysis for Model A
4.2 Test results for Model B
5. Moderating role of education level(MODEL B)
5.1 Dividing Sample by Education Level
5.2 Goodness-of-fit test indices for the hypothetical model by education level
5.3 Results from statistical significance tests of path coefficients
6. DISCUSSION
6.1 Diverging power of media perceptions
6.2 Suspicious paternalism explanation of strategic vote
6.3 Implications for Election Act Article 108
REFERENCES