Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Department of Applied Social Sciences | en_US |
| dc.contributor.advisor | Chen, Juan (APSS) | en_US |
| dc.creator | Shen, Jinliang | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/14710 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Supply and demand of media bias in the U.S. elections | en_US |
| dcterms.abstract | The media has a powerful impact on public opinion and public actions. In political elections, on the supply side, media bias matters because it shapes how political realities are presented to the public, influencing what information voters receive, and affecting voters’ preferences, thereby playing a critical role in the functioning of democracy. On the demand side, media bias is important because it not only reveals how citizens engage with political news but also affects the financial performance of media companies. This thesis seeks to provide insights into understanding both the supply and demand sides of media bias in the U.S. elections. | en_US |
| dcterms.abstract | In Study 1, I use entity sentiment toward presidential candidates in news coverage to construct a daily media bias index for 237 newspapers during the U.S. presidential elections from 1980 to 2020. This study contributes to the literature by introducing election closeness, measured by narrowing pre-election polling margins, as an important factor affecting newspapers’ strategies of media bias on the supply side. I find that the election news coverage was predominantly against Republican presidential candidates, regardless of whether the newspapers were left-leaning or right-leaning. This bias in both newspaper groups consistently decreases when pre-election polls become closer. These findings suggest a general liberal bias and the different strategic responses between left-and right-leaning newspapers. | en_US |
| dcterms.abstract | In Study 2, I construct a series of emotion and information variables for political news tweets from four prominent left-leaning and four right-leaning media outlets during the 2022 U.S. midterm elections, aiming to find out which variables are associated with higher retweet counts. This study contributes to the literature by applying an automated NLP-based method to measure the informational value of news articles, whereas previous research has relied on human coders. I also first apply facial emotion analysis to news emotion measurement, using SOTA techniques from the computer vision field. However, neither the descriptive visualizations nor the regression analyses reveal strong and consistent patterns; only overall emotion intensity and anger show relatively consistent positive effects across most outlets. | en_US |
| dcterms.abstract | In Study 3, I am the first to introduce the divergence between journalists and audiences in evaluating the informational value of news into research on political communication. This divergence distorts the understanding of audiences’ demand for informational value and increases financial pressures on media companies. By addressing the same research question as in Study 2 but adopting a survey experiment design, Study 3 overcomes the limitations inherent in Study 2’s observational data approach. The results show that respondents significantly prefer news they perceive as informational, while disfavoring news classified as informational based on journalistic criteria. This supports my argument that it is the audience’s perceived informational value—rather than objective informational value—that shapes media preferences. | en_US |
| dcterms.abstract | Overall, this thesis uses data science methods to measure media bias in the U.S. elections, highlights differences in bias strategies between left-and right-leaning outlets, and provides empirical evidence of audiences’ preferences for emotional and informational news based on Twitter observations and a survey experiment. | en_US |
| dcterms.extent | 158 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2026 | en_US |
| dcterms.educationalLevel | Ph.D. | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.accessRights | open access | en_US |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14710