News Bias Research Articles

News source bias and sentiment on social media.

As social media becomes a key channel for news consumption and sharing, proliferating partisan and mainstream news sources must increasingly compete for users' attention. While affective qualities of news content may promote engagement, it is not clear whether news source bias influences affective content production or virality, or whether any differences have changed over time. We analyzed the sentiment of ~30 million posts (on twitter.com) from 182 U.S. news sources that ranged from extreme left to right bias over the course of a decade (2011-2020). Biased news sources (on both left and right) produced more high arousal negative affective content than balanced sources. High arousal negative content also increased reposting for biased versus balanced sources. The combination of increased prevalence and virality for high arousal negative affective content was not evident for other types of affective content. Over a decade, the virality of high arousal negative affective content also increased, particularly in balanced news sources, and in posts about politics. Together, these findings reveal that high arousal negative affective content may promote the spread of news from biased sources, and conversely imply that sentiment analysis tools might help social media users to counteract these trends.

Unveiling the hidden agenda: Biases in news reporting and consumption

Abstract Recognizing the presence and impact of news outlets’ biases on public discourse is a crucial challenge. Biased news significantly shapes how individuals perceive events, potentially jeopardizing public and individual well-being. In assessing news outlet reliability, the focus has predominantly centered on narrative bias, sidelining other biases such as selecting events favoring specific perspectives (selection bias). Leveraging machine learning techniques, we have compiled a six-year dataset of articles related to vaccines, categorizing them based on narrative and event types. Employing a Bayesian latent space model, we quantify both selection and narrative biases in news outlets. Results show third-party assessments align with narrative bias but struggle to identify selection bias accurately. Moreover, extreme and negative perspectives attract more attention, and consumption analysis unveils shared audiences among ideologically similar outlets, suggesting an echo chamber structure. Quantifying news outlets’ selection bias is crucial for ensuring a comprehensive representation of global events in online debates.

Reported speech and gender in the news: Who is quoted, how are they quoted, and why it matters

News stories have a well-defined generic structure, consisting of components such as headline, lede, and body, with reported speech a prominent feature, especially in hard news stories. Reported speech serves multiple purposes, from providing evidentiality and intertextuality to contributing to the construction of newsworthiness and to the context creation of news. It is also a site of potential bias in who is cited and how, including with respect to the gender of sources. Using a large corpus of English-language news stories for all of 2023 from the main five mainstream news outlets in Canada (over 370,000 articles from news websites), I examine the gender distribution of those quoted, the syntactic variation in the structure of quotes, and the types of reporting verbs. The study provides a comprehensive overview of the extend of gender bias in contemporary Canadian news, at the same time offering insights into the nature of reported speech in modern news and how it endures and evolves, including in news meant for digital-only publication.

Konstrusi Media Tentang Perempuan Pelaku Kejahatan Predator Seksual Anak

This study seeks to reveal gender bias in the headline news of kompas.com concerning the case of a woman perpetrator of sexual crimes against 11 children in Jambi. The research utilizes Peter L. Berger's social construction theory, Erving Goffman's framing, and Pierre Bourdieu's symbolic violence. Employing Robert Entman's framing analysis model, the findings of this study uncover that the portrayal of women is divided into three constructs imbued with gender bias. Gender bias arises due to the practice of symbolic violence. First, the female perpetrator of sexual crimes is depicted as engaging in unseemly, reprehensible actions that contravene societal values and norms. Second, the woman's vital body becomes a newsworthy asset that is advantageous. Third, the news exhibits ambivalence; on one hand, it provides information on preventing sexual crimes, while on the other hand, it inadvertently offers a tutorial-like narrative on committing sexual offenses against children.

News Bias in Financial Journalists’ Social Networks

ABSTRACTConnected financial journalists—those with working relationships, common school ties, or social media connections to company management—introduce a marked media slant into their news coverage. Using a comprehensive set of newspaper articles covering mergers and acquisition (M&A) transactions from 1997 to 2016, I find that connected journalists use significantly fewer negative words in their coverage of connected acquirers. These journalists are also more likely to quote connected executives and include less accurate language in their reporting. Moreover, they tend to portray other firms in the same network in a less negative light. Journalists’ favoritism bias has implications for both capital market outcomes and their careers. I find that acquirers whose M&As are covered by connected journalists receive significantly higher stock returns on the news article publication date. However, these acquirers’ stock prices reverse in the long term, suggesting market overreaction to news covered by connected journalists. Around M&A transactions, connected articles are correlated with increased bid competition and deal premiums. In terms of future career development, connected journalists are more likely to leave journalism and join their associated industries in the long run. Taken together, the evidence suggests that financial journalists’ personal networks promote news bias that potentially hinders the efficient dissemination of information.

Seasonality and regional disparity in attention to homelessness in UK newspapers between 2001 and 2020

ABSTRACT This research examined the amount of media attention to homelessness in 11 UK newspapers with two main research questions: (1) Does the media coverage exhibit seasonal cyclical patterns in their attention to homelessness? (2) Do England's nine regions differ in the attention they receive after adjusting for their relative homelessness levels? Time series of all 4105 news reports with the keyword “homeless*” in their headline between 2001 and 2020 were tested for meteorological and monthly seasonal effects, revealing significantly greater attention in winter but decreased coverage in spring and summer. Contrary to expectation, further analysis suggested that the increased winter coverage was not related to the colder months but to the Christmas period. Regarding regional differences, London received the highest mentions in the news but after adjusting for the relative burden of homelessness, the North West emerged as receiving the most relative coverage. The East Midlands and East of England had the lowest adjusted attention scores. As suggested by the public arenas model, the findings of this research suggest that media focus on social problems like homelessness be culturally patterned with key events like Christmas shaping issue-prominence. Similarly, regional imbalances in attention may reflect a systemic bias in news reporting.

The Demand for News: Accuracy Concerns Versus Belief Confirmation Motives

Abstract We examine the relative importance of accuracy concerns and belief confirmation motives in driving the demand for news. In experiments with US voters, we first vary beliefs about whether an outlet reports the news in a right-wing biased, left-wing biased or unbiased way. We then measure demand for a newsletter covering articles from this outlet. Right-wing voters strongly reduce their demand for left-wing biased news, but not for right-wing biased news. The reverse patterns hold for left-wing voters. These results suggest a trade-off between accuracy concerns and belief confirmation motives. We quantify this trade-off using a structural model and find a similar quantitative importance of both motives.

Evolving linguistic divergence on polarizing social media

Language change is influenced by many factors, but often starts from synchronic variation, where multiple linguistic patterns or forms coexist, or where different speech communities use language in increasingly different ways. Besides regional or economic reasons, communities may form and segregate based on political alignment. The latter, referred to as political polarization, is of growing societal concern across the world. Here we map and quantify linguistic divergence across the partisan left-right divide in the United States, using social media data. We develop a general methodology to delineate (social) media users by their political preference, based on which (potentially biased) news media accounts they do and do not follow on a given platform. Our data consists of 1.5M short posts by 10k users (about 20M words) from the social media platform Twitter (now “X”). Delineating this sample involved mining the platform for the lists of followers (n = 422M) of 72 large news media accounts. We quantify divergence in topics of conversation and word frequencies, messaging sentiment, and lexical semantics of words and emoji. We find signs of linguistic divergence across all these aspects, especially in topics and themes of conversation, in line with previous research. While US American English remains largely intelligible within its large speech community, our findings point at areas where miscommunication may eventually arise given ongoing polarization and therefore potential linguistic divergence. Our flexible methodology — combining data mining, lexicostatistics, machine learning, large language models and a systematic human annotation approach — is largely language and platform agnostic. In other words, while we focus here on US political divides and US English, the same approach is applicable to other countries, languages, and social media platforms.

Development of a Content Framework of Artificial Intelligence Integrated Education Considering Ethical Factors

The rapid advancement of Artificial Intelligence (AI) technology has brought about significant positive changes across society. However, it has also introduced challenges like privacy breaches, data bias, and spreading fake news. In response, several countries, including South Korea, have provided ethical guidelines and policies for AI. Yet, these measures often fall short of keeping pace with the speed and diversity of AI development. To address these issues, this study developed a comprehensive approach by integrating ethical considerations into AI education, covering the entire AI technology development and application process. It involved analyzing previous research on AI and AI ethics education and creating a draft of an integrated AI education program focused on problem-solving based on computational thinking and ethical practicality. The draft was refined and supplemented through two rounds of expert Delphi surveys, ultimately leading to an “Integrated AI Education Program.” This proposed program emphasizes that the knowledge content of AI technology and ethical considerations should not be treated separately but addressed together. It aims to enhance moral and social capacities alongside the experience of thinking like a computer scientist when solving real-life problems. Through this, the program supports the ultimate goal of AI education: to foster computational thinking while providing an educational experience that considers both technology and ethics. This study is expected to advance the discussion on integrating AI ethics into AI education curricula and contribute to developing socially responsible AI developers and users.

W obronie historii. Piotr Kowalewski Jahromi rozmawia z Antoonem De Baetsem

In this interview, conducted at the XXIII International Congress of Historical Sciences in Poznan, Antoon De Baets (emeritus professor of History, Ethics and Human Rights at the University of Groningen, the Netherlands) addresses key issues for historians and other produ-cers of history. His remarks about the scientific status of historiography and the range of different threats to history seem particularly important. He talks not only about the most direct crimes against historians and history, but also about issues like hindsight bias and fake news. The professional duties of historians and the issue of ethical codes for historians are also discussed.

News bias perceptions as impacted by source cues, content cues, and media bias ratings

ABSTRACT Using three experimental studies, this research considers whether exposure to source cues, content cues, and media bias ratings impacts perceptions of ideological bias in news stories. As a possible news media literacy intervention strategy, viewing a media bias ratings chart after reading a news story had limited influence on assessments of bias. Source cues consistently influenced bias perceptions. Study 1 and Study 2 revealed evidence of the relative hostile media perception, where Republicans and Democrats perceived the news stories as biased in the same direction but to a significantly different degree. This finding did not replicate in Study 3, where content cues were held constant. Thus, content cues, alongside source cues, appear necessary in facilitating the relative hostile media perception.

Crowdsourced Forecasts and the Market Reaction to Earnings Announcement News

ABSTRACT This study examines whether crowdsourced forecasts of earnings and revenues help investors unravel bias in earnings announcement news, which is commonly derived from analyst forecasts. Our results suggest that investors, on average, understand and price the predictive signals reflected in crowdsourced forecasts about the bias in analyst-based earnings and revenue surprises. Using the staggered addition of firms to the Estimize platform, we find that crowdsourced coverage is associated with reductions in the mispricing of forecast bias and declines in earnings announcement premia. We further find some evidence that managers use income-increasing accruals to meet the crowdsourced forecast benchmark and that they respond to crowdsourced coverage through increased downward earnings and revenue guidance. Overall, we conclude that user-generated content on crowdsourced financial information platforms helps investors discount biases in traditional equity research and thereby better process the news in earnings announcements. JEL Classifications: G14; G20; M41.

Computational approaches to developing the implicit media bias dataset: Assessing political orientations of nonpolitical news articles

Research on media bias has been primarily conducted a number of times of news outlets referred on political news articles, but nonpolitical articles can still convey media bias that indicates the political orientation of the news outlet. Using manual human evaluation and computational approaches, we developed and publicly released the Implicit Media Bias Dataset, which contains the political orientations of 24,576 news articles featuring nonpolitical events. News articles published in the information technology and science section of the two most biased Korean news outlets (the most conservative and the most progressive) were collected, and each article was manually evaluated by human annotators in terms of its objectiveness, fairness, unbiasedness, and political orientation. The results revealed significant differences between the articles from the conservative and progressive news outlets in these domains. Next, deep learning models trained with a large corpus of nonpolitical articles were used to identify the political orientations of the first set of articles. They achieved over 98% accuracy in classifying the articles as conservative or progressive. The findings of this study demonstrate the effectiveness of computational methods in identifying and analyzing diverse forms of polarization in society.

Exploring the impact of social media exposure patterns on people’s belief in fake news during COVID-19: A cross-gender study

During COVID-19, fake news on social media seriously threatened public health. As a solution to this problem, this study examined how social media exposure patterns affect people being deeply harmed by fake news. Based on cognitive dissonance theory, this study investigated the effect of intentional and incidental exposure on belief in fake news through the mediating role of confirmation bias. The results show that intentional exposure positively influences confirmation bias and belief in fake news. Incidental exposure is the opposite. Our results also show that intentional exposure and confirmation bias negatively influence incidental exposure. Furthermore, these relationships remain unchanged by gender. This study provides theoretical and empirical contributions to reducing people’s belief in fake news.

Quantify the Political Bias in News Edits: Experiments with Few-Shot Learners (Student Abstract)

The rapid growth of information and communication technologies in recent years, and the different forms of digital connectivity, have profoundly affected how news is generated and consumed. Digital traces and computational methods offer new opportunities to model and track the provenance of news. This project is the first study to characterize and predict how prominent news outlets make edits to news frames and their implications for geopolitical relationships and attitudes. We evaluate the feasibility of training few-shot learners on the editing patterns of articles discussing different countries, for understanding their wider implications in preserving or damaging geopolitical relationships.

Support for “doing your own research” is associated with COVID-19 misperceptions and scientific mistrust

Amid concerns about misinformation online and bias in news, there are increasing calls on social media to “do your own research.” In an abundant information environment, critical media consumption and information validation are desirable. However, using panel survey data, we find that positive perceptions toward “doing your own research” are associated with holding more misperceptions about COVID-19 and less trust in science over time. Support for “doing your own research” may be an expression of anti-expert attitudes rather than reflecting beliefs about the importance of cautious information consumption.

“Minimal” and “Biased”: An Intersectional Analysis of Female Candidates’ Perceptions of Their Local News Coverage

This qualitative study analyzes the perceptions of US women political candidates about their local news coverage at a pivotal moment of change in both journalism and electoral politics. Through an intersectional approach, in-depth interviews with thirty-seven women who recently ran for office from 2016 to 2020 revealed findings in two areas of news research: local news declines and gender bias in political news. First, massive declines in local news capacities have trickled into local political communication, which, along with affordances of social media, point to a decline in relevance of local media to local campaigns. Second, participants perceived bias in their news coverage based on an identity intersecting with their gender, partisan affiliation, or both. Their experiences offered two potential ways forward for local news to reattain relevancy in local elections in countries of local news declines: through accountability journalism, and by diversifying staff to meet the needs of a diversifying electoral slate of candidates.

The Influence of News Consumption Habits and Dispositional Traits on Trust in Medical Scientists.

Public trust in medical institutions is essential for ensuring compliance with medical directives. However, the politicization of public health issues and the polarized nature of major news outlets suggest that partisanship and news consumption habits can influence medical trust. This study employed a survey with 858 participants and used regression analysis to assesses how news consumption habits and information assessment traits (IATs) influence trust in medical scientists. IATs included were conscientiousness, openness, need for cognitive closure (NFCC), and cognitive reflective thinking (CRT). News sources were classified on the basis of factuality and political bias. Initially, readership of liberally biased news was positively associated with medical trust (p < 0.05). However, this association disappeared when controlling for the news source's factuality (p = 0.28), while CRT (p < 0.05) was positively associated with medical trust. When controlling for conservatively biased news sources, factuality of the news source (p < 0.05) and NFCC (p < 0.05) were positively associated with medical trust. While partisan media bias may influence medical trust, these results suggest that those who have higher abilities to assess information and who prefer more credible news sources have a greater trust in medical scientists.

Partisan Bias in Flu News and Its Impacts on Flu Vaccination Uptake in the U.S.

ABSTRACT Partisan gaps exist over flu vaccination. Guided by partisan media bias and motivated reasoning theories, this study examined partisan presentation bias in flu news and its impacts on flu vaccination uptake by analyzing 4150 news articles published in 48 U.S. newspapers from 2011 to 2018. Results revealed that partisan media did not utilize each news frame differently. Left-leaning media used flu news frames significantly differently and preferred to use the frames of risk population, flu prevention tips, and influenza caused deaths or illnesses stories. Neither least biased media nor right-leaning media used flu news frames significantly differently. Left-leaning media increased flu vaccine uptake among children and adults before adjusting for any confounders. After controlling for confounders, left-leaning media were no longer a significant predictor. Least biased media nor right-leaning media did not impact flu vaccine uptake among children and adults. Flu news articles published in all partisan media displayed seasonality.

A COMPARATIVE ANALYSIS ON THE FRAMING OF POLITICS IN JORDANIAN ONLINE NEWS

News bias has a detrimental impact on how the general public and specific individuals perceive the news information, resulting in a significant impact on Jordanian politics. Mainstream media sources continue to be the main sources of information about current events in Jordan, despite the internet's many advancements in the domains of communication. The widespread usage of news polarisation and the interference of the government through laws and policies are two additional reasons why it is regarded as good journalism in Jordan. Through a descriptive and comparative content analysis, this research applies theoretical fieldwork of communication and aims to comprehend and assess how the Al Hamzeh issue was framed in the Jordanian online news agencies, Al Ghad and Al Rai. The results of this study show that the official online news agency Al Rai gave more attention in its coverage of the Al Hamzeh issue than the independent online news Al Ghad. Al Ghad and Al Rai news pieces extensively implement the five different types of news frames in both their headlines and body copy. Additionally, the wording employed in the chosen news stories had a strong slant. The outcome regarding the independent news agency, Al Ghad, interpret contradiction anent news independence in Jordan. This study advances development of knowledge in the areas of press freedom, foreign policy, and communication. This study clarifies the understanding of Jordan's media landscape.