There is an interesting analysis according to which the prevalence of emotional payload in the headlines of popular news outlets has significantly increased in the last 20 years.
While neutrality and joy have greatly decreased, anger, fear, disgust, and sadness have increased significantly.
Rozado/Hughes/Halberstadt's research uses modern Transformer language models, fine-tuned for text classification, to automatically label the sentiment polarity and emotional charge of a large data set of news articles headlines (n=23 million). The set of news outlets analyzed was derived from the AllSides Media Bias Chart 2019 which lists 47 of the most popular news media outlets in the United States.
Emotional Clickbait - how Negative Sentiment in News Headlines Manipulates Public Opinion and Shapes Societal Behaviour.
The Power of Headlines: Making or Breaking Engagement
News headlines are the front door to the world of current affairs, setting the stage for our engagement with an article. Researchers suggest that headlines not only influence our decision to delve deeper into the article, but they also shape our understanding of the content, sometimes leading us towards specific interpretations.
The underbelly of this phenomenon is the play of sentiment and emotionality in the text. Research reveals that content loaded with emotional triggers, especially those arousing high emotions such as anger, is more likely to go viral.
Interestingly, the cycle of virality is not confined to just authentic news. Fake news, heavily peppered with emotional stimuli, tends to proliferate further and faster on social platforms3.
Consider Twitter, for instance, where emotional and moral language holds sway over virality. A study highlighted that each moral or emotional word used in a tweet propels its virality by a staggering 20 percent, on average. The implication is clear: tweets sparking strong, especially negative, emotions reach a wider audience than those with neutral tones.
In this backdrop, it's hardly surprising why news outlets might exploit this aspect to enhance the emotional saliency of their headlines. The formula is straightforward: more emotional content brings in increased user engagement, translating to more web traffic and, in turn, higher revenue.
However, this approach carries significant societal implications. News content, knowingly or unknowingly, influences public mood and shapes public opinion. The manipulation of emotions for financial gains could compromise news objectivity, swaying public sentiment and opinion in possibly harmful directions.
A look back in history brings us to the era of Yellow Journalism in the late 19th century. Publishers like Joseph Pulitzer and William Randolph Hearst manipulated headlines and used sensationalism to sell more newspapers. This strategy, while commercially successful, often led to the spread of misinformation and public unrest.
If media strongly sway public mood and opinion, what societal impact could changing media sentiment have? Are news outlets cognizant of their influence? Has our perceived reality truly declined as dramatically as reflected in news headlines over the past two decades? Or is it merely a perceived degradation? What does it suggest about the evolution of news outlets? And how can the audience shield themselves from toxic news media?
How could the future of news consumption look? The complexity of this question transcends common understanding. However, a more tangible question might be: how can the audience protect themselves? One potential solution could be the development of an app with several features:
Headline Analysis: The app could leverage natural language processing techniques to analyze headlines, identify potential negative bias or sensationalism, and provide a summary of key points without exaggerated language.
Tone Correction: By employing sentiment analysis, the app could detect overly negative language in news articles and headlines, and suggest more neutral phrasing.
Contextualization: To counter potential bias, the app could provide related articles or different perspectives on the same topic for better understanding.
Fact-Checking: The app could include fact-checking capabilities to ascertain the credibility of news articles and headlines.
Personalization and Education: Users could customize their bias threshold to match personal preferences while ensuring a balanced presentation of information. The app could also include resources on media literacy and critical thinking to help users effectively navigate biased news.
Neutralisation
AI transforms biased news into neutral text.
Browser
A plug-in protects you from all news read inside your browser.
Customization
Select which publishers you trust, and not.
Warnings
When news try to bias users, AI is able to detect this in many ways and warns.
Personalisation
Every publisher has it's own tone and style and should be treated individually.
Facts
Automated fact checking is a must in today's world.
Summaries
Summaries are usually less biased than articles.
Contextual
Contextual, topic-based sections to show belonging and complexity of news.