An official website of the United States government
In an increasingly information-dense web, how do we ensure that we do not fall for unreliable information? To design better web literacy practices for assessing online information, we need to understand how people perceive the credibility of unfamiliar websites under time constraints. Would they be able to rate real news websites as more credible and fake news websites as less credible? We investigated this research question through an experimental study with 42 participants (mean age = 28.3) who were asked to rate the credibility of various “real news” (n = 14) and “fake news” (n = 14) websites under different time conditions (6s, 12s, 20s), and with a different advertising treatment (with or without ads). Participants did not visit the websites to make their credibility assessments; instead, they interacted with the images of website screen captures, which were modified to remove any mention of website names, to avoid the effect of name recognition. Participants rated the credibility of each website on a scale from 1 to 7 and in follow-up interviews provided justifications for their credibility scores. Through hypothesis testing, we find that participants, despite limited time exposure to each website (between 6 and 20 seconds), are quite good at the task of distinguishing between real and fake news websites, with real news websites being overall rated as more credible than fake news websites. Our results agree with the well-known theory of “first impressions” from psychology, that has established the human ability to infer character traits from faces. That is, participants can quickly infer meaningful visual and content cues from a website, that are helping them make the right credibility evaluation decision.
more »
« less
“Fake News” Is Not Simply False Information: A Concept Explication and Taxonomy of Online Content
As the scourge of “fake news” continues to plague our information environment, attention has turned toward devising automated solutions for detecting problematic online content. But, in order to build reliable algorithms for flagging “fake news,” we will need to go beyond broad definitions of the concept and identify distinguishing features that are specific enough for machine learning. With this objective in mind, we conducted an explication of “fake news” that, as a concept, has ballooned to include more than simply false information, with partisans weaponizing it to cast aspersions on the veracity of claims made by those who are politically opposed to them. We identify seven different types of online content under the label of “fake news” (false news, polarized content, satire, misreporting, commentary, persuasive information, and citizen journalism) and contrast them with “real news” by introducing a taxonomy of operational indicators in four domains—message, source, structure, and network—that together can help disambiguate the nature of online news content.
more »
« less
- Award ID(s):
- 1742702
- PAR ID:
- 10121164
- Publisher / Repository:
- SAGE Publications
- Date Published:
- Journal Name:
- American Behavioral Scientist
- ISSN:
- 0002-7642
- Page Range / eLocation ID:
- Article No. 000276421987822
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Machine Learning-Based Identifications of COVID-19 Fake News Using Biomedical Information ExtractionThe spread of fake news related to COVID-19 is an infodemic that leads to a public health crisis. Therefore, detecting fake news is crucial for an effective management of the COVID-19 pandemic response. Studies have shown that machine learning models can detect COVID-19 fake news based on the content of news articles. However, the use of biomedical information, which is often featured in COVID-19 news, has not been explored in the development of these models. We present a novel approach for predicting COVID-19 fake news by leveraging biomedical information extraction (BioIE) in combination with machine learning models. We analyzed 1164 COVID-19 news articles and used advanced BioIE algorithms to extract 158 novel features. These features were then used to train 15 machine learning classifiers to predict COVID-19 fake news. Among the 15 classifiers, the random forest model achieved the best performance with an area under the ROC curve (AUC) of 0.882, which is 12.36% to 31.05% higher compared to models trained on traditional features. Furthermore, incorporating BioIE-based features improved the performance of a state-of-the-art multi-modality model (AUC 0.914 vs. 0.887). Our study suggests that incorporating biomedical information into fake news detection models improves their performance, and thus could be a valuable tool in the fight against the COVID-19 infodemic.more » « less
-
null (Ed.)In times of uncertainty, people often seek out information to help alleviate fear, possibly leaving them vulnerable to false information. During the COVID-19 pandemic, we attended to a viral spread of incorrect and misleading information that compromised collective actions and public health measures to contain the spread of the disease. We investigated the influence of fear of COVID-19 on social and cognitive factors including believing in fake news, bullshit receptivity, overclaiming, and problem-solving—within two of the populations that have been severely hit by COVID-19: Italy and the United States of America. To gain a better understanding of the role of misinformation during the early height of the COVID-19 pandemic, we also investigated whether problem-solving ability and socio-cognitive polarization were associated with believing in fake news. Results showed that fear of COVID-19 is related to seeking out information about the virus and avoiding infection in the Italian and American samples, as well as a willingness to share real news (COVID and non-COVID-related) headlines in the American sample. However, fear positively correlated with bullshit receptivity, suggesting that the pandemic might have contributed to creating a situation where people were pushed toward pseudo-profound existential beliefs. Furthermore, problem-solving ability was associated with correctly discerning real or fake news, whereas socio-cognitive polarization was the strongest predictor of believing in fake news in both samples. From these results, we concluded that a construct reflecting cognitive rigidity, neglecting alternative information, and black-and-white thinking negatively predicts the ability to discern fake from real news. Such a construct extends also to reasoning processes based on thinking outside the box and considering alternative information such as problem-solving.more » « less
-
In recent years, the emergence of fake news outlets has drawn out the importance of news literacy. This is particularly critical in social media where the flood of information makes it difficult for people to assess the veracity of the false stories from such deceitful sources. Therefore, people oftentimes fail to look skeptically at these stories. We explore a way to circumvent this problem by nudging users into making conscious assessments of what online contents are credible. For this purpose, we developed FeedReflect, a browser extension. The extension nudges users to pay more attention and uses reflective questions to engage in news credibility assessment on Twitter. We recruited a small number of university students to use this tool on Twitter. Both qualitative and quantitative analysis of the study suggests the extension helped people accurately assess the credibility of news. This implies FeedReflect can be used for the broader audience to improve online news literacy.more » « less
-
Recently, deepfake techniques have been adopted by real-world adversaries to fabricate believable personas (posing as experts or insiders) in disinformation campaigns to promote false narratives and deceive the public. In this paper, we investigate how fake personas influence the user perception of the disinformation shared by such accounts. Using Twitter as an exemplary platform, we conduct a user study (N=417) where participants read tweets of fake news with (and without) the presence of the tweet authors' profiles. Our study examines and compares three types of fake profiles: deepfake profiles, profiles of relevant organizations, and simple bot profiles. Our results highlight the significant impact of deepfake and organization profiles on increasing the perceived information accuracy of and engagement with fake news. Moreover, deepfake profiles are rated as significantly more real than other profile types. Finally, we observe that users may like/reply/share a tweet even though they believe it was inaccurate (e.g., for fun or truth-seeking), which could further disseminate false information. We then discuss the implications of our findings and directions for future research.more » « less
