The Role of Lexical Characteristics in the Dissemination of Misinformation and Truthful News

Authors

Kira Victory, Boise State UniversityFollow
Iryna Babik, Boise State UniversityFollow

Document Type

Student Presentation

Presentation Date

4-19-2024

Faculty Sponsor

Dr. Iryna Babik

Abstract

Recently, there has been a growing concern surrounding misinformation, which spreads about six times faster than truthful information.¹ The purpose of this study is to examine how differences in text characteristics (e.g., stylistic, semantic, sentiment) contribute to the dissemination of truthful vs. false news, especially political news. Truthful political news stories, which are often neutral or negative in sentiment, are significantly more likely to be shared via email if they are positive in sentiment.^{2, 3} Fake news is less frequently neutral and more often exaggerates negative and positive sentiment.⁴ False news which includes a video is more likely to be reported as false, yet positive sentiment in the text decreases the likelihood of it being reported as false.⁵ Including emotion in the title or body of an articles increases the likelihood of virality for both true and fake news. However, the effect is stronger for fake news, especially regarding anger, fear, and anticipation.⁶ Readability, conversational style, grammar elements, and emotion can all impact the spread of true and false news.⁶ Informal style increases the dissemination of fake news, but not true news. Use of plural first person increases rates of diffusion of fake news, whereas use of second and third person reduces it.⁶ In contrast, true news that uses singular first person or singular third person have decreased rates of diffusion.⁶ Understanding these differences in text characteristics that contribute to the dissemination of fake vs. real news could improve organizations’ ability to detect misinformation before it reaches a large audience.

References

1. Soroush, V. Roy, D., & Aral, S. (2018). The spread of true and false news online. Science, 359(6380), 1146–1151. doi:10.1126/science.aap9559
2. Cui, L., Wang., S., & Lee, D. (2019) SAME: Sentiment-aware multi-modal embedding for detecting fake news. Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 41–48, doi:10.1145/3341161.3342894.
3. Kraft, P. W., Krupnikov, Y., Milita, K., Ryan, J. B., & Soroka, S. (2020). Social media and the changing information environment: Sentiment differences in read versus recirculated news content. Public Opinion Quarterly, 84
, 195–215. doi:10.1093/poq/nfaa015
4. Alonso, M. A., Vilares, D., Gómez-Rodríguez, C., & Vilares, J. (2021). Sentiment analysis for fake news detection. Electronics, 10, 1348. doi:10.3390/electronics10111348
5. Wang, S. (Ada), Pang, M., & Pavlou, P. A. (2022). Seeing is believing? How including a video in fake news Influences users’ reporting of fake news to social media platforms. MIS Quarterly, 46(3), 1323–1353. doi:10.25300/MISQ/2022/16296
6. Esteban-Bravo, M., Jiménez-Rubido, L. d. l. M., & Vidal-Sanz, J. M. (2024). Predicting the virality of fake news at the early stage of dissemination. Expert Systems with Applications, 248, 123390. doi:10.1016/j.eswa.2024.123390