Understanding the Role of Nonverbal Tokens in the Spread of Online Information

Abstract

Individuals and society continue to suffer as the fake news infodemic continues unabated. Current research has focused largely on the verbal part (plain text) of fake news, the nuances of nonverbal communication (emojis and other semiotic tokens) remain largely understudied. We explore the relationship between fake news and emojis in this work through two studies. The first study found that information with emojis is retweeted 1.28 times more and liked 1.41 times more than information without them. Additionally, our research finds that tweets with emojis are more common in fake news (49%) than true news (33%). We also find that emojis are more popular with fake news compared to true news. In our second study, we conducted an online experiment with true and fake news (N=99) to understand how the functional usage (replace/emphasize) of emoji affects the spread of information. We find that when an emoji replaces a verbal token, it is liked less (p<0.05) or equal to information without a nonverbal token (control condition), and when an emoji emphasizes a phrase, it is liked more or equal to the control condition. These effects are observed only for fake news. Functional usage of emojis did not affect the diffusion of true news in our study (p >0.05). © 2023 IEEE Computer Society. All rights reserved.