Copy the page URI to the clipboard
Tavakoli, Mohammadali; Alani, Harith and Burel, Gregoire
(2023).
URL: https://ceur-ws.org/Vol-3370/paper6.pdf
Abstract
Psychological studies have demonstrated that much misinformation circulating on the Web tends to be more believable and memorable due to its ease of processing. The readability of a passage is a crucial factor in the ease of processing, as it indicates how easy or difficult it is to read and understand. According to some qualitative research, if online misinformation is easier to read, it becomes stickier and more memorable. In contrast, other studies showed that people are more likely to trust and believe misinformation when it appears to be more complex. As a result of such conflicting findings, it remains unclear how readability is associated with true or false content on the Web in general. This paper aims to gain a deeper understanding of readability through quantitative analysis by applying six readability formulas to four datasets containing both true and false content, as well as across multiple datasets. Our research shows that false claims are generally harder to read than true claims.