The Effect of Types of Language Mistakes on the Persuasiveness of User-Generated Content on Facebook

Naama Meir, Nurit Tal-Or

Research output: Contribution to journalArticlepeer-review

Abstract

Language errors are prevalent on social media. We explored the effect of these errors on perceptions of the writer and the persuasiveness of the content that was posted. In an online experiment, participants (N = 325) were randomly assigned to read one of six identical texts designed as screenshots of Facebook posts that differed only in the types of mistakes they contained. The participants were then asked to report their attributions for the mistakes, perceptions of the writer, and attitudes related to the post. Language errors led to negative perceptions of the writer. In addition, these perceptions depended on the types of errors made and the reasons attributed to them. For example, typographical errors indirectly led to perceptions of the writer as rash, through attributing the errors to the writer’s hastiness. Spelling errors, on the other hand, indirectly led to perceptions of the writer as less intelligent, through attributing the errors to the writer’s inferior intellectual abilities. Moreover, language errors indirectly led to less acceptance of the writer’s claims. The findings are discussed in the context of attribution theory and the heuristic systematic model.

Original languageEnglish
Pages (from-to)224-232
Number of pages9
JournalPsychology of Popular Media
Volume14
Issue number2
DOIs
StatePublished - 15 Feb 2024

Bibliographical note

Publisher Copyright:
© 2024 American Psychological Association

Keywords

  • attribution
  • Facebook
  • language mistakes
  • persuasion
  • social media

ASJC Scopus subject areas

  • Cultural Studies
  • Communication
  • Applied Psychology
  • Psychology (miscellaneous)

Fingerprint

Dive into the research topics of 'The Effect of Types of Language Mistakes on the Persuasiveness of User-Generated Content on Facebook'. Together they form a unique fingerprint.

Cite this