The effect of avatars on perceived credibility of comments posted to online news stories

Abstract

An experiment tested two two-part hypotheses predicting the effect of specific avatar features -- avatar humanness and eye contact -- on perceived credibility of related comments about online news stories. Participants viewed a series of news stories and related comments, and responded to questions regarding the likelihood of their recommending the comments; perceived credibility of the comments; and perceived social presence of the comments. Repeated-measures analysis of variance (ANOVA) was used to test the effects of the manipulations. A significant interaction was identified between avatar humanness and eye contact, such that comments paired with indirect-eye contact human avatars were rated as more credible than comments paired with direct-eye contact human avatars. Hypothesis 1a predicted that comments paired with human-like avatars would be recommended more often than comments paired with animal-like avatars. Hypothesis 1b predicted that comments paired with human-like avatars would be rated as more credible than comments paired with animal-like avatars. Neither hypothesis was supported. Hypothesis 2a predicted that comments paired with direct-eye contact avatars would be recommended more often than comments paired with indirect-eye contact avatars. Hypothesis 2b predicted that comments paired with direct-eye contact avatars would also be rated as more credible than comments paired with indirect-eye contact avatars. Neither hypothesis was supported. The avatar manipulations had no effect on perceived comment social presence.