Non-face emojis and credibility in user-generated content: the roles of review valence and consumer involvement
研究基于感觉即信息理论，通过两个实验发现非面部表情符号通过积极情绪和加工流畅性影响评论可信度，且效果取决于评论效价和消费者涉入度。
Purpose This study applies the Feelings-as-Information Theory (FIT) to investigate how non-face emojis in user-generated content (UGC) influence perceived review credibility through positive affect and processing fluency, and how these effects vary by review valence and involvement. Design/methodology/approach Two experiments were conducted: a 3 × 2 between-subjects design with review valence as a contextual factor, and a 3 × 2 × 2 between-subjects design with involvement acting as a boundary condition for the early-stage impact of emojis on perceived credibility via positive affect and processing fluency. Findings Study 1 shows that in positive reviews, low-relevance (LR) and high-relevance (HR) non-face emojis enhance perceived credibility by boosting positive affect and processing fluency, respectively. In negative reviews, LR emojis reduce perceived credibility by impairing processing fluency, while HR emojis enhance it by supporting processing fluency. Study 2 reveals that these effects are attenuated for highly involved consumers. Research limitations/implications The findings enrich research on emojis, image–text relevance, multimodal communication and FIT. Limitations include the experimental design and focus on positive affect and processing fluency, suggesting opportunities for future research using broader methods and additional mechanisms. Practical implications Marketing managers, UGC platforms and content creators can strategically use non-face emojis to optimise emoji marketing, resulting in more valuable evaluations. Originality/value This research introduces a novel categorisation of non-face emojis into low- and high-relevance types based on emoji-text alignment. Furthermore, it demonstrates that their credibility effects depend on both message-level (review valence) and consumer-level (involvement) contingencies.