UC Santa Cruz

Generating perceptions of others is usually effortless, but determining how others perceive us is more demanding. People often underestimate how much others like them, a cognitive bias known as the liking gap. We examined how the communicative setting affected the liking gap in conversation. We conducted four experiments where participants interacted via different forms of computer-mediated communication. In Experiment 1, participants worked on tangram tasks and communicated first via text, audio, or video-chat, then switched to video-chat. After this, they completed a memory test of the tangrams seen. In Experiment 2, participants engaged in conversations about favorite movies and TV shows and interacted via text, audio, or video-chat, then switched to video-chat. After this, they recalled topics they discussed. In Experiment 3, participants again engaged in conversations about favorite movies and TV shows and interacted via a single modality of text, audio, or video-chat. In Experiment 4, participants engaged in conversations about favorite movies and TV shows, but interacted via mixed modalities where one participant interacted via text, audio, or video-chat, and the other interacted in a different modality of text, audio, or video-chat. Across all four experiments, the liking gap persisted: People thought they were liked less than they actually were. Results also showed a cross modality benefit for recall: switching from text-chat to video-chat (Experiments 1 and 2) and audio-chat to video-chat (Experiment 2) resulted in better memory recall of tangram shapes and conversational topics. Experiments 3 and 4 also investigated moderating effects of experience, preference, personality traits, and social desirability characteristics on the liking gap. Experience and preference had mixed moderating effects, while only the personality traits of Openness to Experience and Agreeableness had moderating effects. Social desirability characteristics were not found to moderate the effect of the liking gap.