Data for "Warmth and Competence in Conversational AI: A Dual-Route Model of Purchase Intention"

This dataset contains experimental data from three studies examining how conversational AI agent characteristics influence consumer purchase intention in e-commerce. Study 1 (N=198) tested the competence-risk pathway using a skincare chatbot. Study 2 (N=200) tested the warmth-engagement pathway with both self-report and behavioral measures (interaction time, words exchanged). Study 3 (N=400) tested the complete dual-route model across two product types (skincare and laptops) using a 2×2×2 factorial design. Variables include perceived warmth, perceived competence, perceived risk, user engagement, purchase intention, and control variables (AI literacy, product knowledge). All experiments used functional LLM-powered chatbots (Gemini-2.5-flash) in simulated e-commerce environments. Participants were recruited from the United States via Prolific.