In-Lab Teen–Chatbot Conversations: Relational vs Transparent Styles and Social Stress

This in-lab study brings adolescents into conversation with a chatbot under controlled conditions to examine how conversational style affects their social and emotional experience.

Research question

When a chatbot adopts a relational style (warm, personal, human-like) versus a transparent style (openly signaling that it is an AI), how do teens’ stress responses, trust, and sense of connection differ — especially during socially stressful moments?

Approach

Adolescents complete a structured interaction with a chatbot while we collect self-report and physiological measures of stress and engagement. Comparing the two styles isolates the effect of relational framing on how teens cope with social stress.

Why it matters

Many AI companions are deliberately designed to feel like friends. This study helps clarify when that relational design supports teens — and when transparency about the AI’s nature may better protect their wellbeing.

← All research