Big thanks to all the young people who co-created these stories with us and to my brilliant co-authors: @orbenamy.bsky.social, @adigitaltanay.bsky.social @valerieyapzy.bsky.social, Ashish Mehta, Anne-Marie Burn, Petr Slovak

🔗 More on the method:

Micro-narratives = short, first-person stories co-created by participants + LLM-powered chatbots.

dl.acm.org/doi/10.1145/...

We also found differences in harm frequency, prevalence, and platform distributions:
📉 Behavioural harms (e.g., negative commenting) happened most often
📱 Instagram & TikTok stood out as key platforms of harm exposure
🌍 Harms were perceived as widespread amongst young people

𝐔𝐬𝐢𝐧𝐠 𝐦𝐢𝐱𝐞𝐝-𝐦𝐞𝐭𝐡𝐨𝐝𝐬 𝐚𝐧𝐚𝐥𝐲𝐬𝐢𝐬, 𝐰𝐞 𝐟𝐨𝐮𝐧𝐝 𝐭𝐡𝐚𝐭 𝐡𝐚𝐫𝐦𝐬 𝐟𝐞𝐥𝐥 𝐢𝐧𝐭𝐨 𝐟𝐨𝐮𝐫 𝐦𝐚𝐢𝐧 𝐜𝐚𝐭𝐞𝐠𝐨𝐫𝐢𝐞𝐬:
1. Other people’s behaviour (e.g. cyberbullying, scams)
2. Own behaviour (e.g. compulsive scrolling, comparison)
3. Content (e.g. graphic or idealised posts)
4. Platform features (e.g. poor moderation)

• Using 𝘀𝗲𝗹𝗳-𝗱𝗲𝘁𝗲𝗿𝗺𝗶𝗻𝗮𝘁𝗶𝗼𝗻 𝘁𝗵𝗲𝗼𝗿𝘆, we outline a behavioural change framework supporting 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝘆, 𝗰𝗼𝗺𝗽𝗲𝘁𝗲𝗻𝗰𝗲, and 𝗿𝗲𝗹𝗮𝘁𝗲𝗱𝗻𝗲𝘀𝘀—especially for adolescent users

We highlight the need for 𝘁𝗵𝗲𝗼𝗿𝗲𝘁𝗶𝗰𝗮𝗹𝗹𝘆 𝗶𝗻𝗳𝗼𝗿𝗺𝗲𝗱 𝗮𝗽𝗽𝗿𝗼𝗮𝗰𝗵𝗲𝘀 that centre users’ 𝗽𝘀𝘆𝗰𝗵𝗼𝗹𝗼𝗴𝗶𝗰𝗮𝗹 𝗻𝗲𝗲𝗱𝘀.

• We distinguish between 𝘂𝘀𝗲-𝗿𝗲𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗶𝗻𝘁𝗲𝗿𝘃𝗲𝗻𝘁𝗶𝗼𝗻𝘀 (e.g., limiting time) and 𝘂𝘀𝗲-𝗶𝗺𝗽𝗿𝗼𝘃𝗲𝗺𝗲𝗻𝘁 𝗶𝗻𝘁𝗲𝗿𝘃𝗲𝗻𝘁𝗶𝗼𝗻𝘀 (e.g., enhancing content and interactions)
• We introduce a 𝗰𝗼𝗻𝘁𝗲𝘅𝘁𝘂𝗮𝗹 𝗺𝗼𝗱𝗲𝗹 mapping interventions across five levels: 𝗽𝗹𝗮𝘁𝗳𝗼𝗿𝗺, 𝗱𝗲𝘃𝗶𝗰𝗲, 𝘂𝘀𝗲𝗿, 𝗳𝗮𝗺𝗶𝗹𝘆, and 𝘀𝗼𝗰𝗶𝗲𝘁𝘆