Normalizing Toxicity: The Role of Recommender Algorithms for Young People's Mental Health and Social Wellbeing
Published in Frontiers in Psychology in November 2025, this peer-reviewed paper by Kaitlyn Regehr and colleagues presents the academic, peer-reviewed analysis behind their 2024 'Safer Scrolling' report on TikTok's recommender system. Using the same mixed-methods design — archetype-account testing of over 1,000 videos plus interviews with young people and school leaders across England and Wales — the authors find that recommendation systems are associated with rapidly escalating exposure to gender-based-violence and misogynistic content framed as entertainment, and link this pattern to harms for young people's mental health and peer relationships. The paper argues that policy, education, and platform-design responses are all needed together. TechKids Foundation includes it as the formally peer-reviewed version of research earlier released as a policy report, adding academic-journal-level scrutiny to the same underlying dataset.
The peer-reviewed analysis reports the same core finding as the underlying 'Safer Scrolling' fieldwork: recommended misogynistic content on TikTok test accounts rose roughly fourfold within five days of use.
Regehr, Kaitlyn, Caitlin Shaughnessy, Minzhu Zhao, Idil Cambazoglu, Alfie Turner, and Nicola Shaughnessy. "Normalizing Toxicity: The Role of Recommender Algorithms for Young People's Mental Health and Social Wellbeing." Frontiers in Psychology 16 (November 6, 2025). https://doi.org/10.3389/fpsyg.2025.1523649