Mapping “Super-Responders” to Algorithmic Comparison: An N-of-1 Micro-Randomized Trial in Adolescents

Build on Valkenburg et al.’s person-specific paradigm (2021) by running a micro-randomized trial that tweaks, at the session level, the feed’s “comparison density” (e.g., proportion of highly idealized posts, engagement metrics visibility, and influencer vs peer content). Adolescents report state self-esteem and social comparison moments (six times/day) while the system logs exposure characteristics. Existing studies largely correlate user-reported use with self-esteem; very few manipulate the algorithm in situ and almost none do it at person-specific granularity. This design moves from correlational to causal, and from average effects to mapping the tails—who benefits, who is harmed, and under what content regimes. It spotlights “super-responders”—adolescents whose self-esteem is unusually elastic to algorithmic changes—creating an empirical basis for targeted protections. It also reveals unexpected positives (e.g., some teens may gain self-worth when exposed to mastery- or community-oriented content). Person-specific response maps could inform adaptive safety by default: for highly sensitive teens, feeds can automatically lower comparison density or hide social metrics during vulnerable windows (e.g., late night).

References:

1. Social Media Use and Adolescents’ Self-Esteem: Heading for a Person-Specific Media Effects Paradigm. Patti M. Valkenburg, Ine Beyens, J. L. Pouwels, Irene I. van Driel, L. Keijsers (2021). Journal of Communications.