Absence of Within-School Cross-Classroom Spillovers
RCTClinical Trial
Policy makers often rely on passive social diffusion to spread health knowledge across untreated groups within the same institution to lower implementation costs. However, assuming widespread cross-classroom sharing without empirical validation can lead to under-funded, partially deployed programs.
Picture this
Think of information diffusion like a Wi-Fi signal in a building. While the signal reaches everyone inside the same room easily, closed doors and wall boundaries between adjacent classrooms block the transmission, preventing knowledge from passively leaking across independent student groups.
What the evidence says
Spillover students exhibited non-significant knowledge improvements of 0.015 SD at 1 week and 0.011 SD at 6 months, and non-significant attitude changes of 0.026 SD at 1 week and 0.022 SD at 6 months compared to pure control schools.
- Who was studied
- N = 1,600 human ninth-grade students in 46 spillover classrooms located in treatment schools across 21 Colombian cities.
- How
- Two-stage randomized controlled trial assigning schools to treatment/control and classrooms within treatment schools to treated or untreated spillover status.
What to do
Enroll entire grade-level cohorts rather than a subset of classrooms when deploying school-based health curricula, as information does not passively diffuse across classroom boundaries.
From the source
"The results do not provide any indication of classroom-level spillovers within the same school, which suggests for policy purposes that programs should target entire cohorts, not a subset of classrooms with expectations of spillovers leading indirectly to treatment."
Do Information Technologies Improve Teenagers’ Sexual Education? Evidence from a Randomized Evaluation in Colombia