Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore how capacity constraints in randomized controlled trials (RCTs) can fundamentally alter treatment effects and statistical power in this 35-minute research presentation. Discover the concept of "operational dosage" - the hidden variation in treatment intensity that occurs when participants share finite service capacity in interventions like teacher outreach, healthcare support, or social worker visits. Learn how queueing theory reveals that treatment effects become both capacity- and sample-size-dependent, with the counterintuitive finding that effects can actually decrease once sample size exceeds a critical threshold. Examine why statistical power in service intervention RCTs can peak at intermediate sample sizes, challenging conventional power calculations and showing how high-capacity/small-sample designs can achieve equivalent power to low-capacity/large-sample approaches. Understand the implications for experiment design through simulations calibrated to a tuberculosis intervention trial in Kenya, and gain insights into why replication failures and implementation challenges occur when scaling interventions.
