Operational Dosage: The Impact of Capacity Constraints on RCT Evaluation
Talkat Bridging Prediction and Intervention Problems in Social Systems (Simons Institute, 2026) · 2026-01-12..01-18·
summary
RCTs of human-delivered services show that finite staffing creates an operational dosage effect: as patient count rises per fixed number of callers, queueing delays (modeled with queueing theory and a continuous-time Markov chain approximation) lower individual exposure so average effects fall — a pilot with 100 showing 15% vs 10% at 1,000 exemplifies this — simulations calibrated to Kahila (Kenya TB) show power can peak at intermediate sample sizes and high-capacity/small-sample designs match low-capacity/large-sample outcomes.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:04RCT size and treatment effects
- 3:17Kahila tuberculosis intervention
- 5:56Capacity constraints and spillovers
- 7:08Operational dosage in services
- 9:25Dosage effects in healthcare
- 13:13How capacity constraints distort RCT estimates
- 15:35AI decision-making implications
- 20:27Stochastic process modeling
- 28:21Capacity constraints in ML interventions
- 30:37Key takeaways
- 31:01Q&A
related talks
29mLearning Outcomes in Queueing Systems
25mParadoxes in model training and evaluation under constraints
21m"Capacity Prediction" instead of "Capacity Planning":
27mA Practitioner’S Guide To Intent-To-Treat Effects From Multisite (Blocked)...
34mMaximizing marketplace experimentation: switchback design for small samples
25mOver- and Under-Estimation of Vaccine Effectiveness
01h 01mIntroduction to Experimental Design
05mDBTB INT Sanghamitra Deb r