Any specification test for an exposure mapping model with power against a larger exposure mapping model has worst-case Type I and Type II errors summing to one; informative tests require restricting the alternative class, as shown by a consistent test for no-interference versus network linear-in-ans
A General Design-Based Framework and Estimator for Randomized Experiments , year =
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Coupling designs extend stratified randomization to complex treatment spaces by matching units into homogeneous groups and dispersing treatments within groups to improve efficiency proportional to dispersion and match quality.
citing papers explorer
-
On the Impossibility of Specification Testing of Interference Models Based on Exposure Mappings
Any specification test for an exposure mapping model with power against a larger exposure mapping model has worst-case Type I and Type II errors summing to one; informative tests require restricting the alternative class, as shown by a consistent test for no-interference versus network linear-in-ans
-
Coupling Designs for Randomized Experiments with Complex Treatments
Coupling designs extend stratified randomization to complex treatment spaces by matching units into homogeneous groups and dispersing treatments within groups to improve efficiency proportional to dispersion and match quality.