Randomization Inference for Peer Effects

Xinran Li, Peng Ding, Qian Lin +2

Many previous causal inference studies require no interference, that is, the potential outcomes of a unit do not depend on the treatments of other units. However, this no-interfere…