Graph Agnostic Randomized Experimental Design

Randomized experiments are widely used to estimate causal effects of proposed "treatments" in domains spanning across physical sciences, social sciences, medicine, and technology industries. However, classical approaches to experimental design rely on critical independence assumptions that are violated when the outcome of an individual a may be affected by the treatment of another individual b, referred to as network interference. Under such network interference, naively using popular estimators and randomized experimental designs can result in significant bias and loss of efficiency. We consider a heterogeneous linear outcomes model that can capture network interference that arises from spillover, peer effects, and contagion. Under this model, we characterize the limitations and possibilities for estimating the total treatment effect, average direct treatment effect, and average interference effect. Given access to average historical baseline measurements prior to the experiment, we propose simple estimators and randomized designs that output unbiased estimates with low variance for these three estimands. Furthermore, our solution and statistical guarantees do not require knowledge of the underlying network structure, and thus can be used for scenarios where the network is unknown and complex. We believe our results are poised to impact current randomized experimentation strategies due to its ease of interpretation and implementation, alongside its provable statistical guarantees under heterogeneous network effects.