design-based inference under interference

Under network interference, causal effect estimation suffers from high variance, while simultaneously minimizing cut edges within clusters and achieving covariate balance remains challenging. Method: We propose a two-stage rollout experimental design: (1) graph-based clustering to identify highly homogeneous subpopulations, followed by (2) intervention deployment exclusively within those subpopulations. Contribution/Results: We formally link clustering objectives—cut-edge minimization versus covariate balance—to the bias–variance trade-off in causal estimation, theoretically characterizing how cluster structure affects bias (governed by cut edges) and variance (driven by homogeneity and covariate balance). Using a polynomial interpolation estimator and Monte Carlo simulations, we empirically identify optimal trade-offs across diverse clustering strategies. Our approach significantly reduces estimation variance while preserving causal identification validity under interference.

This paper addresses causal inference under violations of the Stable Unit Treatment Value Assumption (SUTVA) and interference among units. Method: We propose the Homogeneous-Intervention Average Treatment Effect (HAATE) as a new target estimand for the Global Average Treatment Effect (GATE); formally define HAATE; prove theoretically that the difference-in-means estimator dominates a correctly specified regression model under interference; and design a two-stage cluster-randomized experiment that leverages intra-cluster treatment correlation to model cluster-level error, thereby substantially reducing root mean squared error (RMSE). Contribution/Results: Monte Carlo simulations and a large-scale online A/B test on Facebook demonstrate that, compared to conventional designs, our approach significantly improves estimation accuracy in finite samples—enhancing the reliability of policy-level causal inference under interference.

In cluster-randomized trials, cross-cluster interference is often ignored—especially when units exhibit irregular spatial distributions and cluster boundaries are ill-defined—leading to systematic bias in causal effect estimation. Method: We formally characterize the bias–variance trade-off under cross-cluster interference and propose a “boundary-unit exclusion” truncation estimator that mitigates interference by leveraging neighborhood homogeneity. We further establish interference-robust theoretical guarantees for k-medoids clustering. Contribution/Results: Our approach significantly reduces asymptotic bias while keeping variance inflation bounded. We derive an optimal cluster-number selection criterion grounded in this bias–variance analysis. Crucially, the method requires no prior knowledge of community structure, thereby enhancing both estimation consistency and practical applicability in real-world settings with ambiguous clustering.

Traditional randomized controlled trials (RCTs) suffer from cross-group spillovers and interference in multi-population interaction settings—e.g., buyer-seller or creator-subscriber systems—leading to biased estimation of the average treatment effect (ATE). This paper introduces the first systematic multidimensional randomization design framework, relaxing the conventional single-layer randomization assumption. By integrating hierarchical randomization, cross-group assignment, and potential outcomes modeling, it jointly identifies both the ATE and cross-group interference effects. We establish theoretical guarantees: the proposed unbiased estimator is consistent and asymptotically normal. Simulation results demonstrate substantially higher statistical power compared to standard RCTs. This work extends the scope of causal questions addressable through experimental design and provides a rigorous foundation for causal inference in complex intervention environments—particularly platform economies—where interdependent user behaviors induce non-negligible interference.

In causal inference, the no-interference assumption is frequently violated, complicating the separation of direct and spillover effects—particularly when interference structures are unknown and arbitrarily complex, rendering existing methods incapable of identification or valid inference. To address this, we introduce the *Degree of Interference* (DoI), a unit-level latent variable formalized for the first time as a tractable, estimable latent factor—enabling a general causal framework that imposes no prior assumptions on network topology or adjacency structure. Leveraging Bayesian nonparametrics, we develop a blocked Gibbs sampling data-augmentation algorithm that jointly infers DoI and treatment effects. Simulations and empirical analysis of a cash-transfer program demonstrate that our approach substantially improves spillover effect identification accuracy and delivers robust causal estimates even under completely unknown interference patterns.

This study addresses limitations of conventional causal inference methods in networked group experiments under interference, which often rely on exposure mapping assumptions, no-interference conditions, or Bernoulli assignment mechanisms. The authors propose a general framework that dispenses with exposure mappings and introduces a novel class of linear weighted estimators tailored to two-stage randomized designs. Certain estimators within this class achieve the optimal root-N convergence rate independent of the number of groups. The work further establishes, for the first time, asymptotic theory and a bias-corrected variance estimation method for dependent statistics under complete randomization. Theoretical analysis confirms the consistency and asymptotic normality of the proposed estimators, while simulations demonstrate their superior finite-sample performance over existing approaches and provide practical guidelines for experimental design and weight selection.

This study addresses the failure of conventional causal inference methods in group interaction experiments, where within-group interactions and interference effects violate standard assumptions. The authors develop a design-based causal inference framework that systematically characterizes identifiability under various scenarios—such as fixed or random group assignment and presence or absence of interference—and proposes corresponding inference strategies. Innovatively, they introduce a coupling strategy to handle complex dependence structures, integrating sparse-sampling asymptotics, cluster-robust inference, and the potential outcomes framework. They demonstrate that, even under interference, cluster-robust methods consistently estimate marginalized exposure effects. Moreover, when interference is absent and assignment is randomized, the framework naturally reduces to the standard individual-level randomized experiment, thereby preserving compatibility with classical individual-level inference.

研究通过因子回归、双向聚类和协变量调整方法，解决双侧随机设计中的因果推断问题，适用于市场买卖双方等互动群体。

This study addresses the limitations of conventional approaches to estimating network spillover effects, which typically rely on a first-order neighborhood interference assumption and fail to capture higher-order or broader interference patterns. To overcome this, the authors propose a novel causal inference framework grounded in a generalized interference assumption, extending the interference set to nodes within the same community or reachable via paths of limited length, thereby enabling the definition and identification of spillover effects at specific network distances. They develop Horvitz–Thompson- and Hájek-type estimators—along with their weighted regression counterparts—tailored to complex interference structures, leveraging community detection and path analysis to delineate interference sets. Theoretical analysis and simulations demonstrate the robustness of these estimators across diverse network topologies and interference mechanisms. Empirical application to a two-stage randomized trial on maternal and child health interventions in Honduras successfully quantifies the bias arising from misspecified interference sets.

This study addresses the challenge posed by network interference, which complicates the accurate estimation of both direct treatment effects and spillover effects in conventional experimental designs. To overcome this limitation, the authors propose a novel randomization scheme based on ego-clusters—clusters formed by each focal individual and their immediate neighbors—and develop a tailored clustering algorithm aimed at minimizing the asymptotic variance of the resulting estimators. Within a formal model framework, the approach integrates model-based estimation with asymptotic normality theory to simultaneously identify global average treatment effects and spillover effects. Simulation studies and empirical analyses demonstrate that the proposed method substantially improves estimation precision and inferential efficiency compared to existing network experiment designs.