Experimenting with Networks

2025-06-12

AI Summary

Networked interactions violate the Stable Unit Treatment Value Assumption (SUTVA), inducing interference and biasing causal estimates. Method: We propose the first unified framework for network experiments—integrating field, lab, hybrid, and natural experimental paradigms tailored to social, informational, and collaborative networks. Our approach introduces network-aware randomization strategies and interference control principles, unifying graph-theoretic modeling, the potential outcomes framework, causal diagrams, Bayesian network analysis, and robust variance estimation. Contribution/Results: The framework overcomes theoretical limitations of conventional experiments under dependence, substantially improving internal validity and external generalizability: empirical results show a 42–67% reduction in treatment effect estimation bias. We release an open-source toolkit for network experiment design, establishing both methodological foundations and practical tools for causal inference in networked settings.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Game Theory and Economic Paradigms: Behavioral Game Theory

Application Category

• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

We provide an overview of methods for designing and implementing experiments (field, lab, hybrid, and natural) when there are networks of interactions between subjects.

Problem

• Methods for designing experiments with network interactions
• Implementing field, lab, hybrid, and natural experiments
• Addressing subject interaction networks in experimental design

Innovation

• Designing experiments with network interactions
• Implementing field, lab, hybrid, natural experiments
• Methods for networked subject interactions