Publications
Publications
- 2024
- HBS Working Paper Series
Design of Panel Experiments with Spatial and Temporal Interference
By: Tu Ni, Iavor Bojinov and Jinglong Zhao
Abstract
One of the main practical challenges companies face when running experiments (or A/B tests) over a panel is interference, the setting where one experimental unit's treatment assignment at one time period impacts another's outcomes, possibly at the following time period. Existing literature has identified aggregating units into clusters as the gold standard to handle interference, yet the degree of aggregation remains an open question. In this work, we present a new randomized design of panel experiments and answer this question when all experimental units are modeled as vertices on a two-dimensional grid. Our proposed design has two features: the first feature is a notion of randomized spatial clustering that randomly partitions units into equal-size clusters; the second is a notion of balanced temporal randomization that extends the classical completely randomized designs to the temporal interference setting. We prove the theoretical performance of our design, develop its inferential techniques, and verify its superior performance by conducting an extensive simulation study.
Keywords
Citation
Ni, Tu, Iavor Bojinov, and Jinglong Zhao. "Design of Panel Experiments with Spatial and Temporal Interference." Harvard Business School Working Paper, No. 24-058, March 2024.