Panel Data and Experimental Design EconomicsSSMART

Fiona Burlig Matt Woerman Louis Preonas

Burlig, Preonas, and Woerman develop a power calculation technique and accompanying software package for experiments that use panel data. They generalize Frison and Pocock (1992) to fully arbitrary error structures, thereby extending McKenzie (2012) to allow for non-constant serial correlation. Using Monte Carlo simulations and real-world panel data, they demonstrate that failing to account for arbitrary serial correlation ex-ante yields experiments that are incorrectly powered under proper inference. By contrast, our “serial-correlation-robust” power calculations achieve correctly powered experiments in both simulated and real data.

Publications associated with this project: