Panel Data and Experimental Design

39 Pages Posted: 9 Sep 2019 Last revised: 23 Apr 2021

See all articles by Fiona Burlig

Fiona Burlig

University of Chicago

Louis Preonas

University of California, Berkeley - Department of Agricultural & Resource Economics

Matt Woerman

University of Massachusetts Amherst

Multiple version iconThere are 3 versions of this paper

Date Written: September 2019

Abstract

How should researchers design panel data experiments? We analytically derive the variance of panel estimators, informing power calculations in panel data settings. We 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, we 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. We discuss the implications of these results, and introduce a new software package to facilitate proper power calculations in practice.

Suggested Citation

Burlig, Fiona and Preonas, Louis and Woerman, Matt, Panel Data and Experimental Design (September 2019). NBER Working Paper No. w26250, Available at SSRN: https://ssrn.com/abstract=3450276

Fiona Burlig (Contact Author)

University of Chicago ( email )

5757 S. University Ave
Chicago, IL 60637
United States

Louis Preonas

University of California, Berkeley - Department of Agricultural & Resource Economics ( email )

Berkeley, CA 94720
United States

Matt Woerman

University of Massachusetts Amherst ( email )

Amherst, MA
United States

HOME PAGE: http://https://sites.google.com/site/mattwoerman/

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
7
Abstract Views
263
PlumX Metrics