NATIONAL BUREAU OF ECONOMIC RESEARCH
NATIONAL BUREAU OF ECONOMIC RESEARCH
loading...

Kaspar Wüthrich

University of California at San Diego

E-Mail: EmailAddress: hidden: you can email any NBER-related person as first underscore last at nber dot org
Institutional Affiliation: University of California at San Diego

NBER Working Papers and Publications

December 2019Factorial Designs, Model Selection, and (Incorrect) Inference in Randomized Experiments
with Karthik Muralidharan, Mauricio Romero: w26562
Factorial designs are widely used for studying multiple treatments in one experiment. While “long” model t-tests provide valid inferences, t-tests using the “short” model (ignoring interactions) yield higher power if interactions are zero, but incorrect inferences otherwise. Of 27 factorial experiments published in top-5 journals (2007--2017), 19 use the short model. After including all interactions, over half their results lose significance. Modest local power improvements over the long model are possible, but with lower power for most values of the interaction. If interactions are not of interest, leaving the interaction cells empty yields valid inferences and global power improvements.
 
Publications
Activities
Meetings
NBER Videos
Themes
Data
People
About

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email: info@nber.org

Contact Us