Econometrics Seminar
Time: 14:00 pm – 15:30 pm
Date: 30th of September
Room : 1003

Guillaume POULIOT (University of Chicago) – “Placebo Tests Done Right”

Abstract :

“In order to assess the significance of a regressor, applied regression analysts sometimes rerun their regression replacing the regressor of interest with a randomly perturbed –often permuted– version which is expected to be insignificant, i.e., a placebo. While these procedures are often heuristic in their justification, p-values are often reported. We argue that the common interpretation of such placebo tests – for observational data– as Fisher test –designed for experimental data– is misleading, and that many such placebo tests and p-values are in fact invalid. In particular, we argue that almost all such tests for multivariate linear regression are invalid. A unified treatment of randomization inference and Fisher tests suggests more robust interpretations and designs of such placebo procedures, and allows to handle the multivariate linear regression case”

Organizers:

Xavier D’HAULTFOEUILLE (Pôle économie du CREST)

Sponsors:
CREST