Visitez notre page

 

 

 

 

 

 


Analytic inference with separately exchangeable data

Séminaire
Nom intervenant
Yannick Guyonvarch
Résumé
This paper studies analytic inference for an expectation under dissociated separate exchangeability. In such setups, the commonly used approach of Cameron, Gelbach and Miller (2011) has two drawbacks. First, the corresponding variance estimator is not necessarily positive even asymptotically. Second, inference is invalid in non-Gaussian regimes, namely when the sample mean is not asymptotically Gaussian. We consider a simple fix of the usual approach that addresses both issues. In Gaussian regimes, inference is asymptotically exact and equivalent to the usual one. Otherwise, inference is asymptotically conservative. Inference is also uniformly valid over a certain class of data generating processes. Finally, we highlight specific challenges that arise when dealing with multivariate means or nonlinear estimators.
Lieu
Amphi C2
Date du jour
Date de fin du Workshop