Michael Leung (University of Southern California)

"A Resampling Method for Dependence-Robust Inference."

Oct 30, 2017
from 03:40 PM to 05:00 PM

1131 Social Science and Humanities Gold room

This paper considers the problem of constructing hypothesis tests and confidence intervals that are robust to general forms of weak dependence. We provide a resampling procedure that is asymptotically valid under the weak requirement that the target parameter can be consistently estimated at the root-n rate by a mean-like estimator. Implementation does not depend on the particular correlation structure of the data and is computationally simple. We consider applications to constructing clustered standard errors when the level of clustering is unknown and the number of clusters is potentially small, inference on network data when the network is partially observed or the network-formation model is unknown, and testing for power laws with dependent data. We also propose tests of moment inequalities.



Seminar is open to the public, space is limited.
Filed under: