We give a large-sample analysis of the minimal coverage probability of the usual confidence intervals for regression parameters when the underlying model is chosen by a "conservative" (or ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results
Feedback