|
Activity Number:
|
616
- Multidisciplinary Advances in Computing
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 1, 2019 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Computing
|
|
Abstract #306772
|
Presentation
|
|
Title:
|
Nearly Best Wald Confidence Intervals
|
|
Author(s):
|
George Terrell*
|
|
Companies:
|
VA Poly. Inst. & State Univ.
|
|
Keywords:
|
confidence interval;
inverse Gaussian;
asymptotics;
Wald interval
|
|
Abstract:
|
The best confidence interval of size alpha for an unknown parameter consists of those values that exceed a certain likelihood. This can be expressed in closed form in only a few families. A useful nonsymmetric one is the inverse Gaussian with mean parameter. We generalize the familiar asymptotic Wald approximate confidence interval to families in the normal domain of attraction but arbitrary skewness by matching derivatives of the likelihood to those of the inverse Gaussian, to achieve a higher-order nearly best interval.
|
Authors who are presenting talks have a * after their name.
