Comparison of Test Statistics via Expected Lengths of Associated Confidence Intervals
نویسندگان
چکیده
With reference to a large class of test statistics, higher order asymptotics on expected lengths of associated conndence intervals are investigated in a possibly non-iid setting. The connection with Bartlett adjustability is also indicated.
منابع مشابه
Outer and Inner Confidence Intervals Based on Extreme Order Statistics in a Proportional Hazard Model
Let Mi and Mi be the maximum and minimum of the ith sample from k independent sample with different sample sizes, respectively. Suppose that the survival distribution function of the ith sample is F ̄i = F ̄αi, where αi is known and positive constant. It is shown that how various exact non-parametric inferential proce- ′ dures can be developed on the basis of Mi’s and Mi ’s for distribution ...
متن کاملOn Confidence Intervals Associated with the Usual and Adjusted Likelihoods
In the presence of nuisance parameters, we derive an explicit higher order asymptotic formula to compare the expected lengths of conndence intervals given by likelihood ratio statistics arising from the usual proole likelihood and various adjustments thereof. Highest posterior density regions, with approximate frequentist validity, are also included in the study.
متن کاملMonte Carlo Comparison of Approximate Tolerance Intervals for the Poisson Distribution
The problem of finding tolerance intervals receives very much attention of researchers and are widely used in various statistical fields, including biometry, economics, reliability analysis and quality control. Tolerance interval is a random interval that covers a specified proportion of the population with a specified confidence level. In this paper, we compare approximate tolerance interva...
متن کاملOn Expected Lengths of Predictive Intervals
We consider the problem of comparing predictive intervals for a future observation via their expected lengths at a given con ̄dence level. A higher order asymptotic theory is developed. This yields an explicit formula for expected length comparison and associated admissibility results. Illustrative examples are given.
متن کاملModel-based likelihood ratio confidence intervals for survival functions
We introduce an adjusted likelihood ratio procedure for computing pointwise confidence intervals for survival functions from censored data. The test statistic, scaled by a ratio of two variance quantities, is shown to converge to a chi-squared distribution with one degree of freedom. The confidence intervals are seen to be a neighborhood of a semiparametric survival function estimator and are s...
متن کامل