Abstract
ABSTRACT:
With reference to Stein's two‐stage sampling, it is shown that with a suitably chosen random confidence coefficient, depending on second stage variation, one can be closer in the mean square sense to the indicator specifying whether a normal mean lies in a bounded length confidence interval, than the constant confidence coefficient 1‐α. The efficiency of the constant confidence coefficient 1‐α. as a predictor of this indicator, has also been investigated.
Keywords
Get full access to this article
View all access options for this article.