Exercise 7.42 established that the optimal weights are q*i = (1/Ï2i)/(j 1/Ï2j). A result due to Tukey

Question:

Exercise 7.42 established that the optimal weights are q*i = (1/σ2i)/(ˆ‘j 1/σ2j). A result due to Tukey (see Bloch and Moses 1988) states that if W = ˆ‘i qiWi is an estimator based on another sets of weights qi ‰¥ 0, ˆ‘i gi = 1, then
Exercise 7.42 established that the optimal weights are q*i =

where λ satisfies (1 + λ)/(l - λ) = bmax/bmin, and bmax and bmin are the largest and smallest of bi = qi/q*i.
(a) Prove Tukey's inequality.
(b) Use the inequality to assess the performance of the usual mean ˆ‘i Wi/k as a function of σ2max / σ2min.

Fantastic news! We've Found the answer you've been seeking!

Step by Step Answer:

Related Book For  book-img-for-question

Statistical Inference

ISBN: 978-0534243128

2nd edition

Authors: George Casella, Roger L. Berger

Question Posted: