I have a sample of $n=370$ $5$-dimensional observations which can be reasonably well modeled using a multivariate normal distribution. I want to estimate the mean using this sample, which will be part of a report, and I came across the Stein's paradox. Given that the James-Stein estimator is biased and the sample mean is unbiased, I am unsure on whether I should report the Sample Mean or the James-Stein estimator of the mean (or both?).

What is the practical recommendation in this case?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.