2
$\begingroup$

From Wikipedia

Given a statistic $T$ that is not sufficient, an ancillary complement is a statistic $U$ that is ancillary to $T$ and such that $(T, U)$ is sufficient. Intuitively, an ancillary complement "adds the missing information" (without duplicating any).

I understand that an ancillary statistic wrt a family of possible distributions of the sample, is defined as a statistic whose distribution doesn't depend on any sample distribution in the family of sample distributions.

I was wondering what "a statistic $U$ is ancillary to another statistic $T$" mean, in the definiiton of an ancillary complement?

Thanks and regards!

$\endgroup$

1 Answer 1

3
$\begingroup$

I'm fairly sure what they meant to say was that $U$ is ancillary for $\theta$, that $(T,U)$ is sufficient for $\theta$, & that $U$ is an ancillary complement to $T$ - where $\theta$ is a population parameter. Like you, I can't see what a statistic being ancillary to another statistic could possibly mean.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.