2
$\begingroup$

Let $g:\mathbb{R}^2\rightarrow \mathbb{R}$ be a smooth, but not necessarily bounded function and $X$ and $Y$ two random variables that are not independent. (assuming they yield sufficiently many moments for the following question)

I'm wanting to approximate the mean $\mathbb{E}[g(X,Y)]$ with a multivariate Taylor expansion around the mean values of $X$ and $Y$ (delta method) and would like to have an estimate for the order of the error in terms of the sample size $n$ of an i.i.d. sample of $X,Y$.

Oehlert [1] has given a result for a one-variable function and I am looking for a similar result for my two-variable function $g$.

Does anyone know whether it is possible to establish a result of this type for my problem? If so, under which reference can it be found?

[1] Oehlert (1992), A Note on the Delta Method, The American Statistician, Vol. 46, No. 1, p. 27-29.

$\endgroup$
2
  • $\begingroup$ I am not sure to understand the question but concerning the multivariate Delta method see the book "Testing statistical hypotheses" by Lehmann & Romano $\endgroup$ Feb 21, 2012 at 19:14
  • $\begingroup$ Serfling's book, "Approximation Theorems of Mathematical Statistics", contains a number of results concerning the second-order multivariate delta method, but it is rather a slog. Start around chapter 6. $\endgroup$
    – Steven Pav
    Apr 21, 2014 at 19:14

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.