1
$\begingroup$

Assume $X_i$ are generated with the following distribution:

$$ f(x; \theta, c) = \theta^{-c}cx^{c-1}e^{-(x/\theta)^c}$$ $\theta>0$ and $c>0$ is known.

Further, assume $T(X)=\sum^{n}_{i=1} X_i^c$. How can we derive the distribution of $T(X)$ such, that it would get the form of something along the lines of $\theta^c\cdot \chi^2(2n)$ distribution?

Also, I have a guess that any distribution, that can be expressed as an exponential family, their statistic $T(x)$ can be also expressed as a $\chi^2$ with some scale/mean transformation. Is this true, and maybe there is some literature describing exactly this?

$\endgroup$
1
$\begingroup$

Assuming $X_1,X_2,\ldots,X_n$ are i.i.d with pdf

$$f(x;\theta,c)=\frac{cx^{c-1}e^{-(x/\theta)^c}}{\theta^c}\mathbf1_{x>0}\quad,\,\theta,c>0$$

It is straightforward to show that $$\left(\frac{X_i}{\theta}\right)^c\stackrel{\text{ i.i.d }}\sim \mathsf{Exp}(1)$$

Equivalently, $$2\left(\frac{X_i}{\theta}\right)^c\stackrel{\text{ i.i.d }}\sim \chi^2_2$$

Hence, $$2\sum_{i=1}^n\left(\frac{X_i}{\theta}\right)^c=\frac{2}{\theta^c}T\sim \chi^2_{2n}$$

$\endgroup$

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.