4
$\begingroup$

Among other applications, a gamma distribution answers the question: "If the average time (or other quantity) between events is β, what is the probability that x time will elapse before α events occur?"

The exponential distribution is a special case of the gamma distribution where α = 1. The question is easily rephrased as: "If the average time before an event occurs is β, what is the probability that x time will elapse before the event occurs?"

My textbook defines the chi-squared distribution as a special case of the gamma distribution where α = v/2 and β = 2 where v is a positive integer, but the significance of these numbers is not clear to me. Can the question in the first paragraph be rephrased to fit the chi-squared distribution in the same way I rephrased it to fit the exponential distribution?

$\endgroup$
1
  • $\begingroup$ $\nu$ is the average of the chi-square rv in this case as well $\endgroup$
    – Taylor
    Jul 24 '16 at 20:52
2
$\begingroup$

The $\chi^2_n$ distribution can be seen as a $\Gamma(\frac{\nu}{2},2)$, which is called the scale parametrisation of the $\Gamma$-distribution. There is, however, also a rank interpretation. In that case the $\chi^2_n$ distribution can be seen as a $\Gamma(\frac{\nu}{2},\frac{1}{2})$, which I think lends itself more easily to a rephrasing of your question. Your question would then become:

Given $\nu$ independent processes, each with succes rate $\frac{1}{2}$, how long would it take before exactly half of these processes ($\frac{\nu}{2}$) yield a succes?

$\endgroup$

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.