0
$\begingroup$

For a data set that is Gaussian distributed, $\sigma$ defines the standard-deviation of the distribution. My question is, what is the correct indention or terminology for the following: $$\frac{\sigma}{\sqrt{2N-2}} \text{.}$$

My interpretation is that it is the "error" in the standard deviation, $\sigma$. Is this a correct way to view this?

$\endgroup$
1
  • 3
    $\begingroup$ To help you clarify your question, note that this is not quite a statistic because it's not a function of the data alone (sigma can't be computed from the data, only estimated). Second, where did this quantity arise? $\endgroup$ – RMurphy Jul 7 '19 at 17:49
2
$\begingroup$

It's important to distinguish between the true population value of a parameter of a statistical distribution and the estimate of that value obtained from a sample of size $N$. The symbol $\sigma$ is conventionally used to represent the population standard deviation, the square root of the population variance $\sigma^2$. A sample estimate of the variance is conventionally represented as $s^2$ and a sample estimate of the standard deviation as $s$.

This question shows the formula for the standard error of estimates of the variance, $s^2$, when sampling from a normal distribution. The answer to that question shows a way to convert such standard error estimates of $s^2$ to standard error estimates of $s$ that holds in the limit of large sample sizes for sampling from any of a large class of distributions. A bit of math shows that if you replace $\sigma$ in your formula with $s$ you have the corresponding formula for the standard error of the sample standard deviation in the case of a normal distribution.

$\endgroup$
1
  • $\begingroup$ Thanks @EdM, this was a great answer! $\endgroup$ – Q.P. Jul 7 '19 at 19:26

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.