0
$\begingroup$

I'm reading a paper Mardia (1974) about multivariate normal tests.

There is a line that says "The statistic is asymptotically distributed as N(0,1)."

Now, I have calculated this value for one of my data sets, and it is -5.8, and for another set it is 0.87.

I am guessing it is something to do with Z scores?

For example, if this value is between +/- 2, then I have a 0.954 chance of it being normal (replace proper null hypothesis terminology here).

If i'm correct that would mean my "statistic" is just like a sample std deviation?

$\endgroup$
1
  • $\begingroup$ This is related to convergence in distribution. $\endgroup$ – StubbornAtom Jun 26 '18 at 5:03
0
$\begingroup$

Yes, the sample standard deviation is a statistic. If a statistic s(x) of some data x is asymptotically distributed as N(0,1), that generally just means that it is distributed as N(0,1) sample size(x) approaches infinity.

$\endgroup$
1
  • $\begingroup$ so what i say about it being between -2 and 2 is correct? Basically I have this statistic, I need to know if it then means my test for normal has passed $\endgroup$ – CptLightning Jun 26 '18 at 22:41

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.