how to add confidence intervals in a probability plot?

Dear guys,

I want to know how to add confidence intervals in a probability plot?

Please have a look at this code:

x=[-1.41207247401840 -1.06884479800464 -0.805056161180770 -0.695214836608495 -0.537941409385110 -0.340400323764181 -0.209393059855624 -0.179107587132278 -0.165916099145327 -0.116016693050744 0.292200480227937 0.590116759815710 1.08666272111204 1.42323816718219 2.13774531380767];

normplot(x);

Could someone tell me how to add confidence intervals or prediction intervals in this normal probability plot?

Thank you.

David

Answer 1

[the cyclist]

Confidence interval for what quantity?

Usually, this type of plot is used purely for a visual idea of whether normality is a good assumption or not. If you are looking to do something more quantitative, you might want to use JBTEST or LILLIETEST to test for normality.

[David]

95% two sided confidence intervals

Yes, you are certainly right. What I want to do is to give a visual indication of the data distribution.

Could you tell me how to do it?

Thank you and look forward to your reply.

David

[the cyclist]

I guess my question was not clear enough. There are many quantities that you could calculate the confidence interval for. For example, it could be the mean of x, or the median of x. Or do you just want the middle 95% of x itself?

Also, are you asking how to calculate something, or just how to display something? If the latter, you can just use the "line" command to draw lines on the plot you have.

[David]

Well, 95% conficence intervals for the probability.

When you plot Probability Vs X, using normplot(x), I think it is also possible to plot the confidence intervals for the probability (0 to 1).

Thank you.

David