Confidence Interval for Population Mean (μ): Known vs Unknown σ

[onceinalifetim]

Hello

May I know what is the difference between

Confidence Interval for Population Mean (μ) when σ Known vs Confidence Interval for Population Mean (μ) when σ unknown..

any example to show me??

 

[jedishrfu]

This may help:

http://onlinestatbook.com/2/estimation/mean.html

 

[MarneMath]

What you'll typically see in a beginners course varies. If for some reason you form a CI with the s.d known you'll use a z-test. If it isn't known, you'll typically have to use a t-test. There are some caveats to this, and it really depends on how loose you want to be with it. An important thing to note is that if you are using SAS, it will assume that the s.d is unknown (because that's common) and give you values based on a t-test.

 

[phion]

<br /> \overline{X} \pm (z_{conf})(\sigma_\overline{x})<br />

Try using a sample standard deviation by using the bessel correction in the computation formula for s within \frac{\sigma}{\sqrt{n}}

 

[DrDu]

One important difference is that if the std is known, then the CI's will be all of the same width, while if sigma has to be estimated, the width will also vary between repetitions.