Extreme Z-values for use in normal distribution

[Klikoos]

Hi all,

I did some calculations using the normal distribution curve and Z tables. Normally these calculation use Z-values between -3 and 3.

Are these calculations still reliable/applicable beyond a 99.9% confidence interval?

Cheers,
Alex

 

[GregLocock]

Somewhat.

The problem you have is that you are calculating s from the sample you have. If you can assume a prioiri that your distribution is normal, then s defines the shape.

BUT... how do you know your distribution is normally distributed, especially at high Z values, unless you have the data to back it up?




Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?