# Normality

Viewing 3 posts - 1 through 3 (of 3 total)
• Author
Posts
• #35824

Shashank
Member

I want to design my own ‘Normal Probability Plot Paper’. Can some one help me by explaining the logic of the horizontal (cumulative density function) scale.
It does not appear to be a log scale.
Regards
Shashank

0
#101569

Gabriel
Participant

You are right that it is not a log scale. It is a cummulative normal distribution scale and that’s why you get a normal probability plot paper instead of a log plot papaer.
The “procedure” to make the cummulative probability scale is as follows:
– Draw an equally spaced scale paralel to what will be the cummulative probablity scale.
– In the middel of that scale mark “0”, and then mark negative numbers down (or left) of zero and possitive numbers up (or right) of zero. That numbers will be the standard deviations from the average or , what’s the same, the Z value in the standardized normal distribution (average 0, standard deviation 1). How far to go with the possitive and negative numbers? That depends on how close from 0% and 100% you want to get. You know, at ±3 sigmas you have from 0.135% to 99.865%, at ±4.5 sigmas you have from 0.00034% to 99.99966%, and so on.
– Now you have two choices: Select the sigma value (Z) and find the corresponding cummulative probability (in Excel you can use =NORMSDIST(Z value) ), or select the cummulative probability and find the corresponding sigma value (in Excel you can use =NORMSINV(cummulative probability) )
– Match the cummulative probability values in the cummulative probability axes with the corresponding Z values in the Z values axis.
– You are done.
If you still cannot work it out, then post your e-mail and I will send you an example in excel. But try it yourself first, ok?

0
#101606

Shashank
Member

Dear Gabriel,
Thanks a lot. I was able to construct the scale for the “Normal Probability Plot” in MS-Excel . Thanks once again
Regards
Shashank

0
Viewing 3 posts - 1 through 3 (of 3 total)

The forum ‘General’ is closed to new topics and replies.