# How to judge nonnormal distribution

Six Sigma – iSixSigma Forums Old Forums General How to judge nonnormal distribution

Viewing 8 posts - 1 through 8 (of 8 total)
• Author
Posts
• #37350

ROSS
Member

Hi:
We can use “Stat > Basic Statistics > Display Descriptive Statistics ” in Minitab to judge if data is normal distribution (p-value >0.1 is normal distribution) or not. But how can we judge what nonormal type it is, possion/Binomial…?

0
#109845

Patricio Torres
Participant

Tony:
If data is normal distribution p-value >0.05
If data is  nonormal can use Stat > Quality Tools >  Individual Distribution Identification……so run…..then see session list distribution and p-value. The better p-value is you distribution.
Regards
pato

0
#109846

mjones
Participant

Be careful here… It is true that p 0.05 does not ‘indicate’ data are normal — you cannot ‘prove’ that data are normal; only that they are not normal.
Otherwise, Pato’s suggestion to test for other distribution fits is right on. Excellent contribution!
Tony: I suggest you use Minitab “help.” It is helpful. Many MBB/instructors owe their success to Mtb “help” function.

0
#109847

Patricio Torres
Participant

MJones ,
I

0
#109848

Patricio Torres
Participant

MJones.
I agree. During my training as BB, with Six Sigma Consultans( Dr. Shree Nanguneri- Mike Carnell), they recommended this value p-value 0.05 as acceptably.

Pato

0
#109859

ROSS
Member

Hi all:  thanks for your all help firstly!
I know this function in Minitab-14, but can you explain the detail meaning for all those fourteen distribution, or where can I get the explaination(better is in  minitab help)?
Tks!

0
#109875

Robert Butler
Participant

I don’t know the Minitab program but I suspect their help section will provide some insight with respect to understanding the various distributions.  I say this because the packages I do use (JMP, Statistica, SAS) all have reasonable explanations of the various distributions. If you find Minitab does not give you enough detail I would recommend checking Statistical Models in Engineering – Hahn and Shapiro.  The title of the book should be – Everything you Wanted to Know About Distributions but Were Afraid to Ask. It is well written and can be read for general information as well as for detail.

0
#109943

mjones
Participant

Excellent advice. While I’m not personally familiar with that text, I absolutely agree with the point.
Minitab explains the concepts of what Mtb is doing to the data, how the analysis is performed, how to interpret results and even has some examples; and often gives equations for the analysis that is done. However, Mtb presumes you already know basic stat and the distributions you are interested in before you start using the tools.
Often Mtb gives a short overview of basic information and assumptions, but as Robert suggests, you really must look to other sources to fully appreciate the stat that Mtb is working on for you.
Please consider, while people on this Forum are willing to help with a specific question, when you ask for information on a wide number of distributions that is more than can be reasonably addressed.

0
Viewing 8 posts - 1 through 8 (of 8 total)

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