cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Choose Language Hide Translation Bar
MartinY
Level III

Interpretation of JMP results

Hi Martin Yang here:

 

Recently I did a four ingredients Mixture design, I have some problem with the JMP results.

 

JMP provides normal plot which indicates which factor is significant. Below is the half normal plot generated by JMP.

 

As we can see, Alipre ( one type of cement), OPC ( ordinary portland cement), Alipre*OPC (interaction between the two cements), citric acid ( retarder) are significant factors.

 

 

half normal plot.PNG

 

However, if you read the Effects report listed below: Only Alipre cement and Alipre*OPC qualify as significant factors, since their P-Value is smaller than 0.05.

 

sorted parameter estimate.PNG

 

My questions is why there is discrapancy betwwen normal plot and Effects Report?  which one tells me which factor is significant?

 

Thanks!

 

Martin Y

1 REPLY 1

Re: Interpretation of JMP results

Hi @MartinY,

 

I discovered that this post did not have a reply.

 

For the Half Normal Plot, significance is determined by a cutoff of 0.1 (take a look at the Contrasts table above it). The t-ratios are also based on simulations and calculated a little differently. See the documentation here:

https://www.jmp.com/support/help/14/lenth-t-ratios.shtml

 

For Sorted Estimates, the cutoff is 0.05 as you have noted and the t-ratios have a slighty different calculation. See the documentation here:

https://www.jmp.com/support/help/14/sorted-estimates.shtml

 

By the way, for mixture experiments, it is not recommended to use the half normal plot found in the Fit Two Level Screening Platform. Instead it is best to use the Fit Model platform given the factors must sum to a value and are not orthoginal to each other. See this link:

https://www.jmp.com/support/help/14/fitting-mixture-designs.shtml#105936

 

Hope this helps,

 

Chris

Chris Kirchberg, M.S.2
Data Scientist, Life Sciences - Global Technical Enablement
JMP Statistical Discovery, LLC. - Denver, CO
Tel: +1-919-531-9927 ▪ Mobile: +1-303-378-7419 ▪ E-mail: chris.kirchberg@jmp.com
www.jmp.com