cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Choose Language Hide Translation Bar
SimonFuchs
Level III

How to add a factor in an already planned DoE setup

Dear JMPxperts,

 

I set up a Definitive screening design with 10 factors and we found out that we have to change some factors and factor ranges of this design.

 

However we dont want to start from scratch again and generate a completely new randomized experimental plan, since we already invested some planning for the rest of the factors.

 

How can I change factors to new factors in a existing plan and how can I change factor ranges without srewing the whole plan? If I chnage thing in the DoE setup it doenst give me a new table and I have to start from beginning again.

 

Thanks for your help?

 

Best,
Simon

2 REPLIES 2
Phil_Kay
Staff

Re: How to add a factor in an already planned DoE setup

Hi Simon,

Classic statistician answer: it depends! There is not a general answer to this. It depends how much you need to change the design.

If it is just changing factor ranges, then it would be relatively simple. Changing the values in the design, assuming you stick to low, centre, and high values, will not affect the properties of the design. (Recode would help here)

If you need to add more factors, it will be more complicated. You will most likely need to take the table with the existing factors and use these as covariate factors (candidate runs) in Custom Design with your additional factors added. This will not result in a Definitive Screening Design.  

I hope this helps,

Phil

Re: How to add a factor in an already planned DoE setup

I had thoughts similar to @Phil_Kay. If you want the same design but only change the factor levels, then Cols > Recode is the way to go. Just be sure to edit the Coding column property for that factor, too, or the analysis will be incorrect.