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
notoriousapp
Level III

Can I have a 3-level categorical factor in a definitive screening design?

It seems that when using the JMP definitive screening platform in JMP 11 Pro I can only enter two levels for each categorical factor.  I need to include a 3-level categorical factor in my DOE.  Is this in any way possible?  If so, how?

1 ACCEPTED SOLUTION

Accepted Solutions
louv
Staff (Retired)

Re: Can I have a 3-level categorical factor in a definitive screening design?

Currently the Definitive Screening Designs accommodate only continuous factors and two-level categorical factors. Perhaps there is a slick way to have your categorical factor be treated as a continuous factor?

View solution in original post

4 REPLIES 4
louv
Staff (Retired)

Re: Can I have a 3-level categorical factor in a definitive screening design?

Currently the Definitive Screening Designs accommodate only continuous factors and two-level categorical factors. Perhaps there is a slick way to have your categorical factor be treated as a continuous factor?

notoriousapp
Level III

Re: Can I have a 3-level categorical factor in a definitive screening design?

Well that stinks.  Thanks anyways LouV.

louv
Staff (Retired)

Re: Can I have a 3-level categorical factor in a definitive screening design?

Not sure how many other factors that you have but one possibility is to run a DSD for each of the three level categorical factors. Alternatively you could use the custom design platform to easily handle the 3 level categorical factor and chose a RSM model for the experiment.

Re: Can I have a 3-level categorical factor in a definitive screening design?

You can create 3 sets of runs that are foldover pairs and treat the three groups as a categorical factor with 3 levels. The resulting design is orthogonal, however it may not be possible to have an equal number of runs in each group. Also, the factor you create in this way is not unaliased by two-factor interactions of the other factors.