Restrict combinations in DOE

[#1]

Hello all,

I'm sorry if the answer is completely obvious, I just started using JMP. I'm trying to design a DOE containing 1 response and 10 factors, let's call them A-J for now. The factors are either excluded from the run, or present at a certain concentration. In the eventual design certain combinations are restricted, namely A-B, C-D, E-F, G-H and I-J. I tried playing around with the factor constraint option, but I can't figure it out. Can anyone clarify what to do? Thanks in advance!

[#2]

Are you saying that either A or B is present at some concentration, but not simultaneously? Then you might have one two-level categorical factor to identify the ingredient (e.g., levels A and B) and another factor for the non-zero concentration range.

[#3]

Indeed, A and B can not be present simultaneous.

Ah yes that makes sense! But how does the factor for both ingredients being excluded look like? Or can I also make it 3-level categorical?

[#4]

I generally recommend against using 0 as a level for a continuous factor. Using 0 turns it into a categorical factor (absent, present). It is better to use two non-zero levels so that you can see the effect of changing the level.

Do you still need to cover the case where both A and B are 0 (absent)?

[#5]

I understand the case of making a categorical factor for A or B present, and why it is recommended. But I'm confused on what to do with the "A and B absent" factor.

[#6 · Accepted solution]

I am sorry but I do not understand your confusion. Please explain it so I can clarify my answer for you.

Here is a design based on my approach, so we have something tangible to discuss.

[#7]

Ah yes now I understand your approach!

Would it also work to make each ingredient 3-level categorical, including A, B or 0? Or do you foresee problems with such an approach?

[#8]

I think your scheme works if you have several pairs of ingredients and some formulations should not include either of the two options for a given pair.