Combine full factorial design with surface respone design

[#1]

Hi,

I'm working with clinical mass spectrometry and trying to learn JMP and DOE for optimization of the instruments and methods.

Currently I'm working on optimizing the ion source and its parameters, and want to do a response surface design. One of the factors I want to include is the MRM transition in the design (I have 4 candidates). But since the MS instrument is capable of measuring all 4 candidates in every run, I would like to include a full design for this 4 level categorical factor. Is there a way to set up this design in JMP? A surface response design for my continuous factors and full factorial for the 4 level categorical.

I have looked into making the design without the 4 level categorical, and then modifying the design table afterwards, but that messes with the script generation upon making the table.

Any suggestions?

[#2]

There are many options. How you set up your experiment will be effected by how the experimental runs are constrained. For example, do you want to (or need to) run all four categorical levels for each measurement? If so, can you vary levels of the other factors for each categorical level? Is the measurement process independent from the application of the factor levels? Could you measure four of the same MRM transition level at once? 

If there are no constrains, you can simply use Custom Design, input your continuous factors and a four level categorical factor, then select RSM. You'll get a warning about not having quadratic effects with categorical factors (a mathematical impossibility). Create all your experimental units, then measure them 4 at a time (grouping as needed).

[#3]

To echo @DonMcCormack 's counsel, the underlying software design principle behind JMP's Custom Design platform is in part for just this type of experimental conduct scenario...if the scenario doesn't fit the classic catalog of designs, let's give JMP users a better mouse trap.