Split Plot plot with Easy DOE

[#1]

Using Easy DOE I tried to create a RSM model split plot DOE with 4 easy to change continuous factors and on hard to change 3-level categorical effect cfr. screen shot here below

Why are in the whole plots the hard to change effect X5 levels still randomized cfr. screen shot below? Can easy DOE be used to create a split plot?

[#2]

Hi @frankderuyck,

I did the same steps as you and have the same design problem, with randomization in the whole plots for the "Hard to Change" factor.
However, with the Custom design, I'm able to have a design that do respect the restriction on randomization of hard-to-change factor X5 :

I would use the "Custom Design" platform for this kind of use case, as this seems more accurate and relevant (a split-plot design might not be an "easy" DoE ;) ).

All the best,

[#3]

Hi Victor, I know it works fine with Custom DOE.

As it doen not work It would be better to eleminate the hard to change factor option from the easy DOE platform, pity, I would like to use it as a standard in DOE training but unfortunately not possible. Is easy DOE OK for blocking? 

Regards, Frank

[#4]

My guess is Easy DOE is intended to encourage the use of experimental design.  I can see there might be some benefit getting novices to use experiments instead of OFAT's. While I understand this purpose, in reality it is an oxymoron.  If you are teaching DOE, you must certainly realize this.   It takes me about 20 days to teach how to plan, prepare, design and analyze experiments that are appropriate for the situation.  I start with fundamentals before using custom design as there are so many options available it can become overwhelming.

[#5]

Agree that split plot is more advanced (but still is part of a training) so better eleminate the hard to change factor option from easy DOE