Subscribe Bookmark RSS Feed

DOE question


Community Member


Mar 17, 2015

I am looking for advice in in setting up a DOE.  I have 4 design parameters (2 continuous and 2 categorical) that I would like to explore in relation to a measurable numeric response.  The response is non-targeted but rather the goal is exploratory to understand/predict relationships between the design parameters.  The next part is where I am struggling as I would like to replicate the above design over two environmental parameters and also over time.  Is there a recommended approach in jmp for what I am trying to accomplish?




Jun 23, 2011


Lots of discussion here but after getting more information it appears that a potential solution for you is to handle the two environmental parameters as "hard-to-change" factors in the custom design platform. This will allow you to handle the experiment in a way that is consistent with your constraints. The time as we discussed could be handled as a factor or as a response. Investigating it as a factor allows one to understand the interaction of time with your other factors. Handling it as a response may also be suitable as well depending on your domain knowledge. In the latter case you would analyze each time response independently.