Subscribe Bookmark RSS Feed

RSM, ANOVA - effort estimate


Community Trekker


Apr 11, 2016

Hi, everybody. I am a new user of JMP. I am thinking of using a CCD for a 3 factor Response Surface Model.  My Question is now, how time and work intensive it is to analyze the results once I have generated them from the tests I want to perform. Does fitting the model equation to the data work automated with the model script?  I am also thinking of ANOVA and so on.  Is that Rocket Science or can I learn it by doing?

Thanks for helping me. I hope i posted this in the right position =)



Jun 5, 2014

Generally speaking using the Fit Model platform or the auto-generated Fit Model type scripts will take less than a handful of minutes wrt to getting what you in all likelihood need out of JMP. So the workflow is really pretty quick. This body of work can definitely be learned by 'doing'. I'd start with the JMP documentation, or view some of the DOE oriented JMP On Demand webinars to show you which buttons to push, potential drop down options, hot spots, using the Prediction Profiler, etc.

Now, if your question is more along the lines of "What does all this mean from a statistical and practical point of view?" Well that's a different question and one I'd be reluctant to recommend you 'learn by doing' unless you are under the tutelage of someone with a deeper broader skill set with respect to DOE, modeling. I define 'learn by doing' as just hitting whatever buttons strike your fancy, or blindly accepting JMP default values or menu selections. DOE and subsequent analysis are very powerful...but none of us were born with the skills to efficiently use them. They can only be learned through work on your part. Like riding a bike...none of us knew how at first...but we learned and never forgot!