49898_en-US

liliconcom

what does it mean? the result from custom design

 The custom design involves 2 continuous factors and 1 category factory(,saying, Solvent A,B,C).

 when the data is from a custom design, first I apply stepwise to identify what factors or interaction will be significant, then fit a model, the JMP pump up an output as below:

then I was confused what the mean of Solvent{A&B-C}, solvent{A-B}?

From the result, in order to get the optomal result, I can set x1=80, x2=60, but what's the solvent?

 

 

Re: what does it mean? the result from custom designThanks for your help!Re: what does it mean? the result from custom designHi,

Stepwise has to recode categorical factors with more than 2 levels in this way in order to test significance of different levels. It does make the model difficult to interpret.

You have a couple of options. 1) in the Stepwise platform choose "whole effects" in "Rules" - it will only add or remove the whole effect of Solvent and you will not get the recoded effects in Fit Model. 2) When you make the model from Stepwise and you are in the Fit Model dialog you can remove the recoded effects and replace with the original Solvent effect.