JMP DOE Setting a range for my factors

[#1]

Hello, 

I am new to JMP and am trying to create a simple DOE where I am trying to maximize growth in a reactor by using different agitation rates, thus my agitation is my factor and my growth is my response. My issue is whenever I try to make a custom DOE I can not set a range for my factors, all there is are values and whenever I make my DOE it just uses the two values I entered. I want to set an upper and lower limit for my factor and not just use two values. I am setting my factor as continuous. Any help would be appreciated. 

[#2]

I must say I'm a bit confused by your query.  It sounds like you want to run a single factor experiment (aka OFAT).  When you add 1 factor to the customer design platform, it will create a design using coded levels (-1, 1).  You can use any actual values you want for these coded levels (use the upper and lower limits as the actual actual values).  

Perhaps you don't want to run an experiment, but want to collect variable data on the agitation rate (any actual values between the limits) and use regression to analyze the relationship between agitation rate and Growth?  In which case, collect the corresponding data in 2 columns in a JMP table, one for agitation rate and one for growth, then a simple fit Y by X platform should help to perform analysis.

[#3]

Another issue is that the design depends on the model you want to estimate. The default model will be an intercept and a single term for agitation rate. That is to say, JMP assumes that you expect a linear change in the growth when you change the agitation. The best design in that case is to observe only at the low and the high values. These observations provide the highest leverage in the regression analysis. For example, if you added a center point to this design, it would have zero leverage in estimating the slope.

[#4]

In Custom DOE In the factor definition you only define the range of experimentation. By adding polynomial terms in the model definition you can force extra points (for a quadratic you will get a mid point) and every order higher will give you one more point optimally spaced between your range limits.

[#5]

@pauldeen explained how to get more than two levels for a continuous factor. You pick the terms to represent the effects and JMP picks the levels. An alternate way is to add a Discrete Numeric factor. Now you pick the levels and JMP picks the terms.