The JMP Design Of Experiments Difference
- Create full custom designs. Describe your process variables and constraints, and then the custom designer tailors a design to fit. The Custom Designer gives you the most flexibility. You can have many different factor roles in the same experiment. Factors can take on continuous, categorical, blocking, mixture, and covariate roles. Enter one or more responses, enter factors, enter factor constraints, choose a model, modify sample size alternatives, choose the run order, and add center points and replicates.
- Use screening designs. JMP supports fractional factorial, Plackett-Burman, and mixed-level designs. Screening experiments separate the factors that have a significant influence on the response from the rest of the factors. Often screening designs are a prelude to further experiments.
- Use response surface designs, including central composite and Box-Behnken. Response surface designs were invented to find the optimal response within the specified ranges of the factors. The designs are capable of fitting a second order prediction equation for the response. The quadratic terms in these equations model the curvature in the true response function. If a maximum or minimum exists inside the factor region, RSM can find it.
- Create full factorial designs. Automatically generate all possible combinations of a set of factors. This is the most conservative design approach, but it is also the most costly in experimental resources.
- Employ the Taguchi method. Create orthogonal arrays that are two-level, three-level, and mixed-level fractional factorial designs. The goal of the Taguchi Method is to find control factor settings that generate acceptable responses despite natural environmental and process variability.
- Determine the best model for your data. Fit one or more y variables to a model of x variables. Select the kind of model appropriate to your data to construct linear models that have the complex effects needed for designed experiments.
- Perform complex what-if scenarios. Visualize your model, optimize responses and use the JMP Prediction Profiler to get a closer look at the response surface to find the best settings that produce the response target. Graphically change one variable at a time and look at the effects on the predicted response.
More on DOE
JMP® Design of Experiments (DOE) Points of Interest
(PDF, 970K)
White Paper: Interactive Data Mining and DOE: the JMP Partition and Custom Design Platforms
(PDF, 600K)
Related Data (Zip File, 21K)
"You have to know that you're the fastest guy out there, that you have the best team and that you have the best equipment. We do. And we have JMP software."
Daniel Herrington
Race Car Driver, AIM Autosport and DMH Racing
Read the AIM Autosport Customer Story
