I made a defintif screening design by JMP with 6 factors and 4 response, for each response i have some significant factors and non significant factors, i want finaly to optimise my fabrication with the significants factors on all the response together (Y1, Y2, Y3 and Y4). Which factors I will choose to optimize the 4 responses together?
I recommend that you fit each response separately and save each individual model. That way you can select the best combination of terms. You save the model by clicking the red triangle next to Response at the top and select Save Columns > Prediction Formula. When you have all four models saved, then select Graph > Profiler. Select the four columns with the prediction formula. Now you can optiomize them together, but each one has the best model.