DOE prediction plot

[#1]

Hi, I developed a DOE model based on 3 factors to predict a response (fc). The model is robust and serves my purpose. However, I wanted to know how jmp calculates the confidence intervals (CI) in the 'Actual vs Predicted Plot' because the values I arrive at are very different from the ones provided by jmp (see attachment). When I save the CI values in columns, I get a different plot from the one presented by jmp. Are there any hidden scale factor for the IC values near the extremes? Can someone tell me how can I calculate the CI presented in the figure?

[#2 · Accepted solution]

Hi @aalberto : There is a recent discussion about this here.

https://community.jmp.com/t5/Discussions/Actual-by-predicted-plot-confidence-region/m-p/641519#M8386...

As outlined in the link above, don't use the act by pred plot to predict in your X1, X2, X3, ...Xn  space.

Actual by predicted plot details here. 

Solved: how to read actual by predicted plot - JMP User Community