Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.

Measurement - Center Points not taken into account

#1
Hi everyone,

I am new to minitab, pleas forgive any obiously stupid Questions:
I have done a 2-factorial DOE with center Points with three factors. Looking at my effects diagram, it Looks like the Center point is not taken into account: there is just a straight line between the low and the high value.

Thanks a lot
Martin
DOE-A-Haupteffekte .JPG

DOE-A-Table.JPG
 

Miner

Forum Moderator
Staff member
Admin
#2
This is correct. Inclusion of center points allows you to test for the presence of curvature, but does not allow you to isolate that curvature to the specific factor(s) responsible for it. The next step is to modify this design in Minitab by adding axial points and additional center points (Stat > DOE > Modify Design). This will convert the design into a response surface model, which will allow you to isolate the factor(s) responsible as well as build a model.
 
#3
Thanks for your answer, Miner. I am confused, though. How can I see if I have a non linear effect? And if I add additional axial and Center Points, would I Need to rerun the whole design? Thanks, Martin
 

Miner

Forum Moderator
Staff member
Admin
#4
You cannot see the nonlinear effect. You only know that at least one exists.

If you add the axial points and additional center points. Minitab will add these to the existing design as another block. You only need to run the new points, but will need to analyzed the combined experiments as a response surface design. Once analyzed, you will be able to create countour plots and/or surface plots.
 


Top Bottom