Hi All
I have a quick question to any DoE Experts out there.
We have set up a half factorial design with 5 factors and 4 centre points.
Our 2 responses are highly correlated (r=0.85), does this mean anything?
Will this have any effect on the output/modelling process? Should we be
more cautious and try and reduce this correlation in future? Talking to
the people involved they say this is to be expected regarding the nature
of the experiments, however, I have not seen this before - although I'm
no expert!
The responses are both well modelled, and we are generally happy all round
with the output from the DoE. I'm just wondering if there's something in
that correlation that I should be picking up on.
Thanks for your time
Eloisa
|