Subscribe Bookmark RSS Feed

Intercept value differ between Design Expert and JMP

Hi,

I am evaluating three outcome variables from a 2^(7-3) fractional factorial experiment. All three variables seem to have the same significant parameters, three main effects.

I have run the dataset in both DesignExpert and JMP 8. I get essentially the same results from both in terms of parameter estimates, but not for the intercept.

Does anyone have experience from both software packages and any insight in if this is a mistake on my behalf, or built-in differences?
5 REPLIES
louv

Staff

Joined:

Jun 23, 2011

Without seeing the JMP file it is hard to say exactly. Did you set the experiment up in JMP initially or did you just read in the data? It may be a factor coding issue perhaps.
Do the sums of squares match?
I did not set it up in JMP, but I did make sure that the factors and levels matched.

I will have to check the sums of squares.
louv

Staff

Joined:

Jun 23, 2011

Double click on your Factor columns in your JMP table and using the column properties pull down, code your lower and upper levels appropriately.
Thank you!

I haven't tried it yet, but this is the obvious answer.

I very much appreciate it.

/Anders