topic alfa value for rotatability and orthogonal blocking in response surface designs in Discussions

alfa value for rotatability and orthogonal blocking in response surface designs

MFVIT — Thu, 08 Jun 2023 21:05:13 GMT

I noted that when I use the response surface platform of JMP, despite the value of alfa is indicated as 1.414, I get slightly different results of the actual values of factors for the axial levels for CCD-orthogonal blocks or central composite design for example. Why is that? Thanks

Re: alfa value for rotatability and orthogonal blocking in response surface designs

Dan_Obermiller — Thu, 22 Jul 2021 12:22:44 GMT

In the Response Surface Design platform, the axial value that is displayed is for a rotatable CCD, which is the default. Once you specify to have orthogonal blocks, the axial distance MUST change to the orthogonal axial distance because it is orthogonal. If you want to know the coded axial distance, look at the orthogonal axial distance.

Re: alfa value for rotatability and orthogonal blocking in response surface designs

MFVIT — Thu, 05 Aug 2021 04:53:12 GMT

Thank you for your reply.

I noted that even though JMP shows alfa=1.414 for two different designs (rotatable and blocked) the actual values of the factors are slightly different.

Re: alfa value for rotatability and orthogonal blocking in response surface designs

Dan_Obermiller — Thu, 05 Aug 2021 17:04:18 GMT

Center points will make a difference on the axial distance. Plus, the labeling can be confusing with some of the designs as the labeling in the design catalog is consistent only if the axial distance is left at rotatable.

Re: alfa value for rotatability and orthogonal blocking in response surface designs

MFVIT — Thu, 12 Aug 2021 10:29:08 GMT

Many thanks, this helped me.