Statistical FAQs

Why does JMP use the default option center polynomials and how does this compare with PROC GLM?

If you don't use center polynomials, the parameters on lower-order terms are less meaningful. The benefit of centered polynomials is that they make the test for the main effect independent of the test for the squared term.

If you are using the center polynomials option, the Parameter Estimates you will receive are centered and in the scale of the actual factor settings (as opposed to being orthogonally coded) -- the main effects are NOT centered. All continuous terms that are involved in cross terms will be centered by the mean and the parameters are so labeled. Centering gives consistent tests that are location invariant.

The following SAS code demonstrates how to reproduce the JMP results using PROC GLM.

/******************************************************************************
SAS INSTITUTE INC. IS PROVIDING YOU WITH THE COMPUTER SOFTWARE CODE INCLUDED WITH THIS AGREEMENT ("CODE") ON AN "AS IS" BASIS, AND AUTHORIZES YOU TO USE THE CODE SUBJECT TO THE TERMS HEREOF. BY USING THE CODE, YOU AGREE TO THESE TERMS. YOUR USE OF THE CODE IS AT YOUR OWN RISK. SAS INSTITUTE INC. MAKES NO REPRESENTATION OR WARRANTY, EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT AND TITLE, WITH RESPECT TO THE CODE.
******************************************************************************/

data test;
    input pressure totalflow gasratio thickness ;
    cpres=pressure-180;
    ctflow=totalflow-450;
    cgasrat=gasratio-4.5;
cards;

180	450	4.5	900.87	15.53
160	500	6	1060.157	16.14
200	400	3	1052.14	15.99
160	400	6	1220.278	12.98
180	450	4.5	940.87	15.62
160	500	3	565.6048	16.35
200	500	3	1278.004	15.83
180	450	4.5	912.87	15.73
200	500	6	757.558	16.71
200	400	6	569.0499	15.64
180	450	4.5	935.87	15.55

;
/* GLM reproduces results from JMP Center Polynomial */
proc glm data=test;
    model thickness=pressure totalflow gasratio cpres*ctflow
                               cpres*cgasrat ctflow*cgasrat;
run;
/* GLM reproduces results from JMP without the Center Polynomial */
proc glm data=test;
    model thickness=pressure totalflow gasratio pressure*totalflow
                                pressure*gasratio totalflow*gasratio;
run;

Note: If you don't want polynomial centering, use the menu on the Fit Model dialog title bar to turn it off.

FAQ # 2083
Last Updated: 2004 Jun 03

919.677.8008 (US)

International Support via Worldwide SAS Offices

Statistical FAQs

Why does JMP use the default option center polynomials and how does this compare with PROC GLM?

Browse JMP FAQs

Support Documents

Contact JMP Support