Subscribe Bookmark RSS Feed

Does JMP have the curvilinear logistic regression function?

xiaoyaj0

Community Trekker

Joined:

Apr 25, 2016

Does JMP have the curvilinear logistic regression function? In my dataset, I would like to predict diagnosis of diabetes from body mass index (BMI) using logistic regression, and the quadratic model reduced the Chi-Square by 114.91, and it is a significant model. This result is from SPSS. How can I do the same analysis in JMP? Thank you!

2 REPLIES
txnelson

Super User

Joined:

Jun 22, 2012

If you are referring to having Polynomial terms for a logistic regression, JMP has that capability.  Just set the degree value to the level of curvature you want to include, select the column to apply it to, and then under the "Macros", select "Polynomial to a Degree".  The example below has all 2 Effects set to a quadratic polynomial

12875_pastedImage_0.png

Also, I don't know what a .sav file is, so I don't know how to open it.

Thanks to Ron(below) I now know what a .sav file is.

Jim
ron_horne

Super User

Joined:

Jun 23, 2011

Hey xiaoyaj0

just open the SPSS file you have in JMP

set the modeling type of Diabetes to Nominal

12873_pastedImage_0.png

go to Analyze>>Fit model and set diabetes as you Y. select the z score from the list on the left an from the drop down menu choose Polynomial to degree and click Run.

12874_pastedImage_1.png

then you will get the result you are looking for

12878_pastedImage_2.png

best,

ron