topic Re: Joinpoint regression in Statistical Procedures

Joinpoint regression

gzr2mz39 — Sun, 15 Aug 2010 22:05:39 GMT

Does Base SAS 9.2 have a proc for joinpoint regression?
Thank you.

Re: Joinpoint regression

Cynthia_sas — Mon, 16 Aug 2010 16:30:41 GMT

Just curious....this web site implies that "Joinpoint Regression" is done using the software programs available from this site:
http://srab.cancer.gov/joinpoint/
http://srab.cancer.gov/joinpoint/example.html#STEP3
(National Cancer Institute).

Do you want to perform this kind of analysis with SAS instead of using the JoinPoint software??? Or do you just want to produce this kind of graph with SAS after you have calculated the data points? This might be a question for Tech Support -- I would expect you'd have to know how the math behind the JoinPoint software compares to the math of the several SAS regression procedure techniques.

cynthia

Re: Joinpoint regression

gzr2mz39 — Mon, 16 Aug 2010 16:44:18 GMT

Yes, I'm wondering if SAS can perform the analysis that the joinpoint software program performs.
If not, can Base SAS perform any similar trending statistics?
Thank you.

Re: Joinpoint regression

Dale — Tue, 17 Aug 2010 17:45:30 GMT

You might want to look at a post I made to SAS-L a couple of years ago. In that post, I addressed fitting a segmented regression model. I only presented mechanics for fitting a model with a single join point and I did not show any of the permutation distribution testing. But that post might provide you with some basic methods which you can extend as necessary.

See: http://www.mathkb.com/Uwe/Forum.aspx/sas/33115/segmented-regression-problem