| Date: | Mon, 22 Jun 1998 13:19:18 GMT |
| Reply-To: | bjj103@MY-DEJANEWS.COM |
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU> |
| From: | bjj103@MY-DEJANEWS.COM |
| Organization: | Deja News - The Leader in Internet Discussion |
| Subject: | CHAID w/continuous variables |
|---|
My knowledge of SPSS is nearly a decade out of date. Can someone tell me
whether one can use CONTINUOUS (interval or ratio-level scaled) variables as
independent variables in an SPSS-CHAID tree? Specifically, I have a binary
outcome (Dependent variable) I am looking to classify using a bundle of both
nominal and continuous independent variables. I'd like to see the tree
contain some of these continuous variables split at some optimal "cut point"
should they help classify the outcome variable. However, my research vendor
claims this isn't possible with SPSS-CHAID, that SPSS's CHAID product works
only with nominal independent variables.
My experience with CART and prior expectation of SPSS suggests SPSS can indeed
handle both types. Is this not true?
Much, much thanks in advance.
Brent Johnson,Ph.D.
Intel corp market research
brent.johnson@intel.com
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/ Now offering spam-free web-based newsreading
|