| Date: | Mon, 11 Jan 1999 14:57:25 +0100 |
|---|
| Reply-To: | Esther de Ruyter <E.deRuyter@FSW.RUU.NL> |
|---|
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU> |
|---|
| From: | Esther de Ruyter <E.deRuyter@FSW.RUU.NL> |
|---|
| Subject: | discriminant analysis |
|---|
| Content-Type: | text/plain; charset="us-ascii" |
|---|
Hello,
I want to reproduce some output in SPSS Windows 7.5, that was originally
made in SPSS/PC+. When I use the stepwise technique for discriminant
analysis in Windows (/method = Wilks), I don't get the same model as in
SPSS/PC+. Does anyone know what exact criteria were used default in
SPSS/PC+? When I do the analysis in SPSS 7.5, I get a model with only 2
variables. So, the criteria were less strict in SPSS/PC+, where I got a
model with 10 variables. Can someone help me with this frustrating problem?
Esther de Ruyter
|
