| Date: | Tue, 22 May 2001 20:35:26 +0100 |
|---|
| Reply-To: | "cvo.mail" <cvo.mail@BTINTERNET.COM> |
|---|
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU> |
|---|
| From: | "cvo.mail" <cvo.mail@BTINTERNET.COM> |
|---|
| Subject: | Test for sphericity in univariate approach to repeated measures |
|---|
| Content-Type: | text/plain; charset="us-ascii" |
|---|
I have done a repeated measures experiment.Because the design is
incomplete (latin Square)I can't use GLM-repeated measures. Instead I am
using GLM-univariate, having entered individual measurements on
individual rows of the spreadsheet and added "subject" as an additional
factor,
e.g.:
subject1 measurement1
subject1 measurement2
subject1 measurement3
subject2 measurement1
subject2 measurement2
subject2 measurement3 etc etc
How can I test for the sphericity assumption and use appropriate F-ratio
adjustments? Mauchly's test, Greenhouse-Geisser and Huynh-Feldt
corrections are not available under GLM-univariate!
Carl Van Oldenbeek
Consultant Anaesthetist
Trafford General Hospital
| 
