|Date: ||Wed, 21 May 2003 17:53:34 -0700|
|Reply-To: ||"Michael R. Healy" <email@example.com>|
|Sender: ||"SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>|
|From: ||"Michael R. Healy" <firstname.lastname@example.org>|
|Subject: ||Within-subjects data and the NLR procedure.|
|Content-type: ||text/plain; charset="US-ASCII"|
I have a data set with 8 measurements per subject. I would like to fit
a nonlinear regression model to each subjects data. The NLR procedure
requires data to be in column format where, ostensibly, each row represents
one subject. To accommodate the NLR data requirements, I've set my data
file up in a 'between-subjects' format, where every 8 rows represents one
1. Is it valid to fit NLR models to within-subjects data in this way?
2. If not, are there alternative procedures to fit curves to within-subjects
data in SPSS or any other package? The complexity of the regression model
prohibits contrast coding in GLM.
Thanks in advance,