MMRM: How to test that a marginal mean is different from zero

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

MMRM: How to test that a marginal mean is different from zero

AndreasV
I have  a simple mixed model for repeated measures with visit as the only
factor that analyzes change from baseline for serial measurements of an
outcome of interest controlling for the baseline value. My syntax looks like
this:

MIXED Change BY Visit WITH Baseline
  /FIXED=Visit Baseline Visit*Baseline | SSTYPE(3)
  /METHOD=REML
  /REPEATED=Visit | SUBJECT(Subject) COVTYPE(UN)
  /EMMEANS = TABLES(Visit).

Note that 'Change' at Visit 'x' is the value for the outcome of interest at
Visit 'x' minus the baseline value, i.e., the intra-individual change from
baseline.

In the EMMEANS subcommand, how do I test that Change at Visit 'x' is
different from zero?

Thank you very much!

Best, Andreas




--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD