Bioequivalence and Bioavailability Forum

SAS PROC GLM with group effect [General Statistics]

posted by d_labes – Berlin, Germany, 2014-05-22 16:11 (3623 d 21:07 ago) – Posting: # 12988
Views: 14,113

Dear Silva,

Using your code:

regarding the problem with df=0: verified
change the model statement to
model lnCMAX= group sequence form form*group period subject(group*sequence);
to solve this issue as ElMaestro suggested.
regarding no output of LSMeans: not verified
Error diagnostics without data not feasible.
Did you encounter any errors during submitting your code?
Is the above exactly your code? I see a missing semicolon.
regarding different signs in point estimator and CI: Phoenix clearly calculates Ref-Test (look at the numbers!), so you have to change the sign to obtain Test-Ref. and then all is SAS and phoenix coincide.

BTW: The link in my post below (FDA suggestion to deal with group effects) the model statement became more involved:

model lnCmax = group sequence form subject(group*sequence) period(group) 

               group*sequence group*form;

Not quite clear which denominator has to be choosen for all that terms.

Hope this helps.

@Helmut: This thread has nothing to to with Regulatives/Guidelines I think. You may consider shifting it.

Done. [Helmut]

—
Regards,

Detlew