topic Re: Can there be different number of timepoints for Generalized Linear Mixed Model? in Statistical Procedures

Can there be different number of timepoints for Generalized Linear Mixed Model?

Liz123 — Wed, 03 Jan 2024 23:10:31 GMT

Hello everyone,

For generalized linear mixed model analyses, can there be different number of time points for each individual or study participant? I am aware that it accommodates varying timepoints, but I am not sure about the the number of timepoints. For example, can an individual have 5 timepoints and another 7 timepoints and another individual 30 timepoints or do they need to have equal number of timepoints for one to conduct this type of analyses? Is there a maximum number of timepoints especially for graphical presentation?

Also, can anyone share resource(s) on how to interpret the SAS Output for Generalized Linear Mixed Model/ how to report the results? Thank you.

Re: Can there be different number of timepoints for Generalized Linear Mixed Model?

Ksharp — Thu, 04 Jan 2024 03:13:22 GMT

“can an individual have 5 timepoints and another 7 timepoints and another individual 30 timepoints”
Yes. You can have different timepoints within different subject.
But you need specify a variable in REPEAT , like:
repeat time/subject=subject type=ar(1);

https://stats.oarc.ucla.edu/other/mult-pkg/introduction-to-generalized-linear-mixed-models/
https://blogs.sas.com/content/iml/2019/12/03/longitudinal-data-response-profile-model.html
https://blogs.sas.com/content/iml/2019/12/05/longitudinal-data-mixed-model.html

could give you a beginning of Mixed Model.
or @SteveDenham could give you more advices .

Re: Can there be different number of timepoints for Generalized Linear Mixed Model?

jiltao — Thu, 04 Jan 2024 15:25:37 GMT

Yes, the number of time points can be different across subjects.

I am not aware of a limit on the number of time points in theory, but practically, it might run out of memory if you have too many time points and you use a complex covariance structure such as UN. Graphically too many time points might make the graph hard to reveal anything. But as long as your computer can handle it, you could have as many time points as you wish.

Output interpretations are provided in many examples in the PROC GLIMMIX documentation. For example,

https://go.documentation.sas.com/doc/en/pgmsascdc/v_046/statug/statug_glimmix_gettingstarted01.htm

https://go.documentation.sas.com/doc/en/pgmsascdc/v_046/statug/statug_glimmix_examples.htm

Hope this helps,

Jill

Re: Can there be different number of timepoints for Generalized Linear Mixed Model?

Liz123 — Fri, 05 Jan 2024 16:20:56 GMT

@Ksharp This is great! Thank you so much!

Re: Can there be different number of timepoints for Generalized Linear Mixed Model?

Liz123 — Fri, 05 Jan 2024 16:23:38 GMT

@jiltao Thank you for this! Highly appreciated!