I have biomarker data from a group of subjects repeatedly sampled over time. I'm using PROC MIXED for the model, and Tukeys as a post hoc. How would I generate a connecting letters report from the Differences of Least Squares Means? I've seen macros for PROC GLIMMIX like %MULTI, and connecting lines. Does anyone know of a macro or script that will work with PROC MIXED instead?
proc mixed data=data; class age sex tp name; model biomarker = age sex age*sex; repeated TP / type=CSH r rcorr subject= name; LSMEANS age / pdiff adjust=TUKEY; LSMEANS age*sex / pdiff adjust=TUKEY; LSMEANS sex / pdiff adjust=TUKEY; run;
Week | Weight | Connecting Letter |
4 | 28.8693 | A |
12 | 30.959 | AB |
16 | 33.7827 | B |
20 | 40.0923 | C |
PROC MIXED plus PROC PLM
You favorite internet search engine will find this:
A SAS macro for generating letter displays of pairwise mean ...
I don't know if it matches your problem as my company's firewall won't let me access it.
Yes, I'm trying to find a macro like that but for PROC MIXED. According to the paper that macro only works with PROC GLIMMIX and I have to use PROC MIXED since I have repeating measures.
PROC MIXED plus PROC PLM
It's finally time to hack! Remember to visit the SAS Hacker's Hub regularly for news and updates.
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.