- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hello,
In a proc mixed how do I classify the type of variance structure I want to use (Unstr, variance comp, comp symm, etc)?
Here's my code.
PROC MIXED DATA = work.pracL;
CLASS ID;
MODEL logmemI = demind age age*demind spline gender ses/solution cl outpm= work.plot1;
RANDOM Intercept age/SUBJECT=ID ;
Estimate "Hyp 1 age slope for healthy individuals B2" age 1/cl;
Estimate "Hyp 3A Slope in MCI prior to chgpt B2+B3" age 1 age*demind 1/cl;
Estimate "Hyp 3B Slope in MCI after chgpt B2+B3+B4" age 1 age*demind 1 spline 1/cl;
Estimate "Hyp 4 Comparing slopes between Healthy and MCI before chgpt B3" age*demind 1/cl;
Estimate "Hyp 5 Comparing the slopes before and after chgpt for MCI B4" spline 1/cl;
Estimate "Hyp 6 Comparing slopes between Healthy and MCI after chgpt B3+B4" age*demind 1 spline 1/cl;
run;
title1; title2; title3; title4; title5; title6;
Thanks!
Accepted Solutions
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Greetings.
To choose which variance structure to use include the TYPE= sub-option in the RANDOM line.
For example, RANDOM .... / TYPE=UN. This would be unstructured.
Hope this helps.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Greetings.
To choose which variance structure to use include the TYPE= sub-option in the RANDOM line.
For example, RANDOM .... / TYPE=UN. This would be unstructured.
Hope this helps.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Thank you!