Normality Assumption for Mixed Model Analysis

[twix17]

Hello community,

I am doing some analysis for repeated measure data and using a mixed model method to carry out my analysis to accommodate missing data. Below are my lines of code and i have also attached the sample data  

proc mixed data=data;
class trt sub time (ref=first);
model Y = time trt time*trt/ s chisq;
repeated / type=un subject=sub r;
run;

I understand I needed to carry out model diagnostic before for modelling my data. Can anyone help on how I can perform Normality Assumption for my data with SAS code.

Thanks

[PaigeMiller]

My understanding is different. You test the normality assumption AFTER you fit the model, by seeing if the residuals are normally distributed. There is no need for the response or predictor variables to have a normal distribution.

[lvm]

Best approach is to look at the studentized residuals in graphic form. Add PLOTS=studentpanel on the procedure statement. See the user guide about how to interpret. Normality is not overly important (within reason), but constant variability is important. 

[lvm]

Your choice of an unstructured covariance matrix (UN) could be problematic if you have many times. The number of parameters to estimate for variances and covariances grows geometrically with time. I recommend you check out the user's guide for examples of other choices for type=. 

The second example in the MIXED user's guide is a good place to start, if you haven't already read it. Or read:

https://www.jstor.org/stable/1400366?seq=1

[twix17 · Accepted solution]

Thank you. I find this information helpful and implemented the suggestion.