I want to compare the four different groups, if they have equal variances, I could use ANOVA. However, the groups have unequal variance, I need to model the variances separately for each group instead of using the pooled variance. I know how to do that in SAS:
The code is:
Proc Mixed Data=Temp Plots=All;
Model Y= Group/ DDFM=Satterth Solution outpm=Resid Residual;
repeated / Group=group;
LSMeans group/diff=control('' ') adjust = dunnett pdiff CL AdjDFE=Row ;
How can I do it in R?