I am using a mixed repeated measures with an ARMA(1,1) variance covariance structure. I have a basic understanding of why I am using this model etc but I need some help on articulating a few things:
How does this model correct for the variance in time between each time point? Or does it ignore the problem? In which case then why is it acceptable here?
What happens to variables that are correlated? When it is assigning variance to one variable or the other? How does this analysis do this?
Why this over repeated measure? Less error? More power? How much is imputed?
I have formulated some thought/answers, but I am curious to hear from others who might actually have used this model more than me.
Thank so much