Interlude stuff

Question 1

What is the variance-covariance structure

Answer 1

In simple terms how much some particular values are being correlated with other values. Attempting to model that values from subject A are highly correlated with values of subject A. When you specify your effects structure, you are also specifying this var/covar structure

Question 2

Sphericity

Answer 2

Assumes that all variances in your variance-covariance matrix are the same. The autoregressive assumption can be an example of structure specification. By default, we use the least squares assumptions

Question 3

What is the between-subject variable doing to the MLM?

Answer 3

Allows you to predict observed variations you have across subjects

Question 4

What is shrinkage?

Answer 4

The idea that the group estimate influences individual estimates and vice versa

Question 5

How do you treat time

Answer 5

Pre / post? = categorical

Things over time = continuous