I have a question about estimating between-person associations in multi-level VAR models with two variables (X and Y).
Is it correct that there are two approaches to get insight in the between-person association between X and Y: 1. By including the correlation between the variance of X and Y at the between level: "X with Y" 2.Explicitly predicting Y from X at the between level: "Y on X"
What is the difference between both approaches? And are there situations in which one of these approaches outperforms the other?