From Pearson correlation to LMM Contribution


I have two variables X1 and X2 which are poorly correlated (Pearson; r=0.1).

Since we think the relation between X1 and X2 varies daily, I then use a linear mixed effect model (LMM) to try and predict X1 where
(1) X1 is the dependent variable, X2 is the fixed part and X3 (=date) is the random part.

I wonder if due to the poor correlation between X1 and X2, would I necessarily expect a low contribution of X2 to the explained variance of X1 in the mixed model ?

Thank you very much