Hello,

In my dissertation I'm examining a mediation model which states that social-connectedness mediates the relationship between nostalgia and self-esteem. However, in my study I added the covariates of participant's gender and age.

Both the original model and the model with these covariates are significant, however it seems that the model with the added covariates explains more variance.

I was wondering if this enough to argue that the mediation model with the covariates is better or whether there was test I could conduct to confirm this?

Thank you!