Dear CONN experts.

Please see the following model used for my study. Two groups (exp. and control) underwent two scans (pre- and post- intervention). What I want to examine is whether or not there is an interaction between group and BDI scores when predicting changes in connectivity over time. In other words, whether group moderates the association between BDI and changes in connectivity/outcome. I believe this qualifies as a GLM multiple regression:

Between-subjects contrast
All Subjects
Group A
BDI (quantitative scale)
BDI*Group A (interaction variable)

[0 0 0 1]

Between-conditions contrast
Pre-
Post-
[-1 1; 1 -1]

I've been advised that a linear mixed effects model is optimal. If the interaction term (resilience * group) shows statistical significance, it suggests exploring within groups. Such a model also allows for the inclusion of time or interactions with time. Should these time-related factors prove significant, one can analyze the two time points independently.

It seems that my model above does this as a standard multiple regression. Does my model then qualify instead as linear mixed effects? I would think not because there is no mention of accounting for random effects. Perhaps I just need some clarity as to the labelling of my model and how it differs from a linear mixed effects model.

Thank you!

-Olivier