I ran a Generalized Linear Mixed Model in SPSS and included 3 measurement levels (D1, D2 and D3) with 2 between-subjects groups (A and B) and the interaction effect between time and group. The interaction was not significant, but the main effects (time and group) both were. It is valid to run between and within post-hoc tests to investigate the main effects? For instance, run post-hoc tests to know if on D1, group A is different from group B. Or, a pairwise test to know if group A in D1 is different from group A in D2?