Medy Freddy posted on Wednesday, November 12, 2014 - 1:01 pm
I need some guidance. I am interested in conducting a experimental mediation analysis with a binary IV and repeated measures of the mediator and DV. As a simple illustration:
-Ind Var(Treatment/matched control group)
-Mediator (8 weekly measures of the same variable)
Dep Var (pre/post measures of outcome variable measured before and after all weekly measures of the mediator)
What I expect is that there is a significant improvement in the DV from pre-post in the treatment grp and not the control grp....and that this improvement is mediated by a linear weekly improvement in the mediator. I have some understanding of mediation with LGCM using repeated measures of the mediator, though I'm unsure if this is advisable with a DV with only two time points (I've read that 4 time points are recommended). Am I correct or is there an accepted method of doing this? Thanks in advance.