Hi there,

I've run a mediation analysis based on data collected 12 weeks apart. I therefore have time 1 data, time 2 data, and difference scores, which a third mediation analysis was run from.

I'm measuring the predictor on an outcome and also with the role of a mediator.

At time 1, I found:

- the direct effect between predictor and outcome was significant

- the indirect effect between predictor > mediator> outcome was also significant

- the total effect was significant

At time 2:

- the direct effect between predictor and outcome was significant, again

- this time, the indirect effect was non-significant

- but the total effect was significant once again

How do I interpret these findings?

In time 1, is it showing that the predictor directly affects the outcome, AND so does the mediator (aka, the relationship is mediated by the mediator)?

But, in time 2, the relationship is NOT mediated by the mediator because the indirect pathway is insignificant? And so the predictor affects the outcome measure independently of the mediator?

Kind regards.