0
$\begingroup$

I am conducting a moderated mediation analysis using Hayes' Process Macro Model 7. I am having a significant index of moderated mediation, significant a path and b path and significant moderator X Predictor(X) interaction. Simple slopes show moderation and I also was able interpret mediation b/c a and b path are significant at all levels of moderator. the only thing that isn't significant is the c' path (Direct path).How do I interpret that? the pearson's correlation (r) between IV (x) and DV(Y) is significant and all the studies say that X is significant predictor of Y. what does not having a c' prime mean?

$\endgroup$
0
$\begingroup$

Since the goal of mediation analyses is to test the mechanism through which X affects Y, a non significant direct effect (c') can be interpreted to mean that the total effect between X and Y runs primarily through the indirect effect (i.e., via your hypothesized mediator M).

Some people may call this scenario "full mediation", but I strongly recommend that you avoid using this term because of the reasons given in Zhao, Lynch, & Chen (2010). Namely, (1) you could simply fail to observe a significant c' because of low statistical power. (2) Claiming full mediations may lead to the (often incorrect) conclusion that no alternative mechanisms for the X-Y effect exists. After all, if M fully explains the effect of X on Y, then why do any more study on this effect?

So what should you do instead? I'd reckon that the top psych journals would require that you report the conditional indirect effects. That is, what are your 'ab' coefficients at various levels of the moderator? Also do report the c' coefficient since it is a useful information. Just note that the direct effect is not statistically significant and move on to interpret your main results.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.