Since I don't have much experience in research, I need your help. Is correlation a requirement in mediation analyses? In my study, the Pearson correlation coefficient between DV and IV was not significant. However, when I put gender and age as control variables, this relationship turns out to be significant. So in short, there is a partial correlation between DV and IV. Can I still continue with mediation analysis? Should I share the findings on partial correlation in the article?

From what I've looked into, Hayes says correlation is not a requirement. How can I explain this situation to the jury where I will defend my thesis?

I use PROCESS Macro in my analysis. Significant effects were found for X on Y and M, and for M on Y.

X=IV

Y=DV

M=Mediator

More Yunus Eskiocak's questions See All
Similar questions and discussions