I am conducting a moderated mediation analysis using Hayes' Process Macro Model 15. Although my mediation is significant, both the interaction effects are not significant. I am having an insignificant index of moderated mediation also but at the same time, I am having significant conditional direct as well as an indirect effect of X on Y at all the three levels of the moderator. How does this significant conditional direct and indirect effect come from?
How do I interpret the result?
Dear Sabnam,
Did you make any progress on this topic? I have similar concerns. Please let me know.
Thanks a lot!
Yours sincerely,
Shirley Yang
Dear Shirley,
If the index of moderated mediation is not significant that means our conditional indirect effect is not significant. Although there is a mediation effect, it does not depend on the level of the moderator. The significance at all the three different levels of the moderator means that the moderator is unable to produce a differential effect in the result. Higher level of the moderator leads to similar effect as the lower level of the moderator. Our interpretation will thus be based on the index of moderated meditation.
Hope this helps.
Dear Sabnam Basu , thanks for your generous sharing and it really helped me a lot!
I had the same problem... I found this:
"Using the SPSS macro introduced here and described earlier, we then estimated the coefficients of a moderated mediation model in which this indirect effect was presumed to be moderated by math self-concept. Note that a significant unconditional indirect effect does not constitute a prerequisite for examining conditional indirect effects".
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate behavioral research, 42(1), 185-227.
Hi Adem Kantar
As I can see your index of moderated mediation is significant and also the indirect effects at different levels of the moderator are also significant. This means to understand how your moderator is affecting this: (ECRS_x -> KD_X -> PYAİT_X ) relationship you have to plot the three levels of the moderator as is also done in simple moderation analysis. By plotting the three levels of the moderator you would be able to see which level has the highest effect, which has a moderate effect and which has the lowest effect.
As you have quoted me saying "The significance at all the three different levels of the moderator means that the moderator is unable to produce a differential effect in the result". This is true when the index of moderated mediation is insignificant. As your index of moderated mediation is significant this means moderated mediation is taking place, i.e. your mediation is contingent on the levels of the moderator. In order to understand how, you need to plot this and understand the three levels. As all the three levels of the moderator are having a significant effect and the direction is also negative for all the three levels, you have to see the magnitude i.e. the effect size, in order to understand the effect.
Best Wishes.
In my moderated mediation model interaction term is not significant but conditional direct, indirect and index of moderated mediation are significant. However after performing boot strapping interaction term is also significant. How to interpret the result. Kindly suggest me.
Dear mam ,
In my moderated mediation model interaction term is not significant but conditional direct, indirect and index of moderated mediation are significant. However after performing boot strapping interaction term is also significant. How to interpret the result. Kindly suggest me.
Hi Nisrutha Dulla
As you have said that the index of moderated mediation is significant, this means moderated mediation has taken place. As for the interaction term, I did not understand your question. How did you get an insignificant interaction term before running bootstrapping?
Hi Sabnam Basu
I am having a similar problem. I work with the Hayes' Process Macro Model 7. I am having an insignificant index of moderated mediation also but at the same time, I am having significant conditional indirect effect of X on Y at all the three levels of the moderator. You wrote "The significance at all the three different levels of the moderator means that the moderator is unable to produce a differential effect in the result. Higher level of the moderator leads to similar effect as the lower level of the moderator". Does that mean I can ignore the significant conditional indirect effects in my interpretation? And do you have a reference or something for that?
Thank you so much!
Hello Zoe Olbermann
If you have an insignificant index of moderated mediation that means moderated mediation has not taken place. The index is indeed the main criteria for testing the significance of the overall moderated mediation model. However, your mediator and moderator might be individually significant in explaining the relationship of X on Y. Based on the aim of your study you might report your results. Sorry, but I do not have any reference material for this.
Hi Sabnam Basu - I have a categorical moderator and am using effect coding. This gives me 4 values in the index of moderation- only one of which is significant (representing group 1 for the moderator). But the conditional indirect effects are only significant for groups 4 and 5. How can I interpret this? Why might there be no conditional indirect effect for group 1 if that is the only one showing a significant index of moderated mediation?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size