Hello Swati,

Yes, there can be multiple instances of mediators/moderators in a SEM design. As well, one variable that is endogenous (an outcome) for another (or others) can be exogenous (or an influence on) another (or others).

Here's a RG link that shows one DV being used as an influence on other DVs, directly and indirectly: Article Consumers’ and Physicians’ Perceptions about High Tech Weara...

Good luck with your work.

Thank you for your reply!

Swati,

have a look at the following article in which multiple mediation is explained:

Article Conducting Mediation Analysis in Marketing Research

Although your DV2 is a dependent variable with regard to your IV and M1, it is also a mediator variable, because it is an independent variable with regard to M2, M3, and DV2.

I guess that you model looks like this:

IV -> M1 -> DV1 (M2) - > M3 -> M4 -> DV2

As you can see, the effect of IV is transmitted through M1, M2, M3, and M4 to DV2. A variable that is a dependent variable and an independent variable at the same time is a mediator variable, therefore DV1 is M2.

Hope this also helps,

Karin

It is of course possible that your IV also has direct effects on M2, on M3, and on M4 - this depends on your theoretical background.

Michael R. Frone I do not have sequential mediation but a parallel one. My problem is not with the position of mediators but the variables. I could not figure out if an endogenous variable could become an exogenous one after its relation with another variable has already been mediated by a mediator. Should I present the two instances of mediation in the same model or should I break those down into two models? Its not sequential because I am not trying to establish any relationships between the mediators, nor am I presenting them in a sequence.