This time we'll be talking about
first stage moderated mediation and this is the agenda for today.
First of all, what is this thing?
What is moderated mediation models?
Then we will show you how to theorize and then test for moderated mediation models.
So again, this is a hands on workshop,
a hands on session,
and you'll be able to run models on SPSS and use the process macro developed by Hayes.
So, first of all,
what is moderated mediation?
In this case, different from mediated moderation,
moderated mediation, we theorize about the mechanism.
So, first, is there,
what we want to show is that
this mediating process that links our independent variable to our dependent variable,
that interactive fact varies at different levels of the moderator. Okay?
So, this is different from
mediated moderation because with mediated moderation models we first
have a moderation and then we show
the mechanisms that connect that interactive term to the dependent variable.
At this time we are looking at the process first,
at the mechanism first,
and then theorizing that
that indirect effect varies at different levels of the moderator.
Is that clear? So, there are many different papers that
used mediating process or moderated mediation models, I would say.
And if you want to know a little bit more about
moderated moderation there are two papers that are really good,
Edwards & Lambert, 2007,
and Muller et al., 2005.
You can get more information about moderated mediation models.
With moderated mediation models we have two different types,
actually three different types of moderated mediation model.
One that's called first stage.
The second is second stage moderation or,
or moderated mediation models.
And the other one is dual moderated mediation models.
This last one we'll also cover here in our last session of this workshop.
This dual mediation, moderated mediation models,
it's not very, very used in our field
but we needed to cover it for the sake of completeness.
Okay? So, let's talk about first stage moderated mediation models first, okay?
So, what we are looking here is that the mediating process,
that relationship between job meaningfulness and job performance mediated by engagement,
that relationship is moderated by organization identification,
in particularly the relationship between job meaningfulness and job engagement.
That first path is moderated by our organization identification variable.
In this screen I'm showing you another model that has been published already and in
this particular paper we theorized that identified ambivalence
has an indirect effect on decision effectiveness via contextual awareness.
And this first path between identified ambivalence and
contextual awareness is moderated by trait self-control.
This is the second example.
There are many examples that you can find in top journals,
in journals that publish organizational behavior theories.
So, how do we theorize?
How do we go from an idea to a paper?
So, first I need it to theorize about this indirect effect,
about the mediation model.
So, the independent variable influences your dependent variable via your mediator.
So, this is a very similar figure as the one that you saw in one
of our initial sessions in the workshop about mediation and moderation.
That's a very common, very simple figure.
The second step in theorizing about
first stage moderated mediation models is to theorize about
the interactive relationship between
your independent variable and your moderator on your mediator.
So you'll need it to theorize about the effects of your independent variable on
the mediator and that being a function of your moderator.
Finally, you theorized about the whole theoretical model.
So, you can say, for example,
that job meaningfulness and fullness is job performance via job engagement but
that indirect effect is a function of your moderator organization or identification.
So that's how you'll theorize about first stage moderated mediation models.
And then how do you test for it?
So, again, it's pretty simple.
There are five steps.
In the first step you needed to adopt the current task for mediation.
We covered that in one of our initial sessions in the mediation and moderation workshop.
So, if you are not familiar with this current test for mediation,
please go back and watch that particular session.
The next step, the second step,
is about the traditional test of moderation of the independent
and of the dependent variable and the moderator on this mediator.
So, again, it's a moderation model.
If you have questions you can go back and watch
this session that covers moderation in more depth.
What's new here is the step number three and now we are
adopting PROCESS and running model number eight.
If you recall we used the same model,
model number eight, to test for moderated mediation or mediated moderation.
What changes, again, is
that theoretical development and the steps before getting to this,
step three, this PROCESS model number eight.
Okay? So, theoretical development and the steps that
you get to PROCESS model number eight,
those are the differences.
Then you run a bootstrapping analysis and look at the index of a moderated mediation.