Family Research Quantitative Methods

These are good links. Mediation and moderation are always confusing… Lately, I’ve been thinking about the possibility of mediating moderator variables. This stems from a talk given by Dr. Niall Bolger in the Psychology department here, during Fall, 2007. I always thought of moderator variables as a switch, in which they are either “on� or “off,� and are randomly assigned (such as gender). I’ve always thought of mediating variables as explaining how X causes Y through an alternate path, instead of a direct path, which sometimes can be more helpful in explaining variance than the direct path. But it seems to me like a really good mediating variable is a moderating variable. It’s confusing, but interesting.

I agree that these are very good links. I have not done a great deal of work with mediators and moderators prior to last semester when I took a regression statistics course. I appreciated how the website not only illustrated what a mediator and moderator looked like, but it also provided detailed instructions on how to "test" for their presence in an actual analysis. These links have helped me to solidify my understanding of moderators and mediators and will surely be very useful with future research.

Mediators and Moderators still confuse me a bit because much of their respective definitions can be left to semantics. I think this is because the terms are misused or used differently in places beyond quantitative research. But I am struck that in qualitative research the terms 'mediator' or 'moderator' aren't seen all that often, but the concepts of intervening variables are often discussed in terms of context. I am wondering if it is my own limited attention to these terms that makes me think they are not used cross-methodologically (i.e., in qualitative research) or if the language is really that different for similar concepts...

I think the mediator link points out a very important point which often does not receive the attention needed in quantitative research overall. Any assessment of effects is only as good as its measurement tool. Reliability and validity in intrumentation are crucial in making statements about effects, in terms of thinking about mediating effects or any other area.

As a visual learner, the diagram provided on the mediator web-site was quite helpful. It would have been helpful, however, to provide a concrete example of what M (the intervening or process variable) and c' (the direct effect) might be in an applied situation. On the moderator web-site, the concrete example made it quite clear in terms of how we can apply M, X, and Y in a given situation. In the same situation provided in the moderator example, what would M and c' be in the mediator example? For me, providing concrete examples might help clarify how the two are different.