Model 4 mediation PROCESS with dichotomous mediators

Question

I'm doing a PROCESS Model 4 mediation analysis, where:

• X = Persuasion Knowledge (continuous)

• M1 = Brand Attitude (continuous)

• M2 = Brand Awareness, which consists of aided brand recall (dichotomous) and unaided brand recall (dichotomous)

I tried putting it into PROCESS with three mediating variables, but I get the error that the mediators are dichotomous.

What can I do? I cannot find any literature that merges aided recall and unaided recall in one variable... should I just do that? Or is there a way to use 3 mediators in PROCESS from which two are dichotomous variables?

Answer 1

I'm not sure whether there is a way to do that in process, but what you could do is manually go over the steps that process uses, i.e. to conduct a series of multiple regressions,

1. X predicting Y (path c)
2. X predicting M1 (path a)
3. X predicting M2 (the second path a), and
4. X, M1, and M2 all predicting Y (paths c', and b).

If you have paths a and b significant for a given mediator, I suppose you could say that there was an indirect effect of X on Y through that mediator.

I am not sure how to get confidence intervals from this procedure though. Also, I am not sure whether this procedure is something a hardcore statistician would object.

However it is a way for you to see whether it makes sense to talk about mediation or not. Hope this helps