Not sure that G*Power can handle that. If you know how, the most flexible way to run any power analysis is to run Monte Carlo simulations. I will attach an annotated example that a friend sent me for SAS and MPlus.

Thanks very much for this Carl. Sadly my instition doesn't have either SAS/MPlus but will see if Monto Carlo simulations for power analysis could be run on another package. Thanks also for the syntax paper.

I have done this in R as well! Its totally free and open source, but it is another thing to learn. I will attach some example code (from Lee VanHorn at U of SC) if you want to have a look. Let me know if you have questions (or if I gave you the wrong file)!

Dear Carl and Amanda,

Belated many thanks for these very useful responses and offers to follow this up further with you via email. I am unfamiliar with using R for power analysis but will investigate this now.

Dominic

Let me add my two cents, to the very good above responses. PROCESS uses traditional ordinal least squares (OLS) for the regression at step 1 and 2 of your model, and in consequence you can check the power of each step using G*Power, as you would do in a normal mutlivariate/hierarchical reggression. However, in order to test the conditional indirect effect (moderated mediation), PROCESS uses bootstrapping to construct confidence intervals. Bootstrapping is a type of Montecarlo simulation, which is equal to generating data as previous respondants answered. In consequence, if you set the bootstrapped samples to more than 1000, you are very unlikely to have power issues. Hope this helps, and good luck with your research.

Can anyone recommend how to calculate sample size for Model 7 using G*Power?

Referring to the mentioned literature of Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185-227. doi:10.1080/00273170701341316 above:

For my study I would have Model 2, Path a is moderated by W. There indeed is in Table 4mentioned the empirical power for Model 2. My question now though is how to interpret it or how I would write it: I see the regression coefficients and I would look on BCa results. There are mentioned 4 Regression Coefficients .00, .14, .39 and .59 and then in line go 5 sample sizes and the belonging empirical power. Let´s say I pick a sample size of 50 - the minimum: if I want to know in advance / a priori how big my sample size has to be I look at the Regression coefficients and take .39 and see an empirical power of .705. What does that mean? Is it the effect size Cohen´s d? Which regression coefficient would I even pick? For a regression coefficient of .14 the power would just be .050 , which I suppose is bad and then I would need a bigger sample size?