simr is an R package for conducting power analyses on linear mixed effects models. One might compute power for a model with:
library('simr') fit = lmer(data=data, formula = x ~ y + (y | z) powerSim(fit, nsim=200)
My question is this: does it make sense to use this package assess the power of models that were not built with lmer? For example, on a simple ANOVA:
fit2 = aov(data=data, formula=x~y) powerSim(fit2, nsim=200)
Just to clarify, I know that this CAN be implemented in R (i.e.
powerSim(fit2) does not cause an error, and returns output similar to that of
powerSim(fit) ), but I'd like to know if using
powerSim on an ANOVA model is theoretically flawed.