Open KJames5 opened 1 year ago
Problem is with the test. fcompare
keeps random effects the same and modifies the fixed part of the model, which is ~ tile
in both cases.
If you were trying to test the fixed effect you probably want test = fcompare(~1)
.
If you were trying to test the random effect probably test = random()
.
Hi,
Thanks that makes more sense! Thanks for making that clear!
Hi! I'm having issues running a simple GLM in simr(). I want to test the effect of 1 factor (3 groups), with site included as a random factor (3 groups). I don't have any issues when I include 2 independent factors in the model (as a interaction or just including both in the model) so I'm not sure why it is having errors. When I run a powerSim(), the output has a power of 0%, with a warning of "Test did not return a p-value" for each simulation. Here is my code:
Do you have any idea why this might be? I have no issues when using the same data frame when I include material in the model. Thanks!