Dear FSL experts,
I have an event related design that contains 44 different EVs. However, I'm only interested in the activations associated with ~6 of these EVs.
My question is, should I model all 44 EVs, or would it be more prudent to model just a selection of these variables (lumping relevant variables together) to prevent overfitting my model and to gain more power?
Many thanks in advance for your feedback,
Lara
|