I have a dataset with 16 repreated-measures (values) per subject, 4 belonging to conditon A and 12 Belonging to condition B.
I am not interested in the effect of the repeated measure within condition, and I was wondering if it is necessary to model them, or if having multiple values for the same condition-group-subject trio is ok.
I see 2 ways of running the model: (1) use all values as they are (i.e. 16 values per subject); (2) average the values per condition before modeling (i.e. 2 values per subject, one for condition A and one for condition B).
This is the general model:
y ~ Group*Condition + (Condition|Subject)
Would the first strategy be correct? And if not, why?
I have tried to run the model with each value-subject pair modeled individually (as below), but I was hoping to save time by adopting one of the 2 strategies above, as I am really only interested in these 2 conditions.
y ~ Group*Value + (Value|Subject)
Thank you for you help!!