 |
 |
Hello! We have a 1 x 3 repeated measures design (levels A, B, C) and 4 subjects (g, h, j, & k; not really, but for the sake of simplicity...), and based on the notes on the FSL website and the forum, we have structured our higher level EVs as follows, with k-1 columns for the levels (C is now the "reference" level) and a column for each subject to act as a random effects regressor for subject-wise variability: A B g h j k 1 0 1 0 0 0 0 1 1 0 0 0 0 0 1 0 0 0 1 0 0 1 0 0 0 1 0 1 0 0 0 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 0 0 0 1 0 1 0 0 0 0 1 0 1 0 0 0 1 0 0 0 0 0 1 Based on this we are pretty sure we can get the following contrasts: 1 0 0 0 0 0 This should give A-C 0 1 0 0 0 0 This should give B-C However, we're unsure about whether the following ones are correct, given that we haven't modeled an intercept (we have traded in the intercept column in favor of the 4 subject columns): 1 -1 0 0 0 0 Is this A-B? 0 0 0.25 0.25 0.25 0.25 Is this C? We continued to be a bit flummoxed as to when to include the intercept and when not to, given that we are actually interested in all three levels (there is no "control" so to speak). Confirmation, denial, or other guidance would be greatly appreciated. Best, Elisabeth Ploran Postdoctoral Fellow George Mason University