Dear SPM:
I have a design in which 10 subjects are studied in 3 conditions.
I want to perform a ranfx anova (no constant term) so that each of the
conditions can be examined separately and so I can mask by various
differences among the conditions.
Since the same subject is being used in each of the three conditions I'm
guessing I should use a non-sphericity correction
question 1) replications are over?
I answered groups.
question 2) correlated repeated measures?
I answered no since the tasks are different.
Is this correct?
Thanks,
Darren
|