Dear SPM-experts,
I'm performing analysis on a study which includes 3 groups of different
age and 4 conditions (as contrasts) in a 2X2 design (A1>A1baseline,
A2>A2bl, B1>B1bl, B2>B2bs). I'm interested in the differences among the
three age groups in each of the conditions. There are also two
covariates in the analysis: 1) years of education and 2) the scores of a
behavioral task (Task2) which resulted to be different among the three
groups.
This is what I did so far:
- 4 different multiple regression analyses (spm8), one for each contrast
of interest, including age, education and task2 as covariates. Then, I
looked at the results from the age contrast only.
I want to run categorical analysis as well (since age is not linearly
distributed across the whole population), so my question regards the
most appropriate method to perform this analysis, which of these?
- flexible factorial including just "conditions" and "group" as factors
and including education and task2 as covariates
OR
- full factorial including the three groups, the conditions and
education and task2 as covariates
OR
- 4 one-way anovas, one for each contrast of interest (as I did for the
multiple regression analysis), including "group" as between-groups
factor and including education and task2 as covariates.
I would really appreciate any suggestion.
Thanks!
Federica
|