Dear SPMers,
im using gppi (@Donald: cutting-edge tool – thanks!) to futher explore my data from a GLM Model. Briefly I used an event related design with the following four conditions:
- E_re
- E_nre
- N_re
- N_nre
The second level analysis revealed between-group differences (patients, controls) in the insula for the contrast E_re > E_nre, which I now want to test for between-group differences in connectivity. My main issue is the design of the ppi-contrast and its implementation in gppi.
Running the gppi-script (method = cond / analysis = psy / all 4 task conditions & movement regressors included) without estimating the contrasts works fine and results in a design matrix with the conditions:
E_re / E_nre / N_re / N_nre / PPI_E_re / PPI_Enre / PPI_N_re / PPI_N_nre / constant / 6xmov_parameters
However, setting up the contrasts results in error messages such as ‘P.Contrasts(1).left is not formatted correctly.’
My questions:
1. P.Contrasts(1/2) Left / Right
What does the 1 or 2 and the left or right stand for regarding the setup of the contrasts?
2. How to implement the contrast for the PPI regarding the above mentioned question (comparing the groups regarding differences in
insula-connectivity under the conditions of E_re>E_nre)?
Thanks in advance & best regards
Ben
|