Dear experts,
I would like your help correlating questionnaire scores with BOLD activity in a 2x2 ANOVA model in SPM12. I can't figure out how to set up my contrasts.
I'm analyzing an fMRI attention task with 4 conditions (ExtPos, ExtNeg, IntPos, IntNeg; shown in the 2x2 table below). I have 46 subjects and one questionnaire score of inattention for each participant.
External Attention: Internal Attention:
Positive Valence : ExtPos IntPos
Negative Valence: ExtNeg IntNeg
I wanted to assess if BOLD response during the individual task conditions correlate with inattention at the inter-individual level. Originally, I fed participants' first-level contrast into a 1-sample t-test with their questionnaire score as a covariate of interest. I made four models, one for each of the four task conditions (IntPos, IntNeg, ExtPos and ExtNeg). I used a Bonferroni adjusted p-value to determine significance. This analysis worked but I don't know if it is valid, as my reviewers have advised I make do 2x2 ANOVA instead, with the inattention score as a covariate.
So, now I've made an ANOVA model (within subject) in SPM12 and put the inattention scores as a covariate. My design matrix has 4 columns for the 4 task conditions (ExtPos, ExtNeg, IntPos, IntNeg), followed by one column for the covariate (inattention), followed by 46 columns (one for each participant).
I don't know how to do the contrasts. For example, if if I want to see if there are any significant correlations between participants' whole-brain activity and inattention scores during the internal positive (IntPos) condition, how would I set up my 0s and 1s?
Thank you very much in advance for your help!
Halima
|