Dear SPM Experts,
I hope this email finds you well. I am reaching out for assistance with setting up flexible factorial designs.
I am currently working on a task design which involves listening to auditory stimuli. The participants were required to categorize the sounds into pleasant (P), neutral (N), and unpleasant (U).
For the first-level analysis, I computed three contrasts per subject: U>N, P>N, and U+P>N. These contrasts were chosen to model the effect of affective sounds compared to neutral ones.
Now, I am looking to investigate how brain activity in response to affective sounds varies across different groups (4 levels). I intend to employ a flexible factorial model, incorporating subjects as random effects and groups as a fixed effect.
I have a few questions regarding this:
Are the three first-level contrasts I computed sufficient for second-level analysis? Or do I need to compute contrasts representing the effects of pleasant alone (1 0 0), neutral alone (0 1 0), and unpleasant alone (0 0 1) to be used in the second level?
My primary interest lies in examining the main effect of group and the interaction effect of group (4 levels) * condition (3 levels: P, U, N). Which first level contrasts should be included in the factor condition? I am concerned because if I input contrasts like U>N and P>N, neutral is represented twice. Is this a problem?
Could you please direct me to any resources that guide through the implementation of flexible factorial ANOVA in SPM, including setting up contrasts and the subject factor?
Thank you very much for your time and assistance.
I look forward to your guidance.
Best,
_____________________________
Namitha Jain
PhD student, Speech and Hearing sciences
University of Illinois, Urbana-Champaign
|