Statistical model and estimation method, inference type (mixed/random effects or fixed), e.g., “Mixed effects inference with one sample t-test on summary statistic” (SPM2/SPM5), e.g., “Mixed effects inference with Bayesian 2-level model with fast approximation to posterior probability of activation.” (FSL) |
|
If fixed effects inference used, justify |
|
If more than 2-levels, describe the levels and assumptions of the model (e.g., are variances assumed equal between groups) |
|
Repeated measures? |
|
If multiple measurements per subject, list method to account for within subject correlation, exact assumptions made about correlation/variance |
|
e.g., SPM: “Within-subject correlation estimated at F-significant voxels (P <0.001), then used globally over whole brain”; or, if variances for each measure are allowed to vary, “Within-subject correlation and relative variance estimated…” |
|