Dear Abraham,
The SPM8 manual (p. 313 onwards) gives a short description of each of the BMS parameters.
We looked mostly at the exceedence probabilities (.xp) - it's actually the same data you get in the output graph after running the BMS analysis.
I think you'd find the paper "Bayesian model selection for group studies" (Stephan et al., 2009, Neuroimage) very helpful.
Hope this helps,
Talya
|