Principles of random effect analyses
These analyses are also called second level analyses, and concentrate on the variation of activation between subjects in deciding the significance levels for SPMs.
For a comprehensive description of this problem in SPM, see the [http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/pdfs/Ch12.pdf random effects chapter] of the [http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/ Human Brain Function book] - second edition.
See also the [http://www.fil.ion.ucl.ac.uk/spm/doc/biblio/Keyword/RFX.html RFX section] of the FIL/SPM bibliography.
There is a nice [http://psyphz.psych.wisc.edu/~oakes/spm/spm_random_effects.html random effects discussion on the Wisconsin site]
A useful set of [http://www.cu.mrc.ac.uk/~jaston/mrcjc/sld001.htm powerpoint slides on random effects], by John Aston, at the MRC Cyclotron Unit.
Jesper Andersson's nice talk on [http://www.fmrib.ox.ac.uk/fslcourse/lectures/feat3/MixedEffects.ppt Mixed effects models] and variance components.