If you are unfamiliar with SPM, there are a number of places to begin. The early history of statistical parametric mapping can be found here and an introduction to the basic components of SPM can be found here. This brief overview describes the sorts of neuroimaging experimental designs that can be analyzed using SPM, but does not discuss its capabilities in analyzing data from EEG and MEG sources.
 
More detailed treatments can be found in Human Brain Function 2nd Edition, an on-line resource with chapters providing thorough coverage of many of the workshop topics, including computational neuroanatomy, statistical modeling, and inference.
 
An extensive indexed bibliography of SPM methods papers can be found here. This collection is the best source for detailed accounts of the theoretical background to the various SPM components.