From my understanding qdec fits a glm to each vertex. It also smooths the
data.
From the documentation I do not have a clear understanding of whether qdec
uses smoothed data for input (It appears that is does) or smooths the coefficients or t-tests after analysis, or both.
Additionally, the resulting contrast results appear to have patches of significance. What is the processes that makes patches out of individual vertex results? Do a certain number of contiguous vertexes need to be significant for their values to be connected into a patch?