Sure what I mean is if I want to look at the correlation of activation (BOLD) from one trila to the next, throughout the course of a run, I can extract a ceta weight from each trial and then look at the variability from trial to trial.
Page 754, top left from (http://rissmanlab.psych.ucla.edu/rissmanlab/Publications_files/Rissman_NeuroImage_2004.pdf)
The classic univariate method for estimating activity during the individual stages of a delayed recognition task uses a single cue,
delay, and probe covariate to model the data obtained from many experimental trials, essentially treating any trial-to-trial variability
as noise. In contrast, the multivariate method we propose for modeling functional connectivity during the temporally adjacent
stages of a multistage cognitive task, although relying on the same principles used to model stage-specific univariate activity, capital-
izes on this trial-to-trial variability and uses it to characterize dynamic inter-regional interactions. The premise of this method is
that if two areas of the brain are functionally interacting with each other during a particular stage of a cognitive task, then the amount
of activity that the two areas exhibit during that stage should be correlated across trials. The goal is to obtain a reasonable
measurement of the magnitude of stage-specific activity that each voxel exhibits on each of many task trials, and then search for other
voxels in the brain that show correlated fluctuations across trials. This is accomplished by constructing a GLM in which every stage
of every trial is modeled with a separate covariate, so that trial-to- trial parameter estimates of stage-specific activity can be obtained.
These parameter estimates (beta values) can then be sorted according to the stage from which they were derived (what we
will refer to as a beta series) and correlated across regions to obtain a measure of functional connectivity (beta series correlation)
during each of the individual task components. Validation of the beta series correlation method Before illustrating how this method of obtaining