Poppe, Andrew Bogdan2011-07-142011-07-142011-05https://hdl.handle.net/11299/109199University of Minnesota M.A. thesis. May 2011. Major: Psychology. Advisor: Angus W. MacDonald III, Ph.D.,1 computer file (PDF); v, 34 pages.Typically, fMRI analyses of schizophrenia patients’ family members have employed the general linear model (GLM) to examine clusters of activation in the brain, which provides a region-by-region examination of brain activity. This study additionally examined a newer group independent component analysis (ICA) to test the hypothesis that the unexpressed genetic liability to schizophrenia is reflected in the functional connectivity between brain regions during a context processing task, the expectancy AX task. We compared 20 schizophrenia patients and 32 first-degree relatives to 22 controls and 28 control relatives. The subjects completed the expectancy AX task, a context processing measure, while being scanned in a 1.5T MR scanner. We then performed a group ICA on all participants’ fMRI data in order to examine the functional networks that are active during the AX task. Next, a GLM analysis was performed. Groups’ mean activations were contrasted with each other to obtain differential activation. The group ICA showed significantly different activations between patient probands and control probands in a network constituting dorsolateral prefrontal cortex and posterior parietal lobe. The relative groups differed in an anterior cingulate network. The GLM analysis showed differential functioning between patient and control relatives in the dorsolateral prefrontal cortex, but failed to show any differences between patient and control probands. These disparate findings suggest some potential advantages to functional connectivity relative to region-by-region approaches to understanding the neural basis of genetic liability to schizophrenia.en-USPsychologyThesis or Dissertation