The coordination of design and analysis techniques for functional magnetic resonance imaging data

Abstract

Statisticians and research scientists have significant, but often separate, roles in the process of conducting an experiment. In a functional magnetic resonance imaging (fMRI) study, specific scientific questions are formulated, a design is chosen, and the data are analyzed accordingly. Functional MRI technology is unique in that a large, multi-dimensional dataset is produced, the design works under the constraints of physical principles and the type of analysis desired, and the range of analyses is both relatively new and constantly changing. Communication and collaboration regarding these relationships and concerns are vital to a successful study. This master's thesis will address the issues which arise throughout the course of designing and analyzing an fMRI study. Different design and analysis techniques, real-world examples of each, and the consequences of the decisions made will be explored at all stages of the process.