Skip to main content Skip to navigation

How Not to Analyze Your Data: A Skeptical Introduction to Modeling Methods

A half-day educational course at the 2013 Organization for Human Brain Mapping meeting, 16 June.

Organizers Victor Solo, University of New South Wales and Thomas Nichols, University of Warwick

Description While the explosive growth of neuroimaging over the last 20 years is now a commonplace, less remarked is the similar growth of neuroimaging data analysis methodology. Indeed since the beginning of the HBM conference about 20% of the posters have been on methodology demonstrating emphatically the enduring importance of methodology. Further the intense recent interest in connectivity has put pressure on the methodology to deal coherently with the complementary information supplied by different modalities such as MEG, EEG, DTI and so on. But how to do this for those who lack the expertise without handing all responsibility to the 'quants'? This course will tackle that challenge from a number of angles. But an underlying theme will be a bottoom-up approach that starts with realistic neuroimaging data and allows the issues to thereby emerge naturally. But even though the whole neuroimaging community of necessity uses methods, only a fraction are experts. Yet rigorous science requires the scientisit to be critical of all aspects of the science and this includes methodology.