|
Abstract:
|
The success of any scientific field can be judged by the cumulative understanding it produces. Neuroimaging research has adopted several advantageous standards in this respect, including (a) the reporting of standardized brain coordinates, and (b) a movement towards creating publically accessible, large-scale datasets. In this talk, I first outline problems of power and replicability in neuroimaging datasets. Then, I describe current meta-analytic strategies and how they can yield dramatic benefits in power and the ability to specify precise a priori hypotheses in new studies. Finally, I discuss some ways in which current meta-analytic methods fall short, and point out opportunities for further quantitative development.
|
