TRANSPARENCY AND REPRODUCIBILITY IN THE NEUROIMAGING OF AGING

In concert with broader efforts to increase the reliability of social science research, there are several efforts to increase transparency and reproducibility in neuroimaging. The large-scale nature of neuroimaging data and constantly evolving analysis tools can make transparency challenging. I will describe emerging tools used to document, organize, and share behavioral and neuroimaging data. These tools include: (1) the preregistration of neuroimaging data sets which increases openness and protects researchers from suspicions of p-hacking, (2) the conversion of neuroimaging data into a standardized format (Brain Imaging Data Structure: BIDS) that enables standardized scripts to process and share neuroimaging data, and (3) the sharing of final neuroimaging results on Neurovault which allows the community to do rapid meta-analysis. Using these tools improves workflows within labs, improves the overall quality of our science and provides a potential model for other disciplines using large-scale data.