The causes of non-reproducible results are numerous, but the assessment of these issues is often difficult. In this talk, I will review the natural causes of non-reproducibility and non-re-usability, specifically using brain-imaging research examples to illustrate these problems. I will then lay out the most promising emerging solutions, including those that relate to data science, such as the use of semantic web technologies. I will conclude by considering what seem to be necessary cultural changes to improve the current situation in the long term.
Speaker(s)
JB Poline
I have worked on the development of methods for the analysis of functional imaging data (mostly fMRI), specifically related to statistical modeling and inference. I maintain close interactions with neuroscientists to ensure that analysis methods are answering actual needs, and I am involved in the educational component of the organization for Human Brain Mapping. During the past five years, I was responsible for a large multi-centric neuroimaging genetic database. I realize that neuroinformatics is a fundamental part of neuroimaging, and I work with the International Neuroinformatics Coordinating Facility to improve data sharing and reproducibility in brain imaging. I am the co-editor of Frontiers in Brain Imaging Methods and am interested in experimenting with new ways to conduct and publish research.