Some recent studies have pointed out the heterogeneity of functional MRI data analysis pipelines (Carp, 2012a). These studies quantify the combinatorial explosion of analysis options when many different steps are combined as required when analyzing neuroimaging data. Although they focused on fMRI, the same issue arises for M/EEG. We argue that this fact does not need to become a significant drawback, as long as the details required to make the analysis reproducible are available. A difficulty does arise in that whatever level of detail is provided in a methods section of a paper, it is ultimately unlikely to be sufficient to provide access to all parameters used. However, sharing the proper code provides a better guarantee for reproducible science. The previously mentioned studies also raise the issue that the geographical location of the investigators biases their choice in terms of method and software. Again, this is not wrong per se, as expertise is more easily found from colleagues than mailing lists or user documentation. By favoring on-line collaborative work between international institutions, MNE-Python aims to reduce this geographical bias.
