Abstract

Cognitive Neuroscience is a highly interdisciplinary field – this brings challenges with respect to teaching relevant skills, and communication among researchers from different backgrounds. Neuroimaging data analysis is complex: There are different software packages, different methods, and many parameter choices. You want to know about the limitations of your methods before the reviewers point them out to you. Recently reported “replication crises” may be partly due to methodological limitations. While standardised analysis pipelines exist, many experiments are not “standard”, e.g. when using the latest methods. Furthermore, even non-neuroscientists want to understand neuroscientific literature – this requires at least basic knowledge of the methods. Does this mean we have a problem? I’ve heard many different opinions about the amount and level of methods training we should provide. Evidence is mostly anecdotal. Here, I’m trying to estimate the methods-skills level among over 300 students and post-docs in an on-line survey, and determine some factors that affect it.