Lately, the scientific world has been inundated with failures to replicate and neuroimaging is likely to be affected by the same problems. This crush of false positive results is worrying: increasingly evidence suggests that false positives are proliferating due to unchecked researcher biases, which favour analysing data until a publishable positive result is obtained. No researcher wants to make spurious discoveries, but researchers do not know how to change their practices to prevent them. The goal of this course is to present practical solutions that have been developed, allowing any researchers (not just programmers) to conduct power analyses, analyse data and publish results in a reproducible manner.