False discovery rate correction promotes confounded neuroimaging designs

False discovery rate (FDR) corrections are a popular and pragmatic approach to dealing with the problem of multiple comparisons in neuroimaging research. This paper describes a potentially problematic outcome of FDR correction. Specifically, the use of FDR will tend to promote the publication of poorly controlled experiments. This occurs because the presence of significant hypothesis tests in some parts of the brain will lower the threshold for significance elsewhere in the brain. In general, this adaptive thresholding is a feature of FDR. However, it can be problematic when a confounded design produces spurious activations, because they will lower the threshold for significance elsewhere in the brain, relative to a well-controlled study. Here I demonstrate this basic phenomenon using Monte Carlo simulations. I conduct further simulations to estimate how this phenomenon could affect more realistic confounding scenarios and how it could interact with publication bias. I conclude by suggesting mitigation approaches.