A Bayesian approach to phases for frequency-tagged EEG for the cognitive neuroscience of language

For phase data from frequency-tagged EEG data Bayesian modelling gives results that are clean, it is very data efficient when effects are real and less likely to fool you with noise that looks like an effect but isn't! It allows you to arrange the model around the experiment and describes the data in terms of a clear model and posterior probabilities instead of the often confusing picture presented by hypothesis testing and frequentist statistics.