A method for a genuinely multivariate analysis of statistical phase synchronization phenomena in empirical data is presented. It is applied to EEG data from a psychological experiment, obtaining results which indicate a possible relevance of this method in the context of cognitive science as well as in other fields.