“…In a univariate setting, while using the mixtures to model the MIC distribution (e.g. Jaspers et al., ), the wild‐type subpopulation of microbials was characterized by the mixture component with the lowest mean value. In a multivariate setting, with K mixture means , , it is however not always possible to identify a single component for which all marginal means , , attain their minimal value across the subpopulations.…”