Extended Data Fig. 4: Model diagnostics from the best-fitting model.
From: Location-level processes drive the establishment of alien bird populations worldwide
a, Plot of out-of-sample CPO scores for all data points in rank order used in the model. b, Probability density of the CPO scores. c, Map of CPO scores. CPO is the probability of generating each data point in the dataset from a posterior fitted without this data point. Each panel allows visualization of where in the data the model does not fit well. All plots from a single Bayesian regression of global establishment success of avian introductions (n = 1,530 introductions).
