Global envelope tests for spatial processes. Preprint on arxiv:1307.0239

Envelope tests are a popular tool in goodness-of-fit testing in spatial statistics. These tests graphically compare an empirical test function T (r) with its simulated counterparts from the null model. However, conventionally the type I error probability α is controlled for a fixed distance r only, whereas the functions are inspected on an interval of distances I. We offer two approaches to building global envelope tests on I: (1) construction of envelopes for a deviation test, and (2) ordering the empirical and simulated functions based on their r-wise ranks among each other. These new tests allow choosing the global α a priori and provide p-values. We illustrate the tests through simulated and real point pattern data. Key words: deviation test; global envelope test; Monte Carlo p-value; goodness-of-fit test; simultaneous inference; spatial point patter