Skip to main content
. Author manuscript; available in PMC: 2017 Dec 8.
Published in final edited form as: Psychometrika. 2016 Oct 13;82(4):1052–1077. doi: 10.1007/s11336-016-9527-8

Figure 2.

Figure 2

Model-implied coverage probability as a function of ñ, where “coverage probability” refers to the probability that an analytic p value yields consistent results with a permutation test based on a 99%confidence interval for the permutation p value. The darker line corresponds to the model fit to data generated such that p < 0.05, and the lighter line corresponds to the model fit to data generated freely under H0.