New Working Paper (Short Note)
David Rügamer, Sonja Greven
Statistical inference after model selection requires an inference framework that takes the selection into account in order to be valid. Following recent work on selective inference, we address the issue of conducting valid inference after likelihood- or test-based model selection, which comprises (iterative) selection based on the commonly used model selection criteria AIC and BIC, likelihood-ratio test or F-test-based selection and p-value based model selection via t-tests. We derive an analytical solution for inference after likelihood- or test-based model selection for linear models, representing the selection event as affine inequalities, and make available an R software package for selective inference in such settings.