Inferential Statisticshard
0:00.0

When comparing two nested models using the Likelihood Ratio Test, suppose the full model has k1k_1 parameters and the reduced model has k0k_0 parameters. If the true data generation process is not contained in either model, how does the Wilks' Theorem behave?