This paper utilizes the bootstrap to construct a test using R2 for nonnested regression models. The bootstrap enables us to compute the statistical significance of the diggerences in R2's and to formally test about nonnested regression models. Bootstr...
This paper utilizes the bootstrap to construct a test using R2 for nonnested regression models. The bootstrap enables us to compute the statistical significance of the diggerences in R2's and to formally test about nonnested regression models. Bootstrapped R2 test that this paper proposes is expected to show better finite sample properties since it does not have such cumulated errors in the computation process. Moreover, bootstrapped R2 test will remoce the possibility of inconsistent test results that the previous test suffer from. Because bootstrapped R2 test only evaluates if a model has a significantly higer explanatory power Monte Carlo simulation results to compare the finite sample properties of the proposed test with the previous tests such as Cox test and J-test.