A non-nested test of level-differenced versus log-differenced stationary models
应用模拟Cox检验程序来比较线性与对数线性模型，并扩展到水平差分与对数差分平稳模型的选择，通过蒙特卡洛实验发现简化版模拟Cox检验优于PE和BM检验，但与DL检验无显著差异。
This paper considersthe applicationof the simulated Cox test procedure developed in Pesaran and Pesaran (1993) to test linear versus log-linear models. The test procedure can also be applied to other generalized linear regression models such as level-difference stationary models versus the log-difference stationary models. In order to compare the small sample performanceof the proposed test with other tests extant in the literature, the paper also reports the resultsof a numberof Monte Carlo experiments using the experimental framework of Godfrey et al. (1988). The Monte Carlo results provide strong support for a simplified version of the simulatedCox test over the PE and the BM tests, but suggest that there is little to choose between the simulated Cox test and the DL test.