Hello,
I am trying to choose a method of estimation that better fits my model. I have one model of production function which I regressed it using OLS (reg command) and stochastic frontier (frontier command). The result from the frontier regression shows that the inefficient term is insignificant and the model can be reduced to OLS. However, the estimators from the two models are slightly different and I wonder whether which one would be better represent my model. In this case, can I use AIC and BIC (estat ic command)?
Thank you very much,
Komson
I am trying to choose a method of estimation that better fits my model. I have one model of production function which I regressed it using OLS (reg command) and stochastic frontier (frontier command). The result from the frontier regression shows that the inefficient term is insignificant and the model can be reduced to OLS. However, the estimators from the two models are slightly different and I wonder whether which one would be better represent my model. In this case, can I use AIC and BIC (estat ic command)?
Thank you very much,
Komson