Do you think that the model to be used (SAC, SEM, Durbin or other) should be selected according to the specific issue and problem of study or should several techniques be applied and select the best according to the statistical tests?
Can you please explain your question? It ll help to answer correctly.
Dear Héctor Manuel Malacara Hernández
I think there are some useful papers focusing on this topic very much. Sometimes different models may get different results, and however, we cannot judge which model is better, which can be said model effects are completely unknown until you finish every data process, anyway all are based on your raw data. Statistical test is just a test, it cannot be used to judge you are right or wrong completely and sometimes your results may fail to be used to support your ideas. Seems like statistics sometimes looks like a magic. We should not believe the results until we find the substantiation following your hypotheses. By the way, I think you can also read the theory of these models or find some manuals from the Internet.
Hope this can help you.
Regards.
Can you please explain your question? It ll help to answer correctly.
Dear hector
As Kai Liu said, it will depend on your hypothesis, but also on the results of econometric tests.
Check recent literature in your area of study.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics