The preparation of empirical analysis includes the methods and techniques of statistics and econometrics, software operation, theoretical basis and data, as well as trying to figure out and directly participate. Find an empirical paper that interests you, and make a decomposition study on the content of empirical analysis to understand how the author deals with some problems.
Empirical analysis should focus on testing hypotheses, and the steps are as follows:
1. Prepare the test hypothesis through descriptive statistics and diagnostic tests.
2. Test the hypothesis by benchmark regression.
3. Deal with the "protection" hypothesis through robustness test and related measurement problems.
4. Through further discussion, test the hypothesis in a wider range. If beginners want to write empirical analysis well, they need to catch up with the methods, skills and operations that they should have mastered but have not learned. At the same time, they need to learn, ponder and train with the guidance and help of teachers to form a systematic cognition of empirical analysis.
If the verification results are inconsistent with the theory, the possible reasons should be analyzed:
1, fact does not correspond to theory, and theory should not explain this phenomenon.
2, the theory is not correct, as long as a counterexample can deny a theory (falsified by facts), the theory without counterexample is considered as a temporarily acceptable assumption.
3. The background of this theory is inconsistent with the current reality in China. We should analyze the inconsistencies and then improve the theory or put forward policies to change the reality.