My supervisor have a question, for which I am looking for an answer, the question states that "Since the sample used in this paper is cross-country panel data, but the author's empirical research mainly adopts the event sequence method, which is not suitable. In addition, more importantly, when examining the impact of "export diversification" and "FDI inflows" on economic growth, the reverse causality is more obvious, that is, economic growth will have a reverse effect on exports and FDI, and then in the empirical research. There may be serious endogeneity problems, and the usual time series methods cannot handle such problems well, so the reliability of the empirical regression results in this paper is questionable."
how can I solve this problem without changing the model. I have done the following tests and results are favorable.
4.3.1 Descriptive Statistics.
4.3.2 Cross-sectional Dependency Test
4.3.3 Unit Root Test
4.3.4 Cointegration Test
4.3.5 Lag Length Criteria.
4.3.6 Vector Auto regression Model
4.3.7 Impulse Response Function.
4.3.7.1 Response of Export Diversification.
4.3.7.2 Response of Economic Growth.
4.3.7.3 Response of FDI Inflows.
4.3.8 Variance Decomposition
Secondly my supervisor also asked " The impulse response analysis under the PVAR model does not consider the confidence interval, and the IRF is at zero line, which indicates that the IRF is not statistically significant, however the authors did not notice this and need to make Those who have an in-depth grasp of the technical literature, use the calibration method."
I have added the IMR results, I cant figure out her question. Kindly provide me your valuable suggestions.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis