Welcome. The Pooled Mean Group (PMG) estimator is suitable when heterogeneity exists in slope coefficients. This estimator constrains the long-run coefficients to be identical but allows for short-run coefficients to differ across units. It is particularly useful for estimating nonstationary heterogeneous panels, where the number of groups and the number of time series observations are relatively large. In case of slope heterogeneity see following paper for a most suitable technique. I have also explain how to deploy that technique on this channel. Hussain, Z., Mehmood, B., Khan, M. K., & Tsimisaraka, R. S. M. (2022). Green growth, green technology, and environmental health: evidence from high-GDP countries. Frontiers in Public Health, 9, 816697.
Both can be used but PMG can give biased results in presence of cross sectional dependence. Choose PMG for common slope and group-specific intercepts. Choose CS-ARDL for group-specific slopes and intercepts.
PMG is not suitable for nonlinear effects as it assumes a common slope coefficient for all groups. MG is also limited, as it assumes weak homogeneity and may not capture nonlinear effects accurately. CS-ARDL might be your best bet, since it lets you estimate group-specific coefficients and potentially nonlinear relationships. But, if you're working with a small sample size or prefer a more straightforward approach, you might want to consider other options specifically designed for panel data, like Nonlinear Least Squares (NLS), Generalized Method of Moments (GMM), or Nonlinear Panel Data Regression.
Very well explained and easy to understand. Thanks
Glad it was helpful!
Hello.Sir
I am working on panel data
When i applied pesaran yamagata test ..than Ho accepted
Further, by applying HAC ..Ho rejected
What does it mean?
which software are you using?
Hello Sir, I'm having same problem as well. By the way i'm using stata 15
It means that slope coefficients are homogenous.
I don't think that it has anything to do with the version of Stata
@@drbilalmehmood206 Thank you very much for your quick feedback!
Thanks. in your opinion What is the best linear regression model for heterogenous cross dependence panel data of N=7 and T=21 please?
CS-ARDL is contemporary and fulfills both requirements heterogeneous, cross-sectional dependence.
th-cam.com/video/NZoYw3ZeVvE/w-d-xo.html
@@drbilalmehmood206 I really appreciate answering my questions thank you so much doctor.
Thanx. Do you mean that if we accept H0, then we must use PMG in this case?
Welcome.
The Pooled Mean Group (PMG) estimator is suitable when heterogeneity exists in slope coefficients. This estimator constrains the long-run coefficients to be identical but allows for short-run coefficients to differ across units. It is particularly useful for estimating nonstationary heterogeneous panels, where the number of groups and the number of time series observations are relatively large.
In case of slope heterogeneity see following paper for a most suitable technique. I have also explain how to deploy that technique on this channel.
Hussain, Z., Mehmood, B., Khan, M. K., & Tsimisaraka, R. S. M. (2022). Green growth, green technology, and environmental health: evidence from high-GDP countries. Frontiers in Public Health, 9, 816697.
If the model heterogeneity exists, can I use PMG and cs-Ardl, Dr.? Or either one?
Both can be used but PMG can give biased results in presence of cross sectional dependence.
Choose PMG for common slope and group-specific intercepts. Choose CS-ARDL for group-specific slopes and intercepts.
@@drbilalmehmood206 Thanks Dr. If to examine the nonlinear effect, both can be used? Or need to used MG and PMG only?
PMG is not suitable for nonlinear effects as it assumes a common slope coefficient for all groups. MG is also limited, as it assumes weak homogeneity and may not capture nonlinear effects accurately.
CS-ARDL might be your best bet, since it lets you estimate group-specific coefficients and potentially nonlinear relationships. But, if you're working with a small sample size or prefer a more straightforward approach, you might want to consider other options specifically designed for panel data, like Nonlinear Least Squares (NLS), Generalized Method of Moments (GMM), or Nonlinear Panel Data Regression.