Dear Statalist
I am working on spill-over effects, my objective is (Does innovation level in country 1 has an impact on innovation levels and corruption levels of the neighbouring countries?) or (does the corruption level in country 1 has an impact on the corruption levels and innovations level of the neighbouring countries
I want to measure the spill-over effects of innovation and corruption of the countries.
My plan is to use xsmle (Spatial Panel Data Models), which requires spatial weight matrix to run the regression. I want to create the weight matrix based on the distance using the coordination’s of the capital cities of the countries.
I am using unbalanced panel dataset of 100 countries for 15 years. So I’ve imputed the dataset because xsmle command can work only with strongly balanced pane dataset.
The multiple imputation commands are:
describe
misstable summarize
mi set wide
mi register imputed y x1 x2 x3
mi impute mvn y x1 x2 x3 = x4 x5 x6, add(50) reseed(12345)
***declare that the dataset is panel
encode country, generate(kntry)
mi xtset kntry yr
the regression
mi estimate: xsmle y x1 x2 x3 x4 x5 x6, fe model(sdm) wmatrix (...)
The problem I am facing is that I don’t know how to create the spatial weight matrix for my dataset
Any suggestions how to calculate the weight matrix
regards,
Maryam
I am working on spill-over effects, my objective is (Does innovation level in country 1 has an impact on innovation levels and corruption levels of the neighbouring countries?) or (does the corruption level in country 1 has an impact on the corruption levels and innovations level of the neighbouring countries
I want to measure the spill-over effects of innovation and corruption of the countries.
My plan is to use xsmle (Spatial Panel Data Models), which requires spatial weight matrix to run the regression. I want to create the weight matrix based on the distance using the coordination’s of the capital cities of the countries.
I am using unbalanced panel dataset of 100 countries for 15 years. So I’ve imputed the dataset because xsmle command can work only with strongly balanced pane dataset.
The multiple imputation commands are:
describe
misstable summarize
mi set wide
mi register imputed y x1 x2 x3
mi impute mvn y x1 x2 x3 = x4 x5 x6, add(50) reseed(12345)
***declare that the dataset is panel
encode country, generate(kntry)
mi xtset kntry yr
the regression
mi estimate: xsmle y x1 x2 x3 x4 x5 x6, fe model(sdm) wmatrix (...)
The problem I am facing is that I don’t know how to create the spatial weight matrix for my dataset
Any suggestions how to calculate the weight matrix
regards,
Maryam
Comment