Hi,
I am trying to use the multinomial logit with endogenous selection model (Bourguignon, Fournier, and Gurgand 2007) to study the impact of agricultural technology packages on yield (similar to Teklewold et al. 2013) but running into an error as I explain below.
My data has a panel structure with multiple years and repeated households in the long format. The sampling structure is stratified by region (such as North, South, East and West) followed by districts in each region and finally households in each district.
Before running svyselmlog I run the following (since the help file says that svyselmlog should follow the svyset command): svyset strataid, strata(region) vce(linearized) singleunit(missing) || clusterid, strata(district) || hhid. I follow https://www.youtube.com/watch?v=XYjWCL7IEKU&t=74s for the syntax of svyset.
Output:
Note: Stage 1 is sampled with replacement; further stages will be ignored for variance estimation.
pweight: <none>
VCE: linearized
Single unit: missing
Strata 1: region
SU 1: strataid
FPC 1: <zero>
Next I am trying to fit the following command - svy: svyselmlog depvar varlist, select(depvar_s varlist_s) method(dmf) boot(100)
This following error is shown:
svyselmlog is not supported by svy with vce(linearized); see help svy estimation for a list of Stata estimation
commands that are supported by svy
r(322);
Am I on the right track here? Did anyone else face this error and solve it?
I am very new to the svyselmlog and svyset command and looking forward from someone who has some experience with these commands. Literature on this command seems to be very limited. I have searched this forum as well googled this error and cant really find which direction I should look. The help files of both commands also don't seem to help with this command.
Thank you so much for your time!
Srabashi Ray
I am trying to use the multinomial logit with endogenous selection model (Bourguignon, Fournier, and Gurgand 2007) to study the impact of agricultural technology packages on yield (similar to Teklewold et al. 2013) but running into an error as I explain below.
My data has a panel structure with multiple years and repeated households in the long format. The sampling structure is stratified by region (such as North, South, East and West) followed by districts in each region and finally households in each district.
Before running svyselmlog I run the following (since the help file says that svyselmlog should follow the svyset command): svyset strataid, strata(region) vce(linearized) singleunit(missing) || clusterid, strata(district) || hhid. I follow https://www.youtube.com/watch?v=XYjWCL7IEKU&t=74s for the syntax of svyset.
Output:
Note: Stage 1 is sampled with replacement; further stages will be ignored for variance estimation.
pweight: <none>
VCE: linearized
Single unit: missing
Strata 1: region
SU 1: strataid
FPC 1: <zero>
Next I am trying to fit the following command - svy: svyselmlog depvar varlist, select(depvar_s varlist_s) method(dmf) boot(100)
This following error is shown:
svyselmlog is not supported by svy with vce(linearized); see help svy estimation for a list of Stata estimation
commands that are supported by svy
r(322);
Am I on the right track here? Did anyone else face this error and solve it?
I am very new to the svyselmlog and svyset command and looking forward from someone who has some experience with these commands. Literature on this command seems to be very limited. I have searched this forum as well googled this error and cant really find which direction I should look. The help files of both commands also don't seem to help with this command.
Thank you so much for your time!
Srabashi Ray
