Commands for testing instruments when using ' svy:ivregress'

[Mwenzie Banda]

Hello everyone,

I'm currently working with DHS data and attempting instrumental variable estimations. However, DHS recommends utilizing svyset when conducting any regression. The challenge I'm facing is that I cannot use estat endog, estat firststage, and estat overid when weights are applied with the ivregress command. Does anyone have insights on how I can test the strength of my instruments in this scenario?

 
 syv, subpop(sample1): ivregress 2sls dependent_var independent_var (endogenous_var=instrument_var)

[George Ford]

have you tried ivreg2? not sure it works with svy, but if it does it will give you the test statistics automatically.

There's some command that allows svy estimation with non-allowed commands, but I can't recall what it is.

[Mwenzie Banda]

I tried using ivreg2 but it is not supported by svy

[George Ford]

Try running the model using [pw=weightvar] and see how different the results are. Could use estat based on this model. In my experience, the results are trivially different.

And, the command is -cmdok- but it's used with mi rather than svy.

[Vanessa Owusu Ansah]

Mwenzie Banda, did you find a solution to your problem? If yes, how did you handle it?