Thanks to Partha for the correction. Even my second reading of the help
was not careful enough! Do as I say, not as I do....
Nick
[email protected]
Partha Deb
Nick was correct the first time. -vce(cluster <clustervar>)- is not an
allowed option along with -twostep- . It would work if you did not
specify
-twostep- ,i.e., the -mle- option.
Nick Cox wrote:
> Yes, and sorry:
>
> You are correct in that -vce(cluster <clustervar>)- is one allowed
> syntax.
> So your second example on the face of it should work. Your first
doesn't
> satisfy.
>
> I can't explain why not. I hope somebody else can help further.
>
> Choudry T Shehzad
>
> Thanks Nick:
> In help file for ivprobit, they do ask for cluster variable but even
> without that stata says "vcetype 'cluster' not allowed"
>
> ivprobit dependent x y l.z (zz = mz pz), vce(cluster) twostep first
> vcetype 'cluster' not allowed
> r(198);
>
> ivprobit dependent x y l.z (zz = mz pz), vce(cluster country )
twostep
> first
> vcetype 'cluster' not allowed
> r(198);
>
> Nick Cox wrote:
>
>> Typically options belong together after a single comma, for example:
>>
>> ivprobit dependent x y l.z (zz = mz pz), vce(cluster) twostep first
>>
>> There are exceptions to that rule but they don't make your syntax
> legal.
>>
>> Beyond that you just have to read the help! For example, for
> -ivprobit-
>> the allowable syntax is just to specify a vcetype; nothing is said
> there
>> about a cluster variable.
>>
>> Choudry T Shehzad
>>
>> In stata help pages vce(cluster, clustervariable) and vce (bootstrap)
>> has been mentioned as options but when I try to run it says
>>
>> ivprobit dependent x y l.z (zz = mz pz), vce(cluster, country),
>> twostep, first
>> invalid 'twostep'
>> r(198);
>>
>> . ivprobit dependent x y l.z (zz = mz pz), vce(cluster, country),
> first
>> invalid 'first'
>> r(198);
>>
>> . ivprobit dependent x y l.z (zz = mz pz), vce(cluster, country)
>> option country not allowed
>> r(198);
>>
>> . ivprobit dependent x y l.z (zz = mz pz), vce(boot)
>> time-series operators are not allowed with bootstrap without panels,
> see
>> tsset
>> r(198);
>>
>> Only option it accepts is vce(robust). While I want to bootstrap
>> standard errors on clusters. I understand that ivprobit is not for
> panel
>> then why this option has been given after ivprobit. can anybody help?
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
