Martin wrote:
> Look at the last page of
> http://www.tulane.edu/~economic/RePEc/pdf/tul0901.pdf for the returned
> results...
Ok, -ivtest- seems to return the results as r() scalars and macros. To
add, say, the CLR statistic and its p-value, type
. use http://fmwww.bc.edu/ec-p/data/hayashi/griliches76.dta
. eststo q1: ivreg2 lw s expr tenure rns smsa (iq=med kww age mrt), robust
. ivtest, ci
. estadd scalar r(clr_p)
. estadd scalar r(clr_stat)
. eststo q2: ivreg2 lw s expr tenure rns smsa age mrt (iq=med kww), robust
. ivtest, ci
. estadd scalar r(clr_p)
. estadd scalar r(clr_stat)
. estout q1 q2, cells(b(star fmt(3)) se(par) ) stats(r2_a N widstat jp
clr_p clr_stat, fmt(3))
The confidence sets seem to be returned as string macros. To display
them in -estout- type, e.g.,
.estadd local clr_cset = r(clr_cset)
and then
. estout q1 q2, ... stats(... clr_cset)
ben
> On Tue, Feb 10, 2009 at 12:25 AM, Nirina F <[email protected]> wrote:
>> Hello,
>> I am not very familiar with estout.
>> My questions are:
>> 1-how do I put the matrix results from my ivtest below after my
> regressions?
>> 2-When I use append, the second equation result is reported under the
>> first equation. I saw that I could do estout eq1 eq2 but that means I
>> have to also name the matrices from ivtest after eq2 differently to be
>> put under eq2?
>>
>> . use http://fmwww.bc.edu/ec-p/data/hayashi/griliches76.dta
>>
>> . eststo q1: ivreg2 lw s expr tenure rns smsa (iq=med kww age mrt),
> robust
>>
>> . ivtest, ci
>>
>> Weak instrument robust tests and confidence sets for linear IV
>> H0: beta[lw:iq] = 0
>>
>>
> ----------------------------------------------------------------------------
> -----------------------------------------
>> Test | Statistic p-value
>> 95% Confidence Set
>>
> ------+---------------------------------------------------------------------
> -----------------------------------------
>> CLR | stat(.) = 57.03 Prob > stat = 0.0000
>> [-.257136,-.039481]
>> AR | chi2(4) = 91.35 Prob > chi2 = 0.0000
>> null set
>> LM | chi2(1) = 18.20 Prob > chi2 = 0.0000
>> [-.4880471,-.0349708] U [ .0093415, .0202192]
>> J | chi2(3) = 73.15 Prob > chi2 = 0.0000
>> LM-J | H0 rejected at 5% level
>>
> ------+---------------------------------------------------------------------
> -----------------------------------------
>> Wald | chi2(1) = 4.90 Prob > chi2 = 0.0269
>> [-.021772,-.001322]
>>
> ----------------------------------------------------------------------------
> -----------------------------------------
>> Note: Wald test not robust to weak instruments. LM-J confidence set
>> not available with closed-form estimation (use
>> usegrid option).
>>
>> estout using ivequ.xls, cells(b(star fmt(3)) se(par) ) stats(r2_a N
>> widstat jp, fmt(3))
>>
>> eststo q2: ivreg2 lw s expr tenure rns smsa age mrt (iq=med kww), robust
>>
>> ivtest, ci
>>
>>
>> Weak instrument robust tests and confidence sets for linear IV
>> H0: beta[lw:iq] = 0
>>
>>
> ----------------------------------------------------------------------------
> -----------------------------------------
>> Test | Statistic p-value
>> 95% Confidence Set
>>
> ------+---------------------------------------------------------------------
> -----------------------------------------
>> CLR | stat(.) = 0.03 Prob > stat = 0.8628
>> [-.009783, .010634]
>> AR | chi2(2) = 2.62 Prob > chi2 = 0.2694
>> [-.0089519, .0099214]
>> LM | chi2(1) = 0.03 Prob > chi2 = 0.8673
>> [-.0100928, .0108969] U [ .1412529, 1.348119]
>> J | chi2(1) = 2.60 Prob > chi2 = 0.1072
>> LM-J | H0 not rejected at 5% level
>>
> ------+---------------------------------------------------------------------
> -----------------------------------------
>> Wald | chi2(1) = 0.05 Prob > chi2 = 0.8205
>> [-.008054, .010162]
>>
> ----------------------------------------------------------------------------
> -----------------------------------------
>> Note: Wald test not robust to weak instruments. LM-J confidence set
>> not available with closed-form estimation (use
>> usegrid option).
>>
>>
>> estout using ivequ.xls, append cells(b(star fmt(3)) se(par) )
>> stats(r2_a N widstat jp, fmt(3))
>>
>>
>>
>> Thank you,
>> Nirina
*
* 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/