RE: st: RE: outreg2 combined with xtivreg2

["Giacomo PALLARO" <giacomo.pallaro@studenti.unipd.it>]

Dear Mr. Schaffer,
thank you for your reply and forgive me if I was not clear in my question. 
Continuing your example: 

ivreg2 n (ys = w k), first savefirst 

OLS estimation 
-------------- 
------------------------------------------------------------------------------ 
ys | Coef. Std. Err. t P>|t| [95% Conf. Interval] 
-------------+---------------------------------------------------------------- 
w | .0069317 .0111403 0.62 0.534 -.0149286 .028792 
k | .0038175 .0019351 1.97 0.049 .0000203 .0076147 
_cons | 4.617914 .0351991 131.19 0.000 4.548844 4.686985 
------------------------------------------------------------------------------ 

let's run now 

outreg2 [_ivreg2_ys] using prova, replace 

we get exactly the first stage results 

(1) 
_ivreg2_ys 
VARIABLES	ys 

w	0.00693 
(0.0111) 
k	0.00382** 
(0.00194) 
Constant	4.618*** 
(0.0352) 

Observations	1,031 
R-squared	0.004 
Standard errors in parentheses	
*** p<0.01, ** p<0.05, * p<0.1	



My problem is not here, it is on the use of xtivreg2. 

xtivreg2 n (ys = w k), first savefirst fe 

------------------------------------------------------------------------------ 
ys | Coef. Std. Err. t P>|t| [95% Conf. Interval] 
-------------+---------------------------------------------------------------- 
w | -.106388 .0311445 -3.42 0.001 -.1675134 -.0452626 
k | .1702418 .0119891 14.20 0.000 .1467115 .1937722 
------------------------------------------------------------------------------ 


. outreg2 [_xtivreg2_ys] using prova2, replace 
variable _est__xtivreg2_ys not found 
r(111); 


This problem still remain even after est restore or est replay. 
Stata tells me to store results in _xtivreg2_ys, but outreg2 doesn't access to 
them since, as you noted, the _est_ variable is not generated. 

How can I get, using outreg2 after xtivreg2, similar results with respect to those produced using ivreg2? 

Thank you for your attention, 
Best regards
Giacomo 

Università degli Studi di Padova 




-----Original Message----- 
From: Schaffer, Mark E 
Sent: 27 August 2013 22:02 
To: 'statalist@hsphsun2.harvard.edu' 
Subject: RE: outreg2 combined with xtivreg2 

Giacomo, 

I'm not sure exactly what the problem is. 

I tried using the savefirst option with both ivreg2 and xtivreg2. In both cases, the first-stage regressions were saved and I could replay them or restore them. 

For example, using the abdata set, after 

webuse abdata, clear 
ivreg2 n (ys = w k), savefirst 
xtivreg2 n (ys = w k), savefirst fe 

when I say 

est dir 

I get 

------------------------------------------------------------------------- 
name | command depvar npar title 
-------------+---------------------------------------------------------- 
-------------+- 
_ivreg2_ys | ivreg2 ys 3 First-stage regression: ys 
_xtivreg2_ys | xtivreg2 ys 2 
------------------------------------------------------------------------- 

I can use est replay or est restore with either of these saved results. So I don't see what the problem is. 

I do note the discrepancy in terms of variables in memory. As you note, the saved first-stage estimation with ivreg2 creates an esample() variable: 

storage display value 
variable name type format label variable label 
------------------------------------------------------------------------------------------------------------ 
_est__ivreg2_ys byte %8.0g esample() from estimates store 

And savefirst with xtivreg2 does not create a corresponding variable. 

I would need to check, but this is probably because xtivreg2 uses preserve/restore and newly created variables get lost when the original dataset is restore-d at the end. 

But I don't see what this has to do with "exporting first stage results". 

--Mark 

> -----Original Message----- 
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner- 
> statalist@hsphsun2.harvard.edu] On Behalf Of Giacomo PALLARO 
> Sent: 22 August 2013 14:52 
> To: statalist@hsphsun2.harvard.edu 
> Subject: st: outreg2 combined with xtivreg2 
> 
> Dear all, 
> I'm running on Stata 12 some IV regressions using xtivreg2 and I would 
> like to print the results using outreg2. 
> The help file of xtivreg2 sends you to the help section of ivreg2, 
> where you can 
> find: 
> 
> The savefirst option requests that the individual first-stage 
> regressions be saved for later access using the estimates command. If 
> saved, they can also be displayed using first or ffirst and the ivreg2 
> replay syntax. The regressions are saved with the prefix "_ivreg2_", 
> unless the user specifies an alternative prefix with the savefprefix(prefix) option. 
> 
> Using ivreg2 with first savefirst options, I noticed that Stata 
> generates a variable _est__ivreg2_varname. Then using the command " outreg2 
> [_ivreg2_*] " I'm able to export the first stage results correctly. 
> Unfortunately with xtivreg2 this doesn't occur. In particular no 
> _est__xtivreg2_varname is generated, despite I get (similarly to 
> ivreg2 case) this expression just before the first stage output of 
> xtivreg2 
> 
> Stored estimation results 
> -------------------------------------------------------------------------------- 
> name | command depvar npar title 
> -------------------+-------------------------------------------------- 
> -------------------+--------- 
> _xtivreg2_~d | xtivreg2 L.notemplo~d 7 
> ---------------------------------------------------------------------- 
> ---------- 
> 
> Am I wrong somewhere? Is it possible to export first stage results of 
> xtivreg2 in this way? Are there other ways to do it using outreg2? 
> 
> Thank you all in advance, 
> Giacomo 



----- 
Sunday Times Scottish University of the Year 2011-2013 
Top in the UK for student experience 
Fourth university in the UK and top in Scotland (National Student Survey 2012) 

We invite research leaders and ambitious early career researchers to 
join us in leading and driving research in key inter-disciplinary themes. 
Please see www.hw.ac.uk/researchleaders for further information and how 
to apply. 

Heriot-Watt University is a Scottish charity 
registered under charity number SC000278.