# Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous?

6 messages
Open this post in threaded view
|
Report Content as Inappropriate

## Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous?

 Hi Cam, thank you for your help and valuable reference. In fact, the endogenous variable is continuous in my example. So -ivreg2- is fit for it. My doubt is whether I can use -treatreg- in this way. Concretely, my outcome equation is , y=a0+a1x1+a2x2+e where x1 is continuous and endogenous. Base on the median of x1, I could get x1dummy=(x1>r(p50)).Then put x1dummy as the dependent variable in the treat equation to calculate lambda. And at last, reg y x1 x2 lambda. Is is appropriate? thank you for any help! Best regards, Rose. ----- Original Message ----- From: Cameron McIntosh <[hidden email]> To: STATA LIST <[hidden email]> Subject: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? Date: 2009-8-12 03:24:14 Hi Rose, Maybe look into cmp, which deals with mixed mode endogenous variables in simultaneous equation models: http://ideas.repec.org/c/boc/bocode/s456882.htmlhttp://www.stata.com/statalist/archive/2007-10/msg00646.htmlCam ---------------------------------------- > Date: Wed, 12 Aug 2009 01:35:24 +0800 > From: [hidden email] > To: [hidden email] > Subject: st: Can I apply -treatreg- if the endogenous variable is continuous? > > I know "A treatment-effects model considers the effect of an endogenously chosen binary treatment on another endogenous continuous variable,conditional on two sets of independent variables". > > My doubt is as followings, > > If the endogenous variable is continuous, can I apply treatreg? > Namely, based on mean or median or something else, change the endogenous continuous variable > as a binary variable in the treat equation to calculate lambda, and then regress the outcome\main equation with the endogenous continuous variable and lambda as independents variables. > > Is this application appropriate? > > Thank you for any help! > > Best regards, > Rose. > > * > * 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/_________________________________________________________________ More storage. Better anti-spam and antivirus protection. Hotmail makes it simple. http://go.microsoft.com/?linkid=9671357* * 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/* *   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/
Open this post in threaded view
|
Report Content as Inappropriate

## AW: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous?

 <> Note Austin`s newly published talk in DC in July which should be a valuable resource for your reserach: http://www.stata.com/meeting/dcconf09/dc09_nichols.pdfHTH Martin -----Ursprüngliche Nachricht----- Von: [hidden email] [mailto:[hidden email]] Im Auftrag von [hidden email] Gesendet: Mittwoch, 12. August 2009 13:55 An: statalist Betreff: Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? Hi Cam, thank you for your help and valuable reference. In fact, the endogenous variable is continuous in my example. So -ivreg2- is fit for it. My doubt is whether I can use -treatreg- in this way. Concretely, my outcome equation is , y=a0+a1x1+a2x2+e where x1 is continuous and endogenous. Base on the median of x1, I could get x1dummy=(x1>r(p50)).Then put x1dummy as the dependent variable in the treat equation to calculate lambda. And at last, reg y x1 x2 lambda. Is is appropriate? thank you for any help! Best regards, Rose. ----- Original Message ----- From: Cameron McIntosh <[hidden email]> To: STATA LIST <[hidden email]> Subject: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? Date: 2009-8-12 03:24:14 Hi Rose, Maybe look into cmp, which deals with mixed mode endogenous variables in simultaneous equation models: http://ideas.repec.org/c/boc/bocode/s456882.htmlhttp://www.stata.com/statalist/archive/2007-10/msg00646.htmlCam ---------------------------------------- > Date: Wed, 12 Aug 2009 01:35:24 +0800 > From: [hidden email] > To: [hidden email] > Subject: st: Can I apply -treatreg- if the endogenous variable is continuous? > > I know "A treatment-effects model considers the effect of an endogenously chosen binary treatment on another endogenous continuous variable,conditional on two sets of independent variables". > > My doubt is as followings, > > If the endogenous variable is continuous, can I apply treatreg? > Namely, based on mean or median or something else, change the endogenous continuous variable > as a binary variable in the treat equation to calculate lambda, and then regress the outcome\main equation with the endogenous continuous variable and lambda as independents variables. > > Is this application appropriate? > > Thank you for any help! > > Best regards, > Rose. > > * > * 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/_________________________________________________________________ More storage. Better anti-spam and antivirus protection. Hotmail makes it simple. http://go.microsoft.com/?linkid=9671357* * 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/* *   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/* *   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/
Open this post in threaded view
|
Report Content as Inappropriate

## Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous?

 In reply to this post by gjhxmu No, that does not sound appropriate.  Why would you want to do that? Is it because you have no candidates for excluded instruments and you are hoping to identify off nonlinearities alone?  In that case, you would be better off in general ignoring the endogeneity problem than adopting such a strategy.  What you should do is find excluded instruments and run -ivreg2- or -ivregress-. 2009/8/12  <[hidden email]>: > Hi Cam, > thank you for your help and valuable reference. > > In fact, the endogenous variable is continuous in my example. > So -ivreg2- is fit for it. My doubt is whether I can use -treatreg- in this way. > Concretely, my outcome equation is , > > y=a0+a1x1+a2x2+e > > where x1 is continuous and endogenous. > > Base on the median of x1, I could get x1dummy=(x1>r(p50)).Then put x1dummy as the > dependent variable in the treat equation to calculate lambda. > And at last, reg y x1 x2 lambda. > > Is is appropriate? > > thank you for any help! > > Best regards, > Rose. > > ----- Original Message ----- > From: Cameron McIntosh <[hidden email]> > To: STATA LIST <[hidden email]> > Subject: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? > Date: 2009-8-12 03:24:14 > > Hi Rose, > > Maybe look into cmp, which deals with mixed mode endogenous variables in simultaneous equation models: > > http://ideas.repec.org/c/boc/bocode/s456882.html> http://www.stata.com/statalist/archive/2007-10/msg00646.html> > Cam > > ---------------------------------------- >> Date: Wed, 12 Aug 2009 01:35:24 +0800 >> From: [hidden email] >> To: [hidden email] >> Subject: st: Can I apply -treatreg- if the endogenous variable is continuous? >> >> I know "A treatment-effects model considers the effect of an endogenously chosen binary treatment on another endogenous continuous variable,conditional on two sets of independent variables". >> >> My doubt is as followings, >> >> If the endogenous variable is continuous, can I apply treatreg? >> Namely, based on mean or median or something else, change the endogenous continuous variable >> as a binary variable in the treat equation to calculate lambda, and then regress the outcome\main equation with the endogenous continuous variable and lambda as independents variables. >> >> Is this application appropriate? >> >> Thank you for any help! >> >> Best regards, >> Rose. * *   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/
Open this post in threaded view
|
Report Content as Inappropriate

## Re: Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous?

 In reply to this post by gjhxmu Austin, thank you for your reply. I have ever seen this application, however I am not sure. I want to check it right or wrong. Thank you. An additionally doubt is, treatreg y x1dum x2,treat(x1dum=z1 z2 z3 z4) In the above typing, can it be understood as z1 z2 z3 z4 are instruments for x1dum? Thank you. Best regards, Rose. ----- Original Message ----- From: Austin Nichols <[hidden email]> To: [hidden email] Subject: Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? Date: 2009-8-12 23:22:29 No, that does not sound appropriate. Why would you want to do that? Is it because you have no candidates for excluded instruments and you are hoping to identify off nonlinearities alone? In that case, you would be better off in general ignoring the endogeneity problem than adopting such a strategy. What you should do is find excluded instruments and run -ivreg2- or -ivregress-. 2009/8/12 <[hidden email]>: > Hi Cam, > thank you for your help and valuable reference. > > In fact, the endogenous variable is continuous in my example. > So -ivreg2- is fit for it. My doubt is whether I can use -treatreg- in this way. > Concretely, my outcome equation is , > > y=a0+a1x1+a2x2+e > > where x1 is continuous and endogenous. > > Base on the median of x1, I could get x1dummy=(x1>r(p50)).Then put x1dummy as the > dependent variable in the treat equation to calculate lambda. > And at last, reg y x1 x2 lambda. > > Is is appropriate? > > thank you for any help! > > Best regards, > Rose. > > ----- Original Message ----- > From: Cameron McIntosh <[hidden email]> > To: STATA LIST <[hidden email]> > Subject: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? > Date: 2009-8-12 03:24:14 > > Hi Rose, > > Maybe look into cmp, which deals with mixed mode endogenous variables in simultaneous equation models: > > http://ideas.repec.org/c/boc/bocode/s456882.html> http://www.stata.com/statalist/archive/2007-10/msg00646.html> > Cam > > ---------------------------------------- >> Date: Wed, 12 Aug 2009 01:35:24 +0800 >> From: [hidden email] >> To: [hidden email] >> Subject: st: Can I apply -treatreg- if the endogenous variable is continuous? >> >> I know "A treatment-effects model considers the effect of an endogenously chosen binary treatment on another endogenous continuous variable,conditional on two sets of independent variables". >> >> My doubt is as followings, >> >> If the endogenous variable is continuous, can I apply treatreg? >> Namely, based on mean or median or something else, change the endogenous continuous variable >> as a binary variable in the treat equation to calculate lambda, and then regress the outcome\main equation with the endogenous continuous variable and lambda as independents variables. >> >> Is this application appropriate? >> >> Thank you for any help! >> >> Best regards, >> Rose. * * 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/* *   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/
Open this post in threaded view
|
Report Content as Inappropriate

## Re: Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous?

 To be perfectly clear: using -treatreg- when -ivregress- is the right approach is a bad idea.  Period. clear all prog ftreat, rclass  syntax [,Corr(real .5) z(real .8)]  mat C=(1,0,0,0 \ 0,1,`corr',`z' \ 0,`corr',1,0 \ 0,`z',0,1 )  drawnorm x1 x2 e z, corr(C) n(1000) clear  g y=1+x1+x2+e  reg y x1 x2  ret scalar ols=_b[x2]  ivregress gmm y x1 (x2=z)  ret scalar iv=_b[x2]  su x2, d  g t=x2>r(p50)  treatreg y x1,treat(t=z)  ret scalar treat=_b[t]/normalden(0)/4  eret clear end simul, rep(100) seed(1):ftreat tw kdensity ols||kdensity iv||kdensity treat, xli(1) Even if -treatreg- is the right approach, using -ivregress- does not cost very much: clear all prog ftreat2, rclass  syntax [,Corr(real .5) z(real .8)]  mat C=(1,0,0,0 \ 0,1,`corr',`z' \ 0,`corr',1,0 \ 0,`z',0,1 )  drawnorm x1 x2 e z, corr(C) n(1000) clear  g t=x2>0  g y=1+x1+t+e  reg y x1 t  ret scalar ols=_b[t]  ivregress gmm y x1 (t=z)  ret scalar iv=_b[t]  probit t z  predict that  ivregress gmm y x1 (t=that)  ret scalar ivt=_b[t]  treatreg y x1,treat(t=z)  ret scalar treat=_b[t]  eret clear end simul, rep(100) seed(1):ftreat2 tw kdensity ols||kdensity iv||kdensity ivt||kdensity treat, xli(1) 2009/8/12  <[hidden email]>: > Austin, thank you for your reply. > > I have ever seen this application, however I am not sure. I want to check it right or wrong. Thank you. > An additionally doubt is, > > treatreg y x1dum x2,treat(x1dum=z1 z2 z3 z4) > > In the above typing, can it be understood as z1 z2 z3 z4 are instruments for x1dum? > > Thank you. > > Best regards, > > Rose. > > > ----- Original Message ----- > From: Austin Nichols <[hidden email]> > To: [hidden email] > Subject: Re: RE: st: Can I apply -treatreg- if the endogenous variable iscontinuous? > Date: 2009-8-12 23:22:29 > > No, that does not sound appropriate. Why would you want to do that? > Is it because you have no candidates for excluded instruments and you > are hoping to identify off nonlinearities alone? In that case, you > would be better off in general ignoring the endogeneity problem than > adopting such a strategy. What you should do is find excluded > instruments and run -ivreg2- or -ivregress-. * *   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/