Endogeneity with an exponential regression function - Revision history

Wikiacc: retarget following merge

2020-12-11T05:15:26Z

retarget following merge

← Previous revision		Revision as of 05:15, 11 December 2020
Line 1:		Line 1:
	#REDIRECT [[~~Instrumental~~ ~~variables~~ ~~estimation~~#Endogeneity ~~with~~ ~~an exponential~~ regression ~~function~~]]		#REDIRECT [[Control function (econometrics)#Endogeneity in Poisson regression]]

	{{R from merge}}		{{R from merge}}

Klbrain: Merged content to Instrumental variables estimation#Endogeneity with an exponential regression function, redirecting (easy-merge)

2020-07-06T07:05:17Z

Merged content to Instrumental variables estimation#Endogeneity with an exponential regression function, redirecting (easy-merge)

← Previous revision		Revision as of 07:05, 6 July 2020
Line 1:		Line 1:
	~~{{merge~~ ~~to\|Instrumental variables estimation\|discuss=Talk:~~Instrumental variables estimation#Endogeneity with an exponential regression function~~\|date=May 2019}}~~		#REDIRECT [[Instrumental variables estimation#Endogeneity with an exponential regression function]]
	{{notability\|date=January 2018}}
	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

			{{R from merge}}
	Assume the following exponential regression model, where <math>a_i</math> is an unobserved term in the latent variable. We allow for correlation between <math>a_i</math> and <math>x_i</math> (implying <math>x_i</math> is possibly endogenous), but allow for no such correlation between <math>a_i</math> and <math>z_i</math>.
			{{R to section}}

	: <math>\operatorname E[y_i \nu x_i, z_i, a_i] = \exp(x_i b_0 + z_i c_0+a_i) </math> (1)

	The variables <math>z_i</math> serve as [[instrumental variable]]s for the potentially endogenous <math>x_i</math>. One can assume a linear relationship between these two variables or alternatively project the endogenous variable <math>x_i</math> onto the instruments to get the following reduced form equation:

	: <math>x_i=z_i\Pi+v_i</math> (2)

	The usual rank condition is needed to ensure identification. The endogeneity is then modeled in the following way, where <math>\rho</math> determines the severity of endogeneity and <math>v_i</math> is assumed to be independent of <math>e_i</math>.

	: <math>a_i=v_i \rho+e_i</math> (3)

	Imposing these assumptions, assuming the models are correctly specified, and normalizing <math>\operatorname E[\exp(e_i)]=1,</math>, we can rewrite the conditional mean as follows:

	: <math> \operatorname E[y_i v x_i, z_i , a_i] = \exp (x_i b_0 + z_i c_0 +e_i\rho)</math> (4)

	If <math>e_i</math> were known at this point, it would be possible to estimate the relevant parameters by [[quasi-maximum likelihood estimation]]. Following the two step procedure strategies, Wooldridge and Terza propose estimating equation [2] by standard [[Ordinary least squares\|OLS methods]]. The fitted residuals from this regression can then be plugged into the estimating equation [4] and QMLE methods will lead to consistent estimators of the parameters of interest. Significance tests on <math>\hat\rho</math> can then be used to test for endogeneity within the model.

	The methodology proposed here is often used for exponential regression functions. However, the specific assumptions that need to be made can differ across models. Binary response models impose distributional assumptions on ''y''<sub>''i''</sub> and ''x''<sub>''i''</sub>, whereas this model imposed independence between <math>v_i</math> and <math>e_i</math>.

	== See also ==
	* [[Binary response model with continuous endogenous explanatory variables]]
	* [[Endogeneity in multinomial response model]]

	== References ==
	{{Reflist}}

	== Bibliography ==
	* Wooldridge, J. (1997): Quasi-Likelihood Methods for Count Data, Handbook of Applied Econometrics, Volume 2, ed. M. H. Pesaran and P. Schmidt, Oxford, Blackwell, pp. 352–406
	* Terza, J. V. (1998): "Estimating Count Models with Endogenous Switching: Sample Selection and Endogenous Treatment Effects." ''Journal of Econometrics'' (84), pp. 129–154
	* Wooldridge, J. (2002): "Econometric Analysis of Cross Section and Panel Data", ''MIT Press'', Cambridge, Massachusetts.

	[[Category:Poisson distribution]]		[[Category:Poisson distribution]]

AnomieBOT: Dating maintenance tags: {{Merge to}}

2019-05-22T15:20:09Z

Dating maintenance tags: {{Merge to}}

← Previous revision		Revision as of 15:20, 22 May 2019
Line 1:		Line 1:
	{{merge to\|Instrumental variables estimation\|discuss=Talk:Instrumental variables estimation#Endogeneity with an exponential regression function}}		{{merge to\|Instrumental variables estimation\|discuss=Talk:Instrumental variables estimation#Endogeneity with an exponential regression function\|date=May 2019}}
	{{notability\|date=January 2018}}		{{notability\|date=January 2018}}
	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

Wikiacc: Propose merge to instrumental variables estimation

2019-05-22T14:58:08Z

Propose merge to instrumental variables estimation

← Previous revision		Revision as of 14:58, 22 May 2019
Line 1:		Line 1:
			{{merge to\|Instrumental variables estimation\|discuss=Talk:Instrumental variables estimation#Endogeneity with an exponential regression function}}
	{{notability\|date=January 2018}}		{{notability\|date=January 2018}}

	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

Line 9:		Line 9:
	The variables <math>z_i</math> serve as [[instrumental variable]]s for the potentially endogenous <math>x_i</math>. One can assume a linear relationship between these two variables or alternatively project the endogenous variable <math>x_i</math> onto the instruments to get the following reduced form equation:		The variables <math>z_i</math> serve as [[instrumental variable]]s for the potentially endogenous <math>x_i</math>. One can assume a linear relationship between these two variables or alternatively project the endogenous variable <math>x_i</math> onto the instruments to get the following reduced form equation:

	: <math>x_i=z_i\~~prod~~+v_i</math> (2)		: <math>x_i=z_i\Pi+v_i</math> (2)

	The usual rank condition is needed to ensure identification. The endogeneity is then modeled in the following way, where <math>\rho</math> determines the severity of endogeneity and <math>v_i</math> is assumed to be independent of <math>e_i</math>.		The usual rank condition is needed to ensure identification. The endogeneity is then modeled in the following way, where <math>\rho</math> determines the severity of endogeneity and <math>v_i</math> is assumed to be independent of <math>e_i</math>.

LilHelpa: typo; mnr copy edit

2018-02-28T19:23:17Z

typo; mnr copy edit

← Previous revision		Revision as of 19:23, 28 February 2018
Line 1:		Line 1:
	{{notability\|date=January 2018}}		{{notability\|date=January 2018}}

	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a ~~methodologies~~ to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

	Assume the following exponential regression model, where <math>a_i</math> is an unobserved term in the latent variable. We allow for correlation between <math>a_i</math> and <math>x_i</math> (implying <math>x_i</math> is possibly endogenous), but allow for no such correlation between <math>a_i</math> and <math>z_i</math>.		Assume the following exponential regression model, where <math>a_i</math> is an unobserved term in the latent variable. We allow for correlation between <math>a_i</math> and <math>x_i</math> (implying <math>x_i</math> is possibly endogenous), but allow for no such correlation between <math>a_i</math> and <math>z_i</math>.
Line 30:		Line 30:
	{{Reflist}}		{{Reflist}}

	== ~~Bibliograhpy~~ ==		== Bibliography ==
	* Wooldridge, J. (1997): Quasi-Likelihood Methods for Count Data, Handbook of Applied Econometrics, Volume 2, ed. M. H. Pesaran and P. Schmidt, Oxford, Blackwell, pp. 352–406		* Wooldridge, J. (1997): Quasi-Likelihood Methods for Count Data, Handbook of Applied Econometrics, Volume 2, ed. M. H. Pesaran and P. Schmidt, Oxford, Blackwell, pp. 352–406
	* Terza, J. V. (1998): "Estimating Count Models with Endogenous Switching: Sample Selection and Endogenous Treatment Effects." ''Journal of Econometrics'' (84), pp. 129–154		* Terza, J. V. (1998): "Estimating Count Models with Endogenous Switching: Sample Selection and Endogenous Treatment Effects." ''Journal of Econometrics'' (84), pp. 129–154

Insertcleverphrasehere: Added tags to the page using Page Curation (notability)

2018-01-14T18:23:23Z

Added tags to the page using Page Curation (notability)

← Previous revision		Revision as of 18:23, 14 January 2018
Line 1:		Line 1:
			{{notability\|date=January 2018}}

	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodologies to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodologies to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

86.158.194.215: deleted a duplicate "between"

2018-01-05T20:31:13Z

deleted a duplicate "between"

← Previous revision		Revision as of 20:31, 5 January 2018
Line 1:		Line 1:
	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodologies to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodologies to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

	Assume the following exponential regression model, where <math>a_i</math> is an unobserved term in the latent variable. We allow for correlation between ~~between~~ <math>a_i</math> and <math>x_i</math> (implying <math>x_i</math> is possibly endogenous), but allow for no such correlation between <math>a_i</math> and <math>z_i</math>.		Assume the following exponential regression model, where <math>a_i</math> is an unobserved term in the latent variable. We allow for correlation between <math>a_i</math> and <math>x_i</math> (implying <math>x_i</math> is possibly endogenous), but allow for no such correlation between <math>a_i</math> and <math>z_i</math>.

	: <math>\operatorname E[y_i \nu x_i, z_i, a_i] = \exp(x_i b_0 + z_i c_0+a_i) </math> (1)		: <math>\operatorname E[y_i \nu x_i, z_i, a_i] = \exp(x_i b_0 + z_i c_0+a_i) </math> (1)

Michael Hardy at 18:20, 26 December 2017

2017-12-26T18:20:18Z

← Previous revision		Revision as of 18:20, 26 December 2017
Line 1:		Line 1:
	~~Exponential~~ regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodologies to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodologies to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).

	Assume the following exponential regression model, where <math>~~a_{i}~~</math> is an unobserved term in the latent variable. We allow for correlation between between <math>~~a_{i}~~</math> and <math>~~x_{i}~~</math> (implying <math>~~x_{i}~~</math> is possibly endogenous), but allow for no such correlation between <math>~~a_{i}~~ </math> and <math>~~z_{i}~~</math>.		Assume the following exponential regression model, where <math>a_i</math> is an unobserved term in the latent variable. We allow for correlation between between <math>a_i</math> and <math>x_i</math> (implying <math>x_i</math> is possibly endogenous), but allow for no such correlation between <math>a_i</math> and <math>z_i</math>.

	<math>E[~~y_{i}~~ \nu ~~x_{i~~,} ~~z_{i}~~, ~~a _{i}~~] =exp(~~x_{i}b_{0}~~+~~z_{i}c_{0}~~+~~a_{i}~~) </math> (1)		: <math>\operatorname E[y_i \nu x_i, z_i, a_i] = \exp(x_i b_0 + z_i c_0+a_i) </math> (1)

	The variables <math>~~z_{i}~~</math> serve as instrumental ~~variables~~ for the potentially endogenous <math>~~x_{i}~~</math>. One can assume a linear relationship between these two variables or alternatively project the endogenous variable <math>~~x_{i}~~</math> onto the instruments to get the following reduced form equation:		The variables <math>z_i</math> serve as [[instrumental variable]]s for the potentially endogenous <math>x_i</math>. One can assume a linear relationship between these two variables or alternatively project the endogenous variable <math>x_i</math> onto the instruments to get the following reduced form equation:

	<math>~~x_{i}~~=~~z_{i}~~\prod+~~v_{i}~~</math> (2)		: <math>x_i=z_i\prod+v_i</math> (2)

	The usual rank condition is needed to ensure identification. The endogeneity is then modeled in the following way, where <math>\rho</math> determines the severity of endogeneity and <math>~~v_{i}~~</math> is assumed to be independent of <math>~~e_{i}~~</math>.		The usual rank condition is needed to ensure identification. The endogeneity is then modeled in the following way, where <math>\rho</math> determines the severity of endogeneity and <math>v_i</math> is assumed to be independent of <math>e_i</math>.

	<math>~~a_{i}~~=~~v_{i}~~\rho+~~e_{i}~~</math> (3)		: <math>a_i=v_i \rho+e_i</math> (3)

	Imposing these assumptions, assuming the models are correctly specified, and normalizing <math>E[exp(~~e_{i}~~)]=1,</math>, we can rewrite the conditional mean as follows:		Imposing these assumptions, assuming the models are correctly specified, and normalizing <math>\operatorname E[\exp(e_i)]=1,</math>, we can rewrite the conditional mean as follows:

	<math>E[~~y_{i}~~ v ~~x_{i}~~, ~~z_{i}~~ , ~~a_{i}~~] =exp (~~x_{i}~~ ~~b_{0}~~ + ~~z_{i}~~ ~~c_{0}~~ +~~e_{i}~~\rho)</math> (4)		: <math> \operatorname E[y_i v x_i, z_i , a_i] = \exp (x_i b_0 + z_i c_0 +e_i\rho)</math> (4)

	If <math>~~e_{i}~~</math> were known at this point, it would be possible to estimate the relevant parameters by [[quasi-maximum likelihood estimation]]. Following the two step procedure strategies, Wooldridge and Terza propose estimating equation [2] by standard [[Ordinary least squares\|OLS methods]]. The fitted residuals from this regression can then be plugged into the estimating equation [4] and QMLE methods will lead to consistent estimators of the parameters of interest. Significance tests on <math>\hat\rho</math> can then be used to test for endogeneity within the model.		If <math>e_i</math> were known at this point, it would be possible to estimate the relevant parameters by [[quasi-maximum likelihood estimation]]. Following the two step procedure strategies, Wooldridge and Terza propose estimating equation [2] by standard [[Ordinary least squares\|OLS methods]]. The fitted residuals from this regression can then be plugged into the estimating equation [4] and QMLE methods will lead to consistent estimators of the parameters of interest. Significance tests on <math>\hat\rho</math> can then be used to test for endogeneity within the model.

	The methodology proposed here is often used for exponential regression functions. However, the specific assumptions that need to be made can differ across models. Binary response models impose distributional assumptions on yi and xi, whereas this model imposed independence between <math>~~v_{i}~~</math> and <math>~~e_{i}~~</math>.		The methodology proposed here is often used for exponential regression functions. However, the specific assumptions that need to be made can differ across models. Binary response models impose distributional assumptions on ''y''<sub>''i''</sub> and ''x''<sub>''i''</sub>, whereas this model imposed independence between <math>v_i</math> and <math>e_i</math>.

	== See also ==		== See also ==
Line 29:		Line 29:

	== Bibliograhpy ==		== Bibliograhpy ==
	* Wooldridge, J. (1997): Quasi-Likelihood Methods for Count Data, Handbook of Applied Econometrics, Volume 2, ed. M. H. Pesaran and P. Schmidt, Oxford, Blackwell, pp. ~~352-406~~		* Wooldridge, J. (1997): Quasi-Likelihood Methods for Count Data, Handbook of Applied Econometrics, Volume 2, ed. M. H. Pesaran and P. Schmidt, Oxford, Blackwell, pp. 352–406
	* Terza, J. V. (1998): Estimating Count Models with Endogenous Switching: Sample Selection and Endogenous Treatment Effects. Journal of Econometrics (84), pp. ~~129-154~~		* Terza, J. V. (1998): "Estimating Count Models with Endogenous Switching: Sample Selection and Endogenous Treatment Effects." ''Journal of Econometrics'' (84), pp. 129–154
	* Wooldridge, J. (2002): Econometric Analysis of Cross Section and Panel Data, MIT Press, Cambridge, ~~Mass~~.		* Wooldridge, J. (2002): "Econometric Analysis of Cross Section and Panel Data", ''MIT Press'', Cambridge, Massachusetts.

	[[Category:Poisson distribution]]		[[Category:Poisson distribution]]

Uanfala: typo

2017-12-26T10:25:27Z

typo

← Previous revision		Revision as of 10:25, 26 December 2017
Line 17:		Line 17:
	<math>E[y_{i} v x_{i}, z_{i} , a_{i}] =exp (x_{i} b_{0} + z_{i} c_{0} +e_{i}\rho)</math> (4)		<math>E[y_{i} v x_{i}, z_{i} , a_{i}] =exp (x_{i} b_{0} + z_{i} c_{0} +e_{i}\rho)</math> (4)

	If <math>e_{i}</math> were known at this point, it would be possible to estimate the relevant parameters by [[quasi-maximum likelihood ~~estimatation~~]]. Following the two step procedure strategies, Wooldridge and Terza propose estimating equation [2] by standard [[Ordinary least squares\|OLS methods]]. The fitted residuals from this regression can then be plugged into the estimating equation [4] and QMLE methods will lead to consistent estimators of the parameters of interest. Significance tests on <math>\hat\rho</math> can then be used to test for endogeneity within the model.		If <math>e_{i}</math> were known at this point, it would be possible to estimate the relevant parameters by [[quasi-maximum likelihood estimation]]. Following the two step procedure strategies, Wooldridge and Terza propose estimating equation [2] by standard [[Ordinary least squares\|OLS methods]]. The fitted residuals from this regression can then be plugged into the estimating equation [4] and QMLE methods will lead to consistent estimators of the parameters of interest. Significance tests on <math>\hat\rho</math> can then be used to test for endogeneity within the model.

	The methodology proposed here is often used for exponential regression functions. However, the specific assumptions that need to be made can differ across models. Binary response models impose distributional assumptions on yi and xi, whereas this model imposed independence between <math>v_{i}</math> and <math>e_{i}</math>.		The methodology proposed here is often used for exponential regression functions. However, the specific assumptions that need to be made can differ across models. Binary response models impose distributional assumptions on yi and xi, whereas this model imposed independence between <math>v_{i}</math> and <math>e_{i}</math>.

Uanfala: Uanfala moved page Draft:Endogeneity With An Exponential Regression Function to Endogeneity with an exponential regression function

2017-12-25T23:06:04Z

Uanfala moved page Draft:Endogeneity With An Exponential Regression Function to Endogeneity with an exponential regression function

← Previous revision		Revision as of 23:06, 25 December 2017
(No difference)

← Previous revision		Revision as of 15:20, 22 May 2019
Line 1:		Line 1:
	{{merge to\|Instrumental variables estimation\|discuss=Talk:Instrumental variables estimation#Endogeneity with an exponential regression function}}		{{merge to\|Instrumental variables estimation\|discuss=Talk:Instrumental variables estimation#Endogeneity with an exponential regression function\|date=May 2019}}
	{{notability\|date=January 2018}}		{{notability\|date=January 2018}}
	In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).		In statistics as applied to [[econometrics]], exponential regression models constitute a very large and popular class of [[Regression analysis\|regression]] models. Standard [[econometric]] concerns such as [[Endogeneity (econometrics)\|endogeneity]] or [[Omitted-variable bias\|omitted variables]] can be accounted for in a more general framework. Wooldridge and Terza provide a methodology to both deal with and test for endogeneity within the exponential regression framework, which the following discussion follows closely.<ref>Wooldridge 1997; Terza 1998</ref> While the example focuses on a [[Poisson regression]] model, it is possible to generalize the test to other exponential regression models, although this may come at the cost of additional assumptions (e.g. for binary response or censored data models).