Actually, 'ols regression' is NOT a model. Sorry, that's 3 point off for you, Oleg! :-) A regression is a conditional mean E(y|X), which may be linear or it may not (e.g., probit). OLS is one of several possible estimators of a linear regression model. The model must also include other information about the DGP as in how the variables are generated (both y (unit root?) and x (fixed?, random?, jointly determined?) as well as properties of what is excluded from the model, errors. Software knows nothing of this. That is for the user to consider upon prior to choosing an estimator.
Students aren't being stubborn, its me who is stubborn! And cranky .... ! Cheers, Lee On Wed, Oct 14, 2015 at 7:38 AM, <oleg_komashko(a)ukr.net> wrote: > It's interesting how different languages > speakers treat "model". > > My own language has no articles and I > perceive "model" as "a model". > Of course, "ols" is not a model, but > "ols regression" is. > > I think it's better to ask Gretl team rather than > changing menus every time an instructor > meets a stubborn student. > > For, example, currently I don't know the logic > behind placing "arima..... --x-12-arima" under "model" > and X-12-ARIMA under "variable". > > But, certainly, Gretl team DO know, and, having the time, > could inform me. Then I could pass the answer to "children". > > > Oleh > > > > 14 жовтня 2015, 15:00:13, від "Riccardo (Jack) Lucchetti" < > r.lucchetti(a)univpm.it>: > > > On Wed, 14 Oct 2015, Ignacio Diaz-Emparanza wrote: > > > > > El 13/10/15 a las 20:35, Allin Cottrell escribió: > > >> Similarly for "Model": the idea is not that you're going to see a > list of > > >> models as such, it's "Come here for operations relating to models; > come > > >> here if you want to model something." > > >> > > > It was a very interesting debate. This was not correctly translated > into > > > Spanish. I will use "Modelizar" instead of "Modelo" from now on. > > > > I'm going to keep "Modello" in Italian, as shorthand for "Specificazione > e > > stima di un modello", unless my fellow mother-tongue speakers convince me > > otherwise. > > > > ------------------------------------------------------- > > Riccardo (Jack) Lucchetti > > Dipartimento di Scienze Economiche e Sociali (DiSES) > > > > Università Politecnica delle Marche > > (formerly known as Università di Ancona) > > > > r.lucchetti(a)univpm.it > > http://www2.econ.univpm.it/servizi/hpp/lucchetti > > ------------------------------------------------------- > > > > > > _______________________________________________ > > Gretl-devel mailing list > > Gretl-devel(a)lists.wfu.edu > > http://lists.wfu.edu/mailman/listinfo/gretl-devel > > > _______________________________________________ > Gretl-devel mailing list > Gretl-devel(a)lists.wfu.edu > http://lists.wfu.edu/mailman/listinfo/gretl-devel > -- Lee Adkins Professor of Economics and Department Head Economics and Legal Studies in Business Oklahoma State University Email: lee.adkins(a)okstate.edu Fax: 405-744-5180 Phone: 405-744-5196 URL: www.learneconometrics.com
Actually, 'ols regression' is NOT a model.   Sorry, that's 3 point off for you, Oleg!  :-)   A regression is a conditional mean E(y|X), which may be linear or it may not (e.g., probit). OLS is one of several possible estimators of a linear regression model. The model must also include other information about the DGP as in how the variables are generated (both y (unit root?) and x (fixed?, random?, jointly determined?) as well as properties of what is excluded from the model, errors. Software knows nothing of this. That is for the user to consider upon prior to choosing an estimator. Â
Students aren't being stubborn, its me who is stubborn! And cranky ....  !
Cheers,
Lee Â
On Wed, Oct 14, 2015 at 7:38 AM, <[email protected]> wrote:
It's interesting how different languages
speakers treat "model".
My own language has no articles and I
 perceive "model" as "a model".
Of course, "ols" is not a model, but
"ols regression" is.
I think it's better to ask Gretl team rather than
changing menus every time an instructor
meets a stubborn student.
For, example, currently I don't know the logic
behind placing "arima..... --x-12-arima" under "model"
and X-12-ARIMA under "variable".
But, certainly, Gretl team DO know, and, having the time,
could inform me. Then I could pass the answer to "children".
Oleh
14 жовÑÐ½Ñ 2015, 15:00:13, вÑд "Riccardo (Jack) Lucchetti" <[email protected]>:
>Â Â On Wed, 14 Oct 2015, Ignacio Diaz-Emparanza wrote:
>
> > El 13/10/15 a las 20:35, Allin Cottrell escribió:
> >> Similarly for "Model": the idea is not that you're going to see a list of
> >> models as such, it's "Come here for operations relating to models; come
> >> here if you want to model something."
> >>
> > It was a very interesting debate. This was not correctly translated into
> > Spanish. I will use "Modelizar" instead of "Modelo" from now on.
>
> I'm going to keep "Modello" in Italian, as shorthand for "Specificazione e
> stima di un modello", unless my fellow mother-tongue speakers convince me
> otherwise.
>
> -------------------------------------------------------
>Â Â Riccardo (Jack) Lucchetti
>Â Â Dipartimento di Scienze Economiche e Sociali (DiSES)
>
>  Università Politecnica delle Marche
>  (formerly known as Università di Ancona)
>
>Â Â [email protected]
>Â Â http://www2.econ.univpm.it/servizi/hpp/lucchetti
> -------------------------------------------------------
>
>
>Â Â _______________________________________________
> Gretl-devel mailing list
> [email protected]
> http://lists.wfu.edu/mailman/listinfo/gretl-devel
_______________________________________________
Gretl-devel mailing list
[email protected]
http://lists.wfu.edu/mailman/listinfo/gretl-devel
Lee Adkins
Professor of Economics and Department Head
Economics and Legal Studies in Business
Oklahoma State University
Email:Â [email protected]
Fax: 405-744-5180Professor of Economics and Department Head
Economics and Legal Studies in Business
Oklahoma State University
Email:Â [email protected]
