22 Mar
2007
22 Mar
'07
7:43 a.m.
calculating R^2 and fit statistics in general doesn't make sense for
comparing relogit and logit. Logit will fit better, but relogit will be
less biased. In almost all applications its bias rather than fit that
we're shooting for. its the same for comparing OLS and WLS by the way:
WLS will never fit better than OLS (which after all is _least_ squares)
but if done properly WLS will have better statistical properties.
Best,
Gary
On Thu, 22 Mar 2007, Stijn Bruyneel wrote:
Dear authors of RELOGIT,
I am a Phd student who is currently investigating entrepreneurship. I make use
of a dataset in which only few events are present. To investigate the impact of
rare events on my logistic results, I make use of your RELOGIT package written
for STATA.
In my work I try to compare standard logistic results with the rare events
results in order to evaluate the necessity of using the RELOGIT package. In
order to make a good comparison, I would like to generate R² in the relogit
regression. Does the RELOGIT package supply such modality?
If this is not the case, and I should want to calculate R² myself, is there
anything specific I need to take into account when accurately calculating R²?
Thanks for your advice,
Stijn Bruyneel
Ph.D. student
Hasselt University
Department of Business Administration
Faculty of Applied Economics
Agoralaan - building D
3590 Diepenbeek (Belgium)
tel: +32 11 26 86 38
fax: +32 11 26 87 00
e-mail: stijn.bruyneel(a)uhasselt.be
23 Mar
23 Mar
3:17 a.m.
New subject: RELOGIT package
Gary,
Although you are right when you say R² is not a proper mechanism to compare
logit and relogit, it might still be worthwhile to generate R² of the relogit
regression. When stating that the fit of relogit does less well than the simple
logit, I would like to have an idea of how much worse.
Can I just calculate the R² in the standard way, or do I have to take into
account specific features of the relogit model?
Thanks
Stijn
Quoting Gary King <king(a)harvard.edu>du>:
calculating R^2 and fit statistics in general
doesn't make sense for
comparing relogit and logit. Logit will fit better, but relogit will be
less biased. In almost all applications its bias rather than fit that
we're shooting for. its the same for comparing OLS and WLS by the way:
WLS will never fit better than OLS (which after all is _least_ squares)
but if done properly WLS will have better statistical properties.
Best,
Gary
On Thu, 22 Mar 2007, Stijn Bruyneel wrote:
Stijn Bruyneel
Ph.D. student
Hasselt University
Department of Business Administration
Faculty of Applied Economics
Agoralaan - building D
3590 Diepenbeek (Belgium)
tel: +32 11 26 86 38
fax: +32 11 26 87 00
e-mail: stijn.bruyneel(a)uhasselt.be
--
relogit mailing list served by Harvard-MIT Data Center
List Address: relogit(a)lists.gking.harvard.edu
Subscribe/Unsubscribe: http://lists.gking.harvard.edu/?info=relogit
Dear authors of RELOGIT,
I am a Phd student who is currently investigating entrepreneurship. I make
use
of a dataset in which only few events are
present. To investigate the
impact of
rare events on my logistic results, I make use of
your RELOGIT package
written
for STATA.
In my work I try to compare standard logistic results with the rare events
results in order to evaluate the necessity of using the RELOGIT package.
In
order to make a good comparison, I would like to
generate R² in the
relogit
regression. Does the RELOGIT package supply such
modality?
If this is not the case, and I should want to calculate R² myself, is
there
anything specific I need to take into account
when accurately calculating
R²?
>
> Thanks for your advice,
>
> Stijn Bruyneel
> Ph.D. student
> Hasselt University
> Department of Business Administration
> Faculty of Applied Economics
> Agoralaan - building D
> 3590 Diepenbeek (Belgium)
>
> tel: +32 11 26 86 38
> fax: +32 11 26 87 00
> e-mail: stijn.bruyneel(a)uhasselt.be
>
>
8:26 a.m.
New subject: RELOGIT package
but that's the issue. the fit is 'worse' in relogit, just as the fit is
worse in WLS, but 'worse' in these cases means 'better'. so if you
did
out of sample prediction, relogit or wls would do better than logit or
ols, respectively (under the right conditions of course). in WLS, R^2
needn't even be between 0 and 1, which basically means that its not at all
clear what the quantity you're calculating should be. same in relogit.
so set aside some of the data and predict if you want, or do more formal
cross-validation, but i don't think you want R^2, even if we knew what it
was.
and as a side point, you probably don't even want R^2 in regression for
most purposes. see for example: Gary King. "How Not to Lie With
Statistics: Avoiding Common Mistakes in Quantitative Political Science,"
American Journal of Political Science, Vol. 30, No. 3 (August, 1986): Pp.
666-687, copy at http://gking.harvard.edu/files/abs/mist-abs.shtml.
Gary
On Fri, 23 Mar 2007, Stijn Bruyneel wrote:
Gary,
Although you are right when you say R² is not a proper mechanism to compare
logit and relogit, it might still be worthwhile to generate R² of the relogit
regression. When stating that the fit of relogit does less well than the simple
logit, I would like to have an idea of how much worse.
Can I just calculate the R² in the standard way, or do I have to take into
account specific features of the relogit model?
Thanks
Stijn
Quoting Gary King <king(a)harvard.edu>du>:
calculating R^2 and fit statistics in general
doesn't make sense for
comparing relogit and logit. Logit will fit better, but relogit will be
less biased. In almost all applications its bias rather than fit that
we're shooting for. its the same for comparing OLS and WLS by the way:
WLS will never fit better than OLS (which after all is _least_ squares)
but if done properly WLS will have better statistical properties.
Best,
Gary
On Thu, 22 Mar 2007, Stijn Bruyneel wrote:
Stijn Bruyneel
Ph.D. student
Hasselt University
Department of Business Administration
Faculty of Applied Economics
Agoralaan - building D
3590 Diepenbeek (Belgium)
tel: +32 11 26 86 38
fax: +32 11 26 87 00
e-mail: stijn.bruyneel(a)uhasselt.be
Dear authors of RELOGIT,
I am a Phd student who is currently investigating entrepreneurship. I make
use
of a dataset in which only few events are
present. To investigate the
impact of
rare events on my logistic results, I make use of
your RELOGIT package
written
for STATA.
In my work I try to compare standard logistic results with the rare events
results in order to evaluate the necessity of using the RELOGIT package.
In
order to make a good comparison, I would like to
generate R² in the
relogit
regression. Does the RELOGIT package supply such
modality?
If this is not the case, and I should want to calculate R² myself, is
there
anything specific I need to take into account
when accurately calculating
R²?
>
> Thanks for your advice,
>
> Stijn Bruyneel
> Ph.D. student
> Hasselt University
> Department of Business Administration
> Faculty of Applied Economics
> Agoralaan - building D
> 3590 Diepenbeek (Belgium)
>
> tel: +32 11 26 86 38
> fax: +32 11 26 87 00
> e-mail: stijn.bruyneel(a)uhasselt.be
>
>
6268
days inactive
6269
days old
relogit@lists.gking.harvard.edu
2 comments
2 participants
participants (2)
-
Gary King -
Stijn Bruyneel