|
|  |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: RE: pooled ols interpretation, thanks
I don't want to get too deep into semantics here, but perhaps you should
have included the companion statement that no change in education results in
an 8% increase in family income in 1991, to make it clear that it is not
actually the one year change in education that is making the 20% increase in
income (like your statement says), it is still only making a 12% increase.
Maybe we should just agree to disagree.
Michael
----- Original Message -----
From: "White, Justin" <[email protected]>
To: <[email protected]>
Sent: Friday, January 19, 2007 2:23 PM
Subject: RE: st: RE: pooled ols interpretation, thanks
What I am saying is "If you are saying that is the impact of a year of
education plus the impact of the year being 1991"
That is why I made reference to the year in my statement.
Justin White
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Michael
Blasnik
Sent: Friday, January 19, 2007 2:14 PM
To: [email protected]
Subject: Re: st: RE: pooled ols interpretation, thanks
Here is your last sentence:
Without knowing the constant, you could say that an increase in 1 year
of
education in 1991 results in a (0.12+0.08 = 0.20) 20% increase in family
income.
You have added together the coefficients for a year dummy and education
and
claimed that this is the marginal effect of a year of education. That
is
incorrect. If you are saying that is the impact of a year of education
plus
the impact of the year being 1991 rather than the omitted reference
year,
then you are correct, but that is not how I read that statement (and
what
would the constant have to do with any of this anyway?).
Michael Blasnik
----- Original Message -----
From: "White, Justin" <[email protected]>
To: <[email protected]>
Sent: Friday, January 19, 2007 2:06 PM
Subject: RE: st: RE: pooled ols interpretation, thanks
>You are confusing the point estimate and the marginal effect.
I don't believe I am made a mistake. I stated in my message that "The
years only come into play if you wanted to calculate a point estimate
for the growth in family income"
Justin White
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Michael
Blasnik
Sent: Friday, January 19, 2007 1:59 PM
To: [email protected]
Subject: Re: st: RE: pooled ols interpretation, thanks
This is incorrect. The marginal effect of education would still be
estimated as the coefficient on education regardless of the values of
the
other variables (there are no interaction terms). You are confusing
the
point estimate and the marginal effect.
Please refrain from posting if you aren't fairly certain of your
answer.
I
also think the list may be indulging too much in answering very
elementary
statistics questions which have nothing to do with Stata directly and
can be
answered through many other available resources other than Statalist.
Michael Blasnik
----- Original Message -----
From: "White, Justin" <[email protected]>
To: <[email protected]>
Sent: Friday, January 19, 2007 1:27 PM
Subject: RE: st: RE: pooled ols interpretation, thanks
Take the estimated coefficients for education, d91, and the constant
and
add them together. You can do this b/c all of you independent
variables
are regressed in levels. Without knowing the constant, you could say
that an increase in 1 year of education in 1991 results in a
(0.12+0.08
= 0.20) 20% increase in family income.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
