[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Re: adjusted r square
At 10:34 AM 2/20/2007, Ulrich Kohler wrote:
It is just a simple fact that I can explain more variance with an
One other point: It is true that in a sample, R^2 will always go up
as you add more variables. This is because, even if the population
coefficient is zero, sampling variability will cause the estimated
coefficient to be non-zero. But, if we had the entire population,
there would be no requirement that adding another variable result in
more explained variance.
By way of analogy, suppose we had an estimation procedure that we
knew produced coefficients with an upward bias. Would we complain
that we were being "punished" if a correction was applied that
reduced or eliminated that upward bias?
Incidentally, I'm assuming adjusted R^2 does what it purports to
do. Its formula is something I've always just taken on blind
faith. But regardless of how well it works, I think the rationale
for it makes sense.
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
* For searches and help try: