Re: R-squared in regression

On 3 Kwi, 02:25, Richard Ulrich <Rich.Ulr...@xxxxxxxxxxx> wrote:
On Wed, 2 Apr 2008 02:53:49 -0700 (PDT), Stats Wolf

<stats.w...@xxxxxxxxx> wrote:
On 2 Kwi, 08:41, Ray Koopman <koop...@xxxxxx> wrote:
On Apr 1, 12:44 pm, Stats Wolf <stats.w...@xxxxxxxxx> wrote:

[snip]



I don't see any particular reason for R in multiple and r in simple
regression. I would rather say it should be either R or r in each
case. I don't also think that the regression-correlation distinction
is relevant, but if someone wants to convince me this is r and not R
what should be used, I would base also on this argument. All too often
r as a sqrt(r2) in regression is misunderstood as the correlation
coefficient. Hence the distinction does become (at least slightly)
relevant.

That's not really a misunderstanding, unless they are
implying something unusual.  The r of the prediction
equation with the criterion is the sqrt(R^2) .

Well, we're entering another topic, but I can't fully agree. Pearson's
correlation coefficient, which we're talking about here, is used for
two normal variables, while one of the regression's assumptions is
that the independent variable be a fixed one, not random. Hence, if we
want to be strict, we should not consider Pearson's correlation
coefficient correct in regression. IF we want to be strict, though.


--
Rich Ulrich

http://www.pitt.edu/~wpilib/index.html

.

Relevant Pages

  • Re: interpreting sign of t-value
    ... regression coefficient is equal to the hypothesized value. ... A negative t-value indicates the regression coefficient is LESS THAN the hypothesized value. ... Don't use t-test values to tell you about correlation. ...
    (sci.stat.math)
  • Re: Introducing a 3rd variable makes the correlation coefficient non-significant. Whats happenin
    ... This would be SIMPLE correlation coefficient between A and B. ... one independent variable -- that the regression coefficients are NO ... of http://groups.google.com and search for "Reef Fish" as author, ...
    (sci.stat.consult)
  • Re: R-squared in regression
    ... written with lowercase "R", and NOT the uppercase. ... and for example in "Applied Regression Analysis. ... I don't also think that the regression-correlation distinction ...
    (sci.stat.consult)
  • Re: r-Squared Question
    ... >I am trying to undertand r-squared (the coefficient of determination) ... > of regression lines. ... > the correlation between the predicted Y and the observed Y, ...
    (sci.stat.math)
  • r-Squared Question
    ... I am trying to undertand r-squared (the coefficient of determination) ... of regression lines. ... the correlation between the predicted Y and the observed Y, ...
    (sci.stat.math)
Loading