[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
PSPP-BUG: [bug #51146] Regression: r correlation coefficient sign error
|
From: |
John Darrington |
|
Subject: |
PSPP-BUG: [bug #51146] Regression: r correlation coefficient sign error |
|
Date: |
Wed, 31 May 2017 02:27:03 -0400 (EDT) |
|
User-agent: |
Mozilla/5.0 (Windows NT 6.1; rv:45.0) Gecko/20100101 Firefox/45.0 |
Follow-up Comment #1, bug #51146 (project pspp):
I think you are referring to the "R" in the left hand column of the "Model
Summary" box. This is NOT the Pearson Correlation Coefficient.
It is nothing other than the square root of "R Square" which is a measure of
how much variance in the model is explained by the predictor variables. Some
authors call this "Multiple R^2".
In the special case of *simple* linear regression (when there is only one
predictor variable) R does indeed have the same magnitude as that of the
Pearson correlation coefficient, but it should not be regarded as the same
value.
The coefficent "Beta" in your example is correctly given as -0.76
(-0.7556 if you increase the number of decimals in the output).
_______________________________________________________
Reply to this item at:
<http://savannah.gnu.org/bugs/?51146>
_______________________________________________
Message sent via/by Savannah
http://savannah.gnu.org/